THIRD REPORT ON REGULAR VHF IONOSPHERIC PROPAGATION OBSERVABLE OVER LONG DISTANCES

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP81-01043R003000180001-4
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 194
• Document Creation Date: December 23, 2016
• Document Release Date: June 9, 2014
• Sequence Number: 1
• Publication Date: September 26, 1958
• Content Type: REPORT

Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
laiRD REPORT
ON
REGULAR VHF IONOSPHERIC PROPAGATION
OBSERVABLE OVER LONG DISTANCES
FINAL REPORT ON. SIGNAL CORPS INTERDEPARTMENTAL PROCUREMENTS
No. 821-PHIBP-51-04, R54-73-SC-91 and. R56-0002-SC-91
Covering work carried out from
March 1951 to June 1958
U. S. DEPARTMENT OF COMMERCE
NATIONAL BUREAU OF STANDARDS
BOULDER LABORATORIES
Boulder, Colorado
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
?
THE NATIONAL BUREAU OF STANDARDS
Functions and Activities
The functions of the National Bureau of Standards arc set forth in the Act of
Congress, March 3, 1901, as amended by Congress in Public Law 619, 1950. These
include the development and maintenance of the national standards of measurement,
and the provision of means and methods for making measurements consistent with
these standards: the determination of physical constants and properties of materials;
the development of methods and instruments for testing materials, devices, and
structures; advisory services to Government Agencies on scientific and technical
problems; invention and development of devices to serve special needs of the Govern-
ment; and the development of standard practices, codes and specifications. The work
includes basic and applied research, development, engineering, instrumentation, testing,
evaluation, calibration services, and various consultation and information services. A
major portion of the Bureau's work is performed for other Government Agencies, par-
ticularly the Department of Defense and the Atomic. Energy Commission. The scope of
activities is suggested by the listing of divisions and sections on the inside back cover.
Reports and Publications
The results of the Bureau's work take the form of either actual equipment and de-
vices or published papers and reports. Reports are issued to the sponsoring agency of
a particular project or program. Published papers appear either in the Bureau's own
series of publications or in the journals of professional and scientific societies. The
Bureau itself publishes three monthly periodicals, available from the Government Print-
ing Office: The Journal of Research, which presents complete papers reporting techni-
cal investigations; the Technical News Bulletin, which presents summary and preliminary
reports on work in progress; and Basic Radio Propagation Predictions, which provides
data for determining the best frequencies to use for radio communications throughout
the world. There are also five series of nonperiodical publications: The Applied Math-
ematics Series, Circulars, Handbooks, Building Materials and Structures Reports, and
Miscellaneous Publications.
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
NATIONAL BUREAU OF STANDARDS REPORT
NBS PROJECT NBS REPORT
8200-10-8725 $eptember 26, 1958 6014
THIRD REPORT
ON
REGULAR VHF IONOSPHERIC PROPAGATION
OBSERVABLE OVER LONG DISTANCES
Edited.
by
R. C. Kirby
R. M. Davis, Jr.
FINAL REPORT ON SIGNAL CORPS INTERDEPARMENTAL PROCUREMENTS
No. 821-PBIBP-51-04, R54-73-SC-91 and. R56-0002-SC-91
Covering work carried out from
March 1951 to June 1958
U. S. DEPARTMENT OF COMMERCE
NATIONAL BUREAU OF STANDARDS
BOULDER LABORATORIES
Boulder, Colorado
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Cop Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
J
IMPORTANT NOTICE
NATIONAL BUREAU OF STANDARDS REPORTS are usually preliminary or progress accounting docu-
ments intended for use within the Government. Before material in the reports is formally published it
is subjected to additional evaluation and review. For this reason, the publication, reprinting, reproduc-
tion, or open-literature listing of this Report, either in whole or in part, is not authorized unless per-
mission is obtained in writing from the Office of the Director, National Bureau of Standards, Washington
25, D. C. Such permission is not needed, however, by the Government agency for which the Report has
been specifically prepared if that agency wishes to reproduce additional copies for its own use.
I
:I
'4
I
4
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2
1-01043R003000180001-4
Declassified in Part- Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
UNITED STATES ARMY SIGNAL RESEARCH AND DEVELOPMENT LABORATORY
CHIEF OF RESEARCH & DEVELOPMENT
OCS, DEPARTMENT OF THE ARMY
WASHINGTON 25, D.C.
CHIEF SIGNAL OFFICER
DEPARTMENT OF THE ARMY
WASHINGTON 25, D.C.
ATTN: SIGRD
SIGRD-6-f
CONTRACT ISigtgat 0.91
DISTRIBUTION LIST
THIRD AND FINAL REPORT
COMMANDER
ARMED SERV TECH INFO AGCY
1 ARLINGTON HALL STATION
IRLINGTON 12, VIRGINIA
1
1
DIRECTOR.
US NAVAL RESEARCH LABORATORY
WASHINGTON 25, D.C.
ATTN: CODE 2027 1
1
34437
COMMANDING OFFICER & DIRECTOR
US NAVY ELECTRONICS LAB
SAN DIEGO 52, CALIFORNIA
ATTN: LIBRARY
1
COMMANDER
WRIGHT AIR DEVELOPMENT CTR
WRIGHT PATTERSON AFB, OHIO
ATTN: WCOSI -3 2
DR. PAUL L. TAYLOR 1
COMMANDER
AIR FORCE CAMBRIDGE RES CTR
L.G. HANSCOM FIELD
1 BEDFORD, MASSACHUSETTS
ATTN: CROTLR-2
COMMANDER
ROME AIR DEVELOPMENT CENTER
GRIFFISS AFB, NEW YORK
ATTN: RCSSLD
MR. L. COLIN
DIRECTOR OF COMMUNICATIONS
HQS, USAF
WASHINGTON 25, D.C.
ATTN: AFOAC
COMMANDER
USAF SECURITY SERVICE
SAN ANTONIO, TEXAS
ATTN: S-DC
COMMANDING GENERAL
US ARMY ELECTRONIC PROV OR
FORT HUACHUCA, ARIZONA
ATTN: SIGP3-SCGP-1
COMMANDING OFFICER -
US ARMY SIG EQUIP SPT AGCY
FORT MONMOUTH, NEW JERSEY
ATTN: SIGFM/ES-ASA
US NAVY LIAISON OFFICER
DEPARTMENT OF THE NAVY
USASRDL
FORT MONMOUTH, NEW JERSEY
COMMANDING OFFICER
AIR R&D COMMAND, USASRDL
FORT MONMOUTH, NEW JERSEY
COMMANDING OFFICER
US MARINE CORPS LIAISON OFF
CODE A0-4C USASRDL
FORT MONMOUTH, NEW JERSEY
CHIEF SIGNAL OFFICER
COMBAT DIV & OP DIVISION
1 DEPARTMENT OF THE ARMY
WASHINGTON 25, D.C.
ATTN: SIGCO
1
CHIEF SIGNAL OFFICER
PLANS & PROGRAMS DIV
DEPARTMENT OF THE ARMY
WASHINGTON 25, D.C.
ATTN: SIGPL
COMMANDER
1 AIR R&D COMMAND
PO BOX 1395
BALTIMORE, MARYLAND
ATTN: MAJOR A.L1.0 BAKER
1 DEPUTY CHIEF OF STAFF,
INTELLIGENCE
HQS, TACTICAL AIR COMM
LANGLEY AFB, VIRGINIA
CHIEF BUREAU OF SHIPS
C:ODE 835
DEPARTMENT OF THE NAVY
WASHINGTON 25, D.C.
ATTN: MR. R.S. BALDWIN 1
CHIEF SIGNAL OFFICER
ARMY COMM SERV DIVISION
DEPARTMENT OF THE ARMY
10 WASHINGTON 25, D.C.
COMMANDING OFFICER
ARMY SIG COMM ENGRG AGCY
DEPARTMENT OF THE ARMY
3 WASHINGTON 25, D.C.
ATTN: MR. J.N.CRAVER
COMMANDING OFFICER
HQS, ALASKAN AIR COMMAND
1 DEPT OF COMM & ELECTRONICS
APO 942, c/o POSTMASTER
SEATTLE, WASHINGTON
COMMANDING OFFICER
1 ALASKA COMM SYSTEM
550 FEDERAL OFFICE BLDG
SEATTLE 4, WASHINGTON
2
1
1
1
1 PRESIDENT, BOARD #5
CONTINENTAL ARMY COMMAND
FORT BRAGG, NO. CAROLINA 1
SIG C LIAISON OFFICER
1 MIT, LINCOLN LAB, PO BOX 73
LEXINGTON, MASSACHUSETTS
ATTN: MR. W.G. ABEL, LIBRARIAN
DIRECTOR COMM-ELECTRONICS
JOINT CHIEF OF STAFF
1 WASHINGTON 25, D.C.
1
COMMANDER
AIR FORCE MISSILE TEST CTR
PATRICK AFB, FLORIDA
ATTN: TECH LIBRARY
DIRECTOR
ARMANENT LABORATORY
WRIGHT PATTERSON AFB, OHIO
ATTN: MCREIGD
COMMANDING GENERAL
NORTHEAST AIR COMMAND
PEPPERREL AFB, APO 862
c/o POSTMASTER, NEW YORK N.Y.
2 US NAVAL INSPECTOR OF ORDNANCE
APPLIED PHYSICS LABORATORY
JOHN HOPKINS UNIVERSITY
SILVER SPRINGS, MARYLAND
DIRECTOR, NAVAL ORDNANCE LAB
WHITE OAK
SILVER SPRINGS, MARYLAND
ATTN: LIBRARIAN RM1-327 1
1
1
1
1
1
2
Declassified in Part - Sanitized Copy Approved for Release 50-Yr 2014/06/09 ? CIA-RDP81-01043Rorrinnn1 Annni
Declassified in Part - Sanitized Co .y Ap roved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
-10
UNITED STATES ARMY SIGNAL RESEARCH AND DEVELOPMENT LABORATORY
CONTRACT NOS 104-73-SC-91 and R56-0002-SC-91
DISTRIBUTION LIST
THIRD AND FINAL REPORT ( Cpnt'd)
ELECTRONICS RESEARCH LAB
STANFORD UNIVERSITY
STANFORD, CALIFORNIA
ATTN: ASSOCIATE DIRECTOR 1
DIRECTOR, OFFICE OF NAVAL RES
DEPARTMENT OF THE NAY!
1000 GEAR! STREET
SAN FRANCISCO, CALIFORNIA 1
DIRECTOR, CENTRAL INTELLIGENCE
AGENCY
2430 E STREET N.W.
WASHINGTON, D.C. 1
DIRECTOR, NATIONAL SECURITY AGCY
WASHINGTON 25, D.C. 1
ASSISTANT SECRETARY OF DEFENSE
R&D INFORMATION DIVISION
LIBRARY BR, PENTAGON 3D-1041
WASHINGTON 25, D.C. 1
FEDERAL COMM COMMISSION
WASHINGTON 25, D.C.
ATTN: CH, ENGRG TECH RES DIV 1
DIRECTOR REFERENCE DEPARTMENT
TECH INFO DIV, LIBRARY OF CONGRESS
WASHINGTON 25, D.C. 3
CORNELL UNIVERSITY
SCHOOL OF ELECTRICAL ENGRG
ITHACA, NEW YORK
ATTN: DR. H.G. BOOKER 1
USCONARC LIAISON OFFICER
USASRDL
FORT MONMOUTH, NEW JERSEY 1
UNIVERSITY OF ALASKA
COLLEGE, ALASKA
ATTN: DR. C.T. ELVERY, DIR 1
DEPARTMENT OF STATE
WASHINGTON 25, D.C.
ATTN: MR. F.C. deWOLFE
CH, TELECOMM DIV
DIRECTOR US INFO AGENCY
VOICE OF AMERICA
330 INDEPENDENCE AVE S.W.
WASHINGTON 25, D.C.
ATTN: MR. GEORGE JACOBS
SIG C LIAISON OFFICER
OPERATIONS RESEARCH OFFICE
JOHN HOPKINS UNIVERSITY
SILVER SPRINGS, MARYLAND
1
1
1
COMMANDING OFFICER
USA SIG C RADIO PROP AGCY
DEPARTMENT OF THE ARM
FORT MONMOUTH, NEW JERSEY 2
COMMANDING OFFICER
US ARMY SIGNAL R&D LABORATORY
FORT MONMOUTH, NEW JERSEY
ATTN:
DIRECTOR OF RESEARCH
TECH DOC CENTER, EVANS
TECH INFO DIV
EXPLOR DIV, FILE UNIT 6
BLDG 37a EVANS
DIR, EXPLOR RES DIV "S"
DIR, COMM DEPT ATTN MR BROWN
EXPLOR RES DIV "C"
DIR COMM DEPT
DIR, COUNTERMEASURES DIV
DIA, RADAR DIV
ATTN: MR. WOODYARD
DIR, APP PHYS DIV
TECH DOC CTR, HEX
1
1
5
1
5
1
1
1
1
1
1
9
This contract is supervised by Exloratory Research "S" Division, Surveillance Department, US Army
Signal Research and Development Laboratory, Fort Nbnnouth, New Jersey. For further information
contact Mr. Arthur Harris, Project Engineer. Telephone: Liberty 2-4000, Extension 61344.
2
Declassified in Part - Sanitized Co.y Ap?roved for Release ? 50-Yr 2014/06/09 ? CIA-RDP81-01043R00300018flom -4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
1.*
FOREWORD
iii -
CONTENTS
PAGE
vii
1. Introduction - Scope of Report 1
2. The Program 1
2.1 Variations with Season, Geographical Position and
Solar Activity 1
2.2 Frequency Dependence 3
2.3 Scattering Heights, Dependence of Transmission Loss
on Path Geometry 3
2.4 Sporadic-E Observation 3
2.5 'Development of Low-Power Narrow Band. Recording
Equipment 3
3. Results in Detail 4
3.1 Routine Recording Program 4
3.11 Cedar Rapids to Sterling Test Path - 49.8 Mc/s. 4
3.12 Fargo to Churchill Test Path - 49.7 Me/s 5
3.13 Anchorage to Barrow Test Path - 48.87 mc/s. . . ^ 5
3.14 Routine Recording at Other Frequencies 5
a. Anchorage to Barrow - 24.325 Mc/s 5
b. Cedar Rapids to Sterling - 27.775 mc/s. . . ? 5
c. Ceder Rapids to Sterling - 107.8 Mc/s . ? . ? 6
3.15 Seasonal Variation of Signal Intensity 6
3.16 Annual Variation of Signal Intensity;
Correlation with Geomagnetic Observations
and Ionospheric Conditions Observed at High
Frequencies 6
3.17 Sporadic-E Propagation 7
3.2 Experimental and Developmental Phases of the Program. ? 7
3.21 Frequency Dependence ...... . ... . 7
3.22 Scattering Heights, Dependence of Transmission
Loss on Path Geometry ......... ? ? 8
3.23 Signal Fading Characteristics 9
3.24 Doppler-Shifted Meteor Echoes 9
3.25 Development of Low-Power Narrow Band Recording
Equipment for Routine Observations 10
4. Discussion and Conclusions 10
4.1 General . . . ? ........ ..... . . . 10
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
iv
PAGE
v-
APPENDICES
4.2 New Results
11
I.
Part I
Experimental Study of Frequency-Dependence --
a. Scattering Heights
11
R. C. Kirby and R. M. Davis, Jr.
b. Solar Cycle Dependence
11
C. Useful Range of Frequencies
12
Part II
Determination of Frequency Dependence Exponent
d. Meteoric Contribution
13
by True Power Measurement -- G. E. Boggs and
e. Sporadic-E Propagation
13
N. C. Hekimian
4.3 Outstanding Problems
13
II.
Part I
Experimental Investigations of Ionospheric For-
a. Frequency Dependence
14
ward Scattering at VHF using Pulse Techniques --
b. Optimum Antenna Design, Includingleam Slewing
14
V. C. Pineo
c. Modulation Studies
15
d. Geographical Dependence
16
Part II
Summer Heights of VHF Scattering -- V. C. Pineo
ACKNOWLEDGMENTS
17
Part III
Scattering Angle Dependence Tests -- V. C. Pineo
REFERENCES
18
III.
Part I
Correlation of VHF Signal Intensities with Mag-
netic K-Indices -- H. I. Leighton
FIGURES 1 through 40
Part II
Correlation of VHF Scatter Intensities with HIP
Records -- R. M. Davis, Jr.
IV.
Oblique-Incidence VHF Sporadic E Observations --
R..M. Davis, Jr.
?
I
V.
Meteor Whistles on the Cedar Rapids to Sterling
Path -- E. K. Smith
VI.
Part I
Off-Path Measurements of Azimuth of Arrival of
49.8 Mcis Signal -- V. C. Pineo
Part II
EXperimental Observations of the Contribution of
Meteoric Ionization to the Propagation of VHF
Radio Waves by Ionospheric Forward Scatter --
V. C. Pineo
VII.
A Narrow-Band Recording Receiver ? G. F. Montgomery
VIII.
Three Kilowatt VHF Transmitter for Radio Propaga-
tion Studies -- W. B. Harding and D. C. Whittaker
IX.
Stable Frequency Control for Narrow-Band Recording
Receiver and Transmitter -- P. G. Sulzer
Part I
One-Megacycle High Stability Oscillator
Part II
Filter-Type Exciter
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
-vi-
-
IX.
IX. Part III Phase-Locked-Oscillator Exciter
X.
XI.
XII.
Part IV Direct Multiplier Exciter
Solar Cycle Influence on the Lower Ionosphere
and on VHF Forward Scatter -- C. D. Ellyett and
H. I. Leighton
Tabulated Monthly Median and. Decile Values of
Hourly Median Signal Intensity Recorded. in the
Routine Recording Programs
Summary Technical Paper, D. K. Bailey, R. Bateman
and B. C. Kirby, "Radio Transmission at VHF by
Scattering and other Processes in the Lower Ionos-
phere," reprinted (by-permission) from Proc. I.R.E.,
43, October 1955.
?
- vii -
FOREWORD
This is the third and final of a series of reports in which the
results of studies of regular VHF ionospheric propagation (scattering)
are presenttd; it is prepared in fulfillment of the requirement for a
final report on a program of work covered by Signal Corps Interdepart-
mental Procurement Numbers 821-PHIBP-51-041 R54-73-SC-91, R56-0002-SC-
91, and associated amendments, representing financial support from the
Departments of the Army, Air Force and Navy.
The principal additional data obtained from the routine recording
program during the period September, 1952, to December, 1955 are con-
tained herein, together with results of the experimental and develop-
mental phases of the program before their termination in February, 1956,
under the support of the subject MIPR's; work carried out after that
time under the MIPR's was in connection with analysis and reporting of
data. (Beginning in July, 1954, the National Bureau of Standards had
begun direct support of a continuing basic research program on VHF ion-
ospheric scattering, which continues to the present time.)
A summary of most of the important results of the program was pub-
lished as a technical paper in the October, 1955, "Scatter Propagation"
issue of the Proceedings of the I.R.E.; to avoid the necessity for re-
petition or paraphrasing much of the material contained in that paper,
it has been referred to frequently throughout, and reprints have been
included as Appendix: XII in a limited number of copies of this report.
The scope of the present report, therefore, is largely to supplement
that summary paper with detailed technical data which was not included,
as well as to report on the results obtained in limited experimental
work subsequently.
As indicated in previous reports, emphasis throughout the program
was toward obtaining information directly useful for assessing communi-
cation possibilities of the propagation mechanism. It is believed that
the data presented in this and the preceding reports are sufficient,
from the propagation point of view, to provide a basis for system en-
gineering in temperate and high latitudes. Although the principal phe-
nomena associated with ionospheric scatter propagation are now well-
known, many questions still remain unanswered, and the nature of some
of these is suggested in the text.
Consistent with downgrading of the classification of the earlier
two reports of the series, by authority of letter from. U. S. Army Sig-
nal Supply Agency, February 3, 1958, signed by the Contracting Officer,
and approvals for publication in technical journals of summaries of ma-
terial contained herein, the security classificatiOn of this report is
UNCLASSIFIED.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
THIRD REPORT
ON
REGULAR VHF IONOSPHERIC PROPAGATION
OBSERVABLE OVER LONG DISTANCES
Edited
by
R. C. Kirby
R. M. Davis, Jr.
1. Introduction - Scope of Report
The purpose of this report is to provide a formal final report in
connection with the financial support of the work extended by the various
military departments.*
A summAry of most of the important results of the progsrat through
July, 1955, has already been published as a technical paperLand is re-
ferred to frequently throughout this report. Two earlier NBS Reports2,3
gave detailed data obtained from the program through September, 1952.
4 The present report contains the principal additional detailed data
obtained from the routine recording program during the period September,
1952, to December, 1955) as well as further results of the experimental
? and developmental phases of the program before their termination in Feb-
ruary, 1956. The latter are reported in the various appendices.
2. The Program
Emphasis throughout the program was on obtaining information appli-
cable to the design of communication systems; it was also intended to ob-
tain the kind of results necessary for an understanding and development
of the theory of the phytical conditions and processes giving rise to the
observed propagation. The program may be described under the following
headings:
2.1 Variations with Season, Geographical Position and
Solar Activity
Table I should be studied in connection with the program in general
and this aspect in particular. Work under this topic required the regu-
lar operation of three experimental paths at just below 50 NC/s: Cedar
Rapids to Sterling at 49.8 MC/s, Fargo to Churchill at 49.7 NC/s) and
Anchorage to Barrow at 48.87 MC/s. The frequency and transmitter power
were substantially the sane for these three paths. The antenna systems
were all long rhombic antennas (25 wavelengths on each leg), of the sane
nominal design. Probable effects of departures from ideal siting, and
*See foreword.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
TABLE 1
Details of Itcperhaental Paths Ccmprising Routine Recording Program
Cedar Rapids-Sterling
Anchorage -Harrow
Fargo-Churchill
Frequency, Mcla
r's 49.630l
107.e30
27.775
mum
48.e70
24.325
49.700
Dute of commencement
January 23, 1951
December 4,
1951
May 6, 1553
August 28, 1951
Yarch 2,
1953
February 16,
1953
Auguat 29, 1951
Date of termination
June 30, 1958
January 31,
1953
February 29,
1956
June 30, 1953
June 30,
1953
June 30,
1953
March 31, 1953
Transmitter location coordi-
nates of site
Cedar Rapids, Iowa
41?52'N; 91041'W
Same
Same
Anchorage, Alaska
61017'N; 149?42W
Sane
Same
Fargo, North
... Dakota ?.
46w55'N; 96-46'W
Receiver location coordinates
of site
Sterling, Virginia
38?59'N; 77?29'W
Same
Same
Barrow, Alaska
71?18.11; 156?45'W
Same
Same
Churchill
Manitoba
58?44'N; 94?05W
Surface path length (great
circle)
1,243 in.
773 at. mi.
Same
Same
1,156 km.
718 at. ad.
Same
Some
1,326 in.
824 at. ni.
Geographic coordinates of
path midpoint
40?39'N; 84?26'W
Same
Same
66?206; 152?31'W
Same
Some
52?50 'N; 95?56,,
Geomagnetic latitude of path
midpoint
5,?38,N
Same
Same
65?02'N
Same
Same
62?42'N
True azimuth of transmitter
from receiver
289?30'
Sane
Same
160055'
Same
Same
188?56'
True azimuth of receiver
from transmitter
100?17'
Same
Same
347?24'
Sam
Same
6?47'
Type of antennas
Rhombic-Rhombic
Some
Same
Sere
Yagi -Vaal
Same
Rhombic-Rhombic
Elevation of maximum of
main lobe, degrees
7.0
Same
Same
5.72
6.42
6.42
5.4
Height, feet
41
19
73
50
45
90
52
Leg length, feet
500
230
897
500
--
--
500
Estimated plzule-vave gain,
relative to dipole at same
height, decibels
11;
18
18
18
9
9
18
Horizontal beamvidth,
degrees
6
6
6
60
60
60
6
1.0 microvolt received3 signal
intensity on graphn for 30 Ev
antenna pover,-etz....ota
Attenuation relative to
inverse distance (perfect
reflection in ionosphere)db
Transmission loss, db
122
199
115
199
127
199
123
199
105
199
111
199
122
199
149.600 Me/s uncd from January 17, 1552 through March 31, 1952.
2Includes effect of sloping site at Anchorage,
3Received open-circuit antenna voltage measured at 600 ohms impedance (-154 din). Reference power to transmitting antenna is 30 Ku.
I.
- 3 -
the somewhat lower heights of the antennas used on the original installa-
tions at Cedar Rapids and Sterling, are discussed.
2.2 Frequency Dependence
This part of the program involved operation of the Cedar Rapids to
Sterling test path at 107.8 Mcis to January 31, 1953, followed by opera-
tion at 27.775 Eqs to February 29, 1956. From March through June, 1953,
variation of signal intensity with frequency was also investigated for
the arctic path from Anchorage to Barrow by provision of test transmissions
at 24.325 Mc/s.
2.3 Scattering Heights, Dependence of Transmission Loss on
Path Geometry
Heights of scattering from the lower ionosphere were deduced ini-
tially from one-way measurements of the relative transit times of the
tropospheric component and the ionospheric component of a pulse signal
at 49.7 MLcis. The observed scattering heights were compared with simul-
taneous observations of virtual heights of low frequency reflections near
the midpoint of the path.
For path lengths up to about 1,000 Kt the scattering angle is fairly
large (>20?) and depends on path length. The extent to which the trans-
mission loss depends upon this angle was investigated. Round-trip pulse
delay measurements were made finally to obtain heights of ionospheric and
tropospheric scattering.
To obtain information about the possibilities of off-path interfer-
ence in VBF propagation, the signal intensities of the CORY' Rapids trans-
missions at 49.8 Mcis beamed toward Sterling were received at off-path
receiving sites and compared with the signal intensities observed simul-
taneously at Sterling. Off-path observations were also made to investi-
gate experimentally the contribution of meteoric ionization to the ob-
served propagation.
2.4 Sporadic-E Observation
During the early phases of the program occurrences of sporadic-E pro-
pagation at 50 Mc/s were observed and reported. The program was expanded
to include regular observation and analysis of the occurrence of sporadic-
E on all the paths and at all the frequencies used.
2.5 Development of Low-Power Narrow Band. Recording ,Equipment
In order to reduce the power requirement for routine observations of
transmission loss, development of a narrow band recording receiver was
undertaken along with the necessary stable frequency control for a record-
ing receiver and transmitter; a 3 kilowatt VHF transmitter was developed
for unattended operation.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
- 4 -
3. Results in Detail
3.1 Routine Recording Program
Table 1 supplies many details of the various test paths, and frequent
reference to it will be found useful in connection with the following re-
sults. Appendix XX of the second. report3 gave details of the measurements
and methods for reporting the data from the field stations, and indicates
how and why various corrections were made. Appendix VIII of the first re-
port2 described the complete recording setup; and Appendix XV of the se-
cond report3 described a new type of gain-stable converter which was de-
veloped and placed in service. The newly developed narrow band record-
ing systems were not used for the routine recording program during the
period coyered by this report.
Throughout the present report, except where otherwise indicatedl sig-
nal-intensity data are given in decibels above one microvolt received open-
circuit antenna voltage for 600 ohms terminal impedance. Corrections have
been made for any transmission line losses; and all hourly values have
been normalized to the estimated antenna power shown. Available power at
the receiving antenna is E2/4R, so that one microvolt (zero decibel level)
corresponds to an available power of -154 dbw. For the data given in
Figures 1 to 40 in the main text and the tables in Appendix XI, the re-
ference antenna power is 30 kilowatts; the reference power is specified
in the various appendices where it may differ from the standard reference
power. One microvolt (zero decibels) corresponds to a transmission loss
of 199 decibels, or to the values of attenuation relative to inverse dis-
tance (perfect reflection in the ionosphere) given in Table 1; these values
depend on path length.
The signal-intensity results are given in plots and tables showing
for each hourly period the signal levels equAlled or exceeded 10%, 50%
(the monthly median), and 90% of the days for each month. These maybe
regarded, as final results. The previous report3 gave corresponding data
for the period September, 1951, through September, 1952.
The frequency and path length were substantially the same for the
Cedar Rapids-Sterling, Fargo-Churchill and. Anchorage-Barrow paths; the
differences observed in diurnal and seasonal variation of transmission
loss therefore represent 'mainly geographical dependence. There are pro-
bably residual differences of up to 6 ab which are attributable to de-
partures from ideal siting; furthermore, the Cedar Rapids-Sterling an-
tennas were designed to aim the first lobe at 110 Km in the initial ex-
istence test, and their heights remained unchanged throughout the program.
The antennas for the other two paths were designed to aim the main lobes
at the 85 Km scattering height.
3.11 Cedar Rapids to Sterling Test Path - 49.8 Me/s
Forty-months of data are presented covering completely the period
from September, 1952, through December, 1955, in Figures 1-10 and in the
tables of Appendix XI.
- 5 -
The cumulative distribution of hourly median values of signal inten-
sity, by season, are given in Figures 23 through 27. Because of their
direct usefulness in evaluating communication performance, these distri-
butions have also been expressed in terns of transmission loss and loss
relative to inverse distance transmission (that is, relative to perfect
reflection in the ionosphere). Such a distribution for a complete year
of data has been published as Figure 10 of Reference 1.
3.12 Fargo to Churchill Test Path - 49.7 Me/s
Seven months of data, for September, 1952, through March, 1953, are
given in Figures 11 and. 12 and in the tables of Appendix XI.
Cumulative distributions of the hourly values of signal intensity,
by season, for the complete Observing period on the Fargo-Churchill path,
September, 195I throughMarch, 1953-, are given in Figures 30 and 31.
3.13 Anchorage to Barrow Test Path - 48.87 mc/s
Ten months of data, September, 1952, through June, 1953, are pre-
sented in Figures 13 through 15 and in the tables of Appendix XI. Cumu-
lative distributions of the hourly median values of signal intensity, by
season, are given in Figures 28 and 29 for the complete observing period
on this path, September, 1951, through June, 1953.
3.14 Routine Recording at Other Frequencies
a. Anchorage to Barrow - 24.325 Mc/s
Four months of data, March, 1953, through June, 1953, representing
diurnal variation of hourly median signal intensity using Yagi transmit-
ting and receiving antennas, are given in Figure 16 and in the tables of
Appendix XI. The cumulative distributions of the hourly median values
for two spring months and two summer months are given in Figure 35.
b. Cedar Rapids to Sterling - 27.775 Me/s
A total of 18 months of data, covering the periods April, 1954,
through January, 1955, and May, 1955, through December, 1955, are given
in Figures 17 through 21 and in the tables of Appendix XI. Other re-
quirements for the Cedar Rapids transmissions caused interruption Of the
observations from. February through April, 1955. (Transmissions from May,
1953, through April, 1954, were recorded, but questions regarding trans-
mitter power and antenna adjustment made the data unreliable.) Cumulative
distributions of the hourly median values, by season, are given in Fig-
ures 33 and 34. Rhombic antennas, scaled relative to the 49.8 Me/s an-
tenna, were used at 27.775 Me/s.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
- 6 -
c. Cedar Rapids to Sterling - 107.8 ItiCis
Four additional months of data are given in Figure 22 and in tables
of Appendix XI for .the period. October, 1952, through January, 1953. The
previous report3 gave corresponding data for the months January through
September, 1952. The cumulative distribution of the hourly median values,
by season; fromliarch, 1952, through January, 1953, is given in Figure 32.
3.15 Seasonal Variation of Signal Intensity
A,discussion of the important seasonal variation of signal intensity
was given in Reference 1, p. 1188. Figures 36 and 37 herein show running
plots of monthly median values of signal intensities for three-hour periods
centered at 00, 06, 12, and 18 hours local time at the path midpoint, as
well as a running plot of the level exceeded 95% of the hours during each
month. The daytime signal (noon) for the temperate latitude path from
Cedar Rapids to Sterling (Figure 36) shows clear semiannual maxima during
summer and winter with equinoctial minima; these are observed at all three
frequencies. At 18 hours, the winter maximum is somewhat suppressed. but
still in evidence; the sur maximum remains distinct. The summer maxi-
mum is evident day and night.
For the arctic and subarctic paths from Anchorage to Barrow and. Fargo
to Churchill respectively (Figure 37) very strong summer maxima are ob-
served. For the Fargo-Churchill path it is interesting that during the
daytime, both at 12 and 18 hours, a secondary winter maximum appears near-
ly as strong as the summer maximum.
3.16 Annual Variation of Signal Intensity; Correlation with
Geomagnetic Observations and Ionospheric Conditions Ob-
served. at High Frequencies
Figures 38 and 39 show a comparison of five years of observations of
signR1 intensity during the months of March, June, September and December.
These data should. be considered in connection with the more comprehensive
discussion of solar 'cycle dependence given in Appendix X. Appendix III
summnrizes the study of correlation of the observed VHF propagation with
geomagnetic indices and with ionospheric propagation conditions observed
at high frequencies by vertical incidence sounders.
The long-term variation of signal intensity was also observed. by
plotting twelve-month running means of monthly medians of received sig-
nal intensity (Figure 4o) for three-hour periods centered at 00, 06, 12
and 18 hours, and of the level exceeded 95% of the hours during the month
(See also Reference 4). It was interesting to observe whether the long-
term variation of signal intensity was different for noon values, for ex.-
ample, when a dominant solar influence night be expectedl as contrasted
to the valves observed during early morning (at 06 hours) when meteoric
influence is presumed to be predominant. The variation of the noon value
- 7..
was found indeed to have a greater long-term, variation than the early morn-
ing values, and to correspond more closely with the twelve-month smoothed
index of magnetic activity than directly with smoothed relative sunspot
number. A strong long-term variation also appears at 18 hours and tends
to influence the levels of the weakest signals observed., as indicated. by
Figure 4o (e); the weakest signals are usually observed in the early eve-
ning hours. Appendix X discusses in greater detail the physical signifi-
cance of the kind of data presented in Figure 4o.
3.17 Sporadic-E Propagation
Reference 1 (pp. 1200-1204) discusses the interpretation of the very
strong signals dbserved in the records, considered to be a result of spor-
adic-E propagation; a brief statistical summary of the observations of
sporadic E to 1954 is given in the Figures 25 and 26 of Reference 1. Ap-
pendix IV of the present report gives a detailed summAry of the data from
1951 through 1955. A. more comprehensive report of the sporadic-E observa-
tions on the Cedar Rapids to Sterling path, and their interpretation rela-
tive to vertical incidence ionosphere soundings, has also already been pub-
lished. separately.5
Cumulative distributions of the signal intensities due to sporadic-E
propagation are presented in Reference 5 for various time intervals. Evi-
dence is presented that the occurrence of sporadic E is associated with
geomagnetic activity. During the observing period 1952 through 1954 at
least twice as many hours of oblique incidence sporadic-E propagation oc-
curred on the five magnetically quiet days per month as occurred on the
five disturbed days. The preference for quiet days was weaker in 1951 and
disappeared in 1955. The dependence of received power on frequency during
the periods of sporadic-E propagation is strikingly different from that
observed during conditions of normal scatter propagation. The ratio of
signal intensities observed at 49.8 and 27.775 Mcis during conditions of
Es propagation corresponded to a median frequency exponent for received
power in the neighborhood of 12, although the spread of values was very
great. This is noted in Appendix I, in connection with analysis of the
frequency dependence of ionospheric scattering.
3.2 Experimental and Developmental Phases of the Program
3.21 Frequency Dependence
The results of the observations of frequency dependence, using trans-
missions at 107.8, 49.8 and 27.775 Neis are discussed in Reference 1
(pp. 1196-1199). The only additional point to mention is that certain
limitations of the present program prevented obtaining the data under the
very desirable conditions of (a) simultaneous observations at all fre-
quencies, (b) an adequate number of frequenciesto estabiTah the curva-
ture of frequency dependence law if it genuinely departs from a constant
exponential, (c) ideal sites at transmitter and receiver to assure uniform
-
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
- 8 -
and identical illumination of the scattering volume at each frequency,
(d) frequency stability and receiving bandwidth appropriate for reliable
observations at low values of signal-to-noise ratio, especially encoun-
tered at the higher frequencies, and to permit studies of the fading
structure at each frequency.
Appendix I gives detailed data of frequency dependence not reported
earlier, including the results of true power measurements in comparing
the 107,8 and 49.8 Mc/s transmissions.
3.22 -Scattering Heights, Dependence of Transmission Loss on
Path Geometry
The results of experimental investigations of scattering at VHF us-
ing pulse techniques are reported in Appendix II. Under this topic scat-
tering heights and dependence of transmission loss on scattering angle
in the plane of the great circle path were investigated; observed scat-
tering heights were compared with simultaneous observations of virtual
heights of 460 Kc/s reflections observed near the midpoint of the path.
Reference 1 summarizes the results obtained from the early measurements
of heights of scattering, from one-way measurements of relative transit
times of the tropospheric component and the ionospheric component
(pp. 1194-1195), as well as the angle dependence results (pp. 1195-1196).
The most important additional contribution included in the present
report involved the use of round-trip techniques to obtain the heights
for ionospheric and tropospheric scattering. Observations were made dur-
ing December, 1954, and October, 1955, at 49.8 Mc/s over a 624 Km path,
in a cooperative experiment between the National Bureau of Standards and
Lincoln Laboratories of the Massachusetts Institute of Technology. In
this experiment it was concluded that the scattering heights were slightly
higher than had been deduced from the earlier.observations relative to
the tropospheric component. The hourly median values of equivalent ionos-
pheric heights based on round-tiip delays were,-during December (1955)
86 Km during midday (12 to 14 hours) and 90 Kra at 10 and 17 hours (see
Figure 10 of Appendix II).- There were not enough data to calculate me-
dians for the other hours of the day. The median value of all measured
daytime delays gave an equivalent midpoint height of 87 Km. The median
value of all the daytime observations during October was also 87 Km.
Effective midpoint heights for tropospheric scattering as high as 29 Km
were observed; the hourly median values for December varied between 20
and 28 Km, with the median of all observations equal to 24 Km. The ob-
served effective heights in this experiment were undoubtedly influenced
by the radiation patterns of the antennas which were designed to illumi-
nate the ionospheric scattering volume at 90 Km, so that the intensity
of illumination in the troposphere increased very rapidly with height.
As these were substantially the conditions for the earlier one-way pulse
transmission experiments in which ionospheric transmission delays were
referred to the tropospheric delays, it was concluded that the effective
tropospheric heights used in the earlier models had been underestimated;
this would have had the effect of giving ionospheric scattering heights
which were several Km below actual heights.
- 9 -
To obtain information about the possibilities of off-path interference,
signal intensities of the Cedar Rapids transmissions at 49.8 MCA, beamed
towards Sterling, were received at off-path receiving sites and compared
with signal intensities observed simultaneously at Sterling.
An interesting and probably significant correspondence was observed
between the diurnal variations of the effective midpoint ionospheric scat-
tering height and the virtual height of reflections observed at 46o Kc/s
(see Appendix II, Figure 2), even though the actual heights observed dif-
fered by a few Km.
By use of narrow beam receiving antennas at off-path receiving sites,
directed at the midpoint of the Cedar Rapids to Sterling path, an experi-
mental study was made of the contribution of aspect-sensitive meteor re-
flections in the total signal observed as ionospheric scattering. A sum-
mary of early results with a single off-path receiving location was re-
ported in Reference 1. Appendix VI, Part 2, gives a complete report on
this study using two off-path receiving locations; the results confirmed
that there are two principal modes involved in ionospheric scatter propa-
gation. The dominant mode during the daylight hours is apparently mainly
of solar control, and is most effectively utilized by employing antennas
directed along the great circle path between the transmitter and receiver.
The dominant mode during the nighttime hours is attributed to meteoric
ionization, and received signal intensities can be enhanced by the use of
off-path transmission. It is worth emphasizing that the results given in
Appendix VI are representative of wintertime conditions and an east-west
path; results reported earlier in Reference 1 from a summertime experiment,
using only one off-path recording station, with the same geometry, showed
the off-path transmission to be dominant for fewer hours of the day. The
fading of the off-path signal is more rapid than for great circle trans-
mission; further experiments are necessary to determine whether the fading
distribution is as favorable as great circle transmission, in relation to
error rates in binary systems. Further experimental work is also necessary
to determine the optimum azimuth for the total scatter signal, which varies
with time of day and season of the year; observations of the aspect of the
larger meteor bursts are not necessarily directly applicable.
3.23 Signal Fading Characteristics
Summaries of the studies of short-term fluctuations of the observed
signal are given in Reference 1, pp. u6-u88 and pp. 1207-1211, as well
as in Reference 4 and a paper by Sugar.
3.24 Doppler-Shifted Meteor Echoes
Observations of the rate of occurrence of Doppler components (meteor
whistles) differing from the carrier frequency by at least 200 cps were
made at Sterling during 1954 and 1955. As an example of the results, the
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
-10-
average rates for the month of April, 1955, were given in Reference 1,
Figure 57. Appendix V summarizes the observations for the complete pe-
riod, giving diurnal and seasonal variation of the rate of observation
of meteor whistles.
3.25 Development of Low-Power Narrow Band Recording Equip-
ment for Routine Observations
In order to reduce the power requirement for routine observations
of transmission loss, development of a narrow band recording receiver
was undertaken along with the necessary stable frequency control for a
recording receiver and transmitter; a 3 kilowatt VHF transmitter was de-
veloped_ for unattended operation. Appendix VII describes the fixed-fre-
quency narrow band receiver; direct heterodyning of the VHF signal down
to 1,000 cps intermediate frequency allows simple plug-in filters to be
used. The receiver image response is eliminated by a twin-mixing pro-
cess with phase shifting And sensing networks. A stable frequency con-
trol for the receiver and the transmitter is described in Appendix IX;
a 1 14c/s high stability frequency standard is described, as well as addi-
tional units to derive frequencies which are not multiples of 1 licis but
which have the frequency stability of a 1 lic/s reference oscillator. The
three exciter units described are of the filter type, the phase-locked
oscillator type, and the direct multiplier type. The last one, for appli-
cations where the operating frequency can be made a direct multiple of
the crystal frequency, has proven the most satisfactory for field use,
and crystals having the proper characteristics are now readily obtainable
at sub-multiples of arbitrary operating frequencies.
The 3 KW transmitter, described in Appendix VIII, was designed to
operate at fixed frequencies in the range 30 to 75 Ncis, and is capable
of unattended operation for long periods. Compactness compatible with
ease of servicing was an important objective, since many of the trans-
mitters would be installed In small field laboratory buildings or trail-
ers. The prototype unit was placed in operation at Cedar Rapids in Perch,
1956, upon termination of the high-power transmissions furnished by the
Lincoln Laboratories' contract with Collins Radio Company; the unit re-
mained in substantially continuous service until termination of regular
observations on this path June 301 1958. Subsequently, the number of
additional units have been built and operated -in connection with other
projects.
4. DISCUSSION AND CONCLUSIONS
4.1 General
The comprehensive summary to 1955 of the important features of trans-
mission by VHF scattering has been published in References 1 and 4 and will
not be repeated here. The physical processes and ionizing agents operating
in the lower ionosphere giving rise to the observed propagation are dis-
cussed; short and long-term characteristics of the observed signals are
a.
described as well as the results of experiments with spaced and other an-
tenna arrangements. Dependence of the strength of the signals on path
length, frequency, scattering angle, and geographical position of the
transmission path are summArized. Results of the measurements of real-
ized gain of directive antenna systems are interpreted in relation to
parameters of the scattering geometry, the influence of meteoric ioniza-
tion, and the existence of large-scale inhomogeneities in the ionosphere.
Application of VHF propagation to communication systems operating over
distances from about 1,000 to 2,300 kilometers is discussed. Design con-
siderations are given for antenna systems, including siting, choice of
polarization, and space diversity. The useful range of frequencies and
certain modulation techniques are discussed, as well as multipath effects
associated with the scattering process and with the presence of meteoric
and auroral ionization. Reliability of performance of typical systems is
estimated.
Attention is also drawn to a paper7 reporting studies in the extreme
distance range for transmission by scattering.
4.2 New Results
a. Scattering Heights
The round-trip pulse observations have indicated scattering heights
of about 87 Km, a few Km higher than deduced from earlier one-way pulse
observations; the lower heights, in the range 70 to 80 Km, as observed
during the earlier tests, were not observed either during October or De-
cember during the round-trip tests. The reason is not entirely under-
stood but may indicate a somewhat greater angle dependence associated
with scattering from the lower heights, making transmission loss over the
very short (624 KM path) too great for observability of the low-lying
strata. Furtkier observations should be carried out over paths longer
than 1,000 kilometers, with a systematic program for observation of di-
urnal and seasonal variation of scattering heights and contributions from
the various strata.
b. Solar Cycle Dependence
Noon values of signal intensity, after smoothing to remove effect
of seasonal variations, exhibit a long-term variation of received power,
of the order of 6 db, which tends to follow more closely the index of
magnetic activity than direct smoothed relative sunspot number. Values
at 18 hours show a similar though slightly smiler variation, while the
nighttime values and early morning values, which presumably include the
effects of meteoric ionization, show substantially less long-term varia-
tion.
As a practical matter, it was of interest to learn whether the higher
transmission losses which could be expected with the use of higher fre-
quencies in the law range during times of maximum solar activity to avoid
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
-12-
long-distance F2 propagation, might be Offset to some extent bY increased
signal intensities due to solar dependence of the VHF scatter signal. One
must conclude that the rise in signal intensity is not enough to compensate
for losses involved in the use of the higher frequencies. Only 1 or 2 de-
cibels increase in signal intensity was observed during the periods of
weakest signal intensity, that is the level exceeded 95% of the hours dur-
ing the month. Furthermore, the time lag between the rise in signal in-
tensity and the rise in sunspot number, which leads to the use of higher
frequencies, is such that the practical advantage is reduced.
c. Useful Range of Frequencies
The data given in Reference 1 on frequency dependence of scattered
signal intensity and the background galactic radio noise suffices to in-
dicate bow required power varies with frequency, for a given required sig-
nal-to-noise ratio. It is recommended that the data for scaled antennas
be used. In practice, because of the failure to realize gain in propor-
tion to increased directivity, antenna apertures will generally-be scaled
in proportion to frequency. In the frequency-range 30 to 50 Mc/s, approx-
imately 4* db increase in transmitter power is required for 10 Mcis in-
crease in operating frequency.
A slightly lower exponent for frequency dependence may be deduced
from the cnnulative distribution of transmission loss at various frequen-
cies, as given in figures. However, the difference from the median value
(Appendix I) obtained by comparing siwultaneous hours of transmission is
small and should be regarded with caution. The cumulative distribution
of signal intensity does not necessarily involve simultaneous observations
for all hours at the various frequencies.
In spite of the greater signal-to-noise ratio realized at the lower
frequencies of this range, applications requiring the highest trder of
reliability will need to use, at times, frequencies in the higher part
of the range. On rare occasions intense absorption in arctic regions can
cause effective blackout at the lower frequencies in the range, say near
30 Mc/s or below; no such occurrences have been observed at 50 NC/s. It
is unlikely, even at lower frequencies in the range, that such circuit
interruptions would occur on more than a few occasions during a solar cy-
cle. -
In the absence of suitable international frequency allocations and
assignment plans at VHF, designed to avoid mutual interference from long-
distance F2 propagation/ operation at frequencies above the F2 maximum
usable frequency (EDF) must be considered to avoid long range interference.
As far as the effects of ground. backscatter propagated by F2 are con-
cerned, aS a source of self-interference to high speed telegraph systems,
development of sUitable modulation techniques and/or antenna characteristics
tr
-13
gives definite promise of alleviating the difficulties.8'9 It can be
noted, therefore, that civil circuits, or any which can tolerate the
rare interruptions mentioned above, could quite satisfactorily use fre-
quencies below the F2 EDF down to as low as 30 En/s, or perhaps a bit
lower. There is certainly a valid question as to the wisest frequency
allocation policy, involving considerations which are beyond the scope
of this report.
Reference 10 gives world-wide contour naps of frequencies which ex-
ceed the F2 EDF for long-distance propagation 1% and 10% of the hours dur-
ing the various seasons of the year, for conditions of maximum and mini-
mum solar activity.
d. Meteoric Contribution
The off-path transmission studies have shown that hourly median trans-
mission losses may be reduced during the night hours (and during the eve-
ning hours of weakest signal intensity) by slewing the antenna beans to
one side or the other of the great circle path, to take advantage of specu-
lar reflections from meteor trails. The advantage would appear to be great-
er during the winter when the hours of sunlight on the path are fewest;
the advantage is realized during only a few night hours during the summer
months. Further studies need to be made of the diurnal and seasonal vari-
ation of (a) the gain obtainable from off-path transmission, (b) the opti-
mum angles for beam sieving, and (c) the multipath structure and character-
istic short-term fading of the off-path signal as these may affect utility
of modulation techniques. It is known that the observed fading rates are
much higher for off-path transmission than for great circle transmission,
and it may well be that the amplitude distribution of the short-term fading
is less favorable for conventional continuous modulation systems.
e. Sporadic-E Propagation
This report includes a comprehensive statistical summary of diurnal,
seasonal, and annual variations of occurrence of sporadic-E propagation
for the temperate latitude and arctic paths. From the point of view of
mutual interference with other services, it is evident that terminals of
scatter services must be separated from other services by greater than
one hop sporadic-E distances, i.e. > -2000 Km. During the summer months
June and July at temperate latitudes 30 to 40 hours per month of sporadic-
E propagation are typically observed at 50 Mc/s, during May and August some-
what less than 20, and during other months of the year only a few hours per
month. It is important to note that the transmission loss associated with
sporadic-E propagation varies over a wide range and iS typically much greater
(by at least 30 decibels) than that corresponding to a perfect reflection
in the ionosphere.
4.3 Outstanding Problems
During the course of the work the need. has become evident for addi-
tional investigations, or expansion and refinement of studies reported
herein, to obtain data which will provide a better basis for optimizing
performance of systems.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
- 14 -
a. Frequency Dependence
-
The conclusions relative to frequency dependence of ionospheric
scattering have not been left in an entirely satisfactory condition
based on the data obtained in this program (See Reference 1 and Appen-
dix I of this report). It was possible to compare only two frequencies
at one time; the data showed. a curvature in the frequency dependence law,
but it is now believed the data may have been inadequate to assure gen-
uineness of the curvature. Neither the transmitting or receiving site
(but especially the transmitting site) was ideal enough to assure uni-
form illumination of the scattering volume at all the frequencies. The
background galactic noise observed at 27 Mc/s was about 3 decibels be-
low the level expected. based. on frequency dependence measurements of
other observers; this led. to considerable investigation of the losses in
the receiving antenna, with no satisfactory explanation resulting. The
question is left whether the signal intensities observed, at 27 MC/s were
also somewhat lower than should have been observed.
Wheelonil,12 has indicated. the very-important requirement for simul-
taneous observations at a number of frequencies, say five, with careful
attention to uniform illumination of the ionosphere at all the frequencies;
he has shown that a genuine curvature in the frequency dependence data can
be expected on the basis of the theory of scattering from a spectrum of
turbulent irregularities, and that the general slope and curvature of the
frequency dependence law can be used. to deduce important meteorological
parameters of the scattering region. A comparison or characteristics of
envelope and phase fluctuation in fading at the various frequencies can
also yield important information on the physical structure and. motions
in the scattering volume; it is also important to know how the short-term
fluctuation of signal varies with frequency in order to specify limita-
tions on modulation techniques, or even required. power versus frequency.
?
Another project has been undertaken13 along these lines, using trans-
missions over the Long Branch, Illinois to Boulder, Colorado path at fre-
quencies near 30, 40, 50, 74, and 108 Mc/s.
b. Optimum Antenna Design, Including Beam Slewing
Further propagation studies, at least throughout one full year, are
required. to learn the real extent of any advantages of beam slewing to uti-
lize off-path meteor reflections during hours when-transmission loss may
be less by this mechanism than by great circle transmission. The diurnal
and seasonal variation of optimum directions of transmission should be
studied, as well as an explicit investigation made of the multipath and
fading characteristics of received signals as a function of direction of
transmission and beam widch of the antennas. Earlier observations have
shown that during night hours use of low-gain broad beam antennas does not
increase the average transmission loss, relative to high-gain narrow beam
antennas, by anything like what might be expected from actual plane wave
gains of the antennas. The broad beam antennas, of course, make greater
use of the off-path contributions of scattering from irregularities and
t.
4
-15-
aspect-sensitive meteor trails. The fading rate is higher for the of
contributions to the signal than for the great circle transmission
and the amplitude distribution of the total signal, which contains the
effect of many bursts, may not be as favorable for continuous transmission
systems as great circle transmission.
In connection with another project, theoretical computations of opti-
mum antennp. height for scatter propagation in the lower ionosphere have
been made-"-5 by integrating over the scattering volume, taking into ac-
count the radiation pattern of the antennas and the effect of scattering
angle. These results indicate that the optimum antenna height is lower
than the height which aligns the first lobe at the path midpoint of a
symmetrical path, and that there is a considerable range of lower heights
which show a gain over the lobe alignment height. The important conclu-
sion is that the transmission loss is a very broad function of the height
of the transmitting and receiving antennas and that substantial compromises
may be tolerated below lobe alignment height. Experimental verification of
this theoretical result should be obtained by establishing a pair of fixed
symmetrical antennas as a reference and a pair of adjustable antennas over
a given path at two slightly different frequencies; verification of the
theoretical height gain curves could lead to significant savings in the
cost of ionospheric.. scatter systems and allow the use of a broad range of
heights in compromised situations.
Further work should be done to improve the design of antennas with
respect to suppression of radiation outside the main lobe, an important
avenue of attack on the multipath problem. Suppression of radiation out-
side the main lobe by at least 40 decibels relative to the maximum seems
a desirable design goal for high speed telegraph systems, and considerable
progress has already been made along this line in another project.1?
c. Modulation Studies
Further modulation studies need to be made not only to define the
transmission capacity using conventional techniques in ionospheric scat-
ter propagation, but also to develop more efficient techniques making
effective use of the special characteristics of the signal including the
bursts; effort should be made to eliminate the difficulties imposed by
Doppler-shifted meteor echoes and long-delayed. backscatter propagated
via the F2 layer during times of high level of solar activity.
Matters such as correlation of envelope and phase over a frequency
bandwidth deserve further investigation in relation to limitations im-
posed on modulation. A lpited amount of work has already been under-
taken along these lines.1?
Detailed study of the fading and multipath characteristics of re-
ceived signals as a function of frequency may alter the idea that the
transmission capacity via ionospheric scattering varies with frequency
in the same way as the average transmission loss varies with frequency.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09 ? CIA-RDP81-01043R003000180001 4
-16-
Declassified in Part - Sanitized Copy Approved for Release a 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
d. Geographical Dependence
". ? .4n, ...MI`S&1113=
Experiments presently being conducted are expected to yield. valuable
information on ionospheric scattering observed in equatorial regions)-7
Observations in other parts of the world at temperate latitudes are also
required to determine whether the weakest signal intensities encountered
are universally about the sane (within a few decibels), at least at 50 Mc/s
where the rare occurrences of intense anomalous absorption have practically
no effect. (Published resultsl? showing much greater transmission loss ob-
served over a .European path at 48 me/s, from Gibraltar to the United. King-
dom, are difficult to understand in the light of other measurements and the
basic ideas of the ionospheric properties giving rise to regular VHF propa-
gation. It is virtually certain that experimental difficulties, especially
the effects of the transmitting antenna arrangement at Gibraltar, can ac-
count for the very unsatisfactory results dbserved on this path.)
The scope of understanding of the physical mechanisms important in
ionospheric scattering can be enhanced. by observing clear relationships
to other measurable ionospheric parameters. Early attempts to relate the
data observed on the Cedar Rapids-Sterling path to critical frequencies
of the E region were unsuccessful; it was expected that day-to-day fluctua-
tions in transmission loss might be more closely related to absorption ob-
servations at HP over the same path, but attempts to correlate VHF signal
variations with day-to-day absorption on a path from Ohio to Virginia gave
no clear-cut correlation. Nor did a study of stratification at low ionos-
pheric heights recorded. by sweep frequency observations from 100 Kcis to
2 Mc/s at Sterling, Virginia, give any clear-cut correlation. It is felt,
however, that all of these efforts to relate the VHF signal to other ob-
servable ionospheric characteristics were limitPd. by not observing the
ionospheric characteristics in the same geographical location as that in
which the scattering takes place. Observations should definitely be uer-
taken near the midpoint of a scattering path. In New Zealand, Gregory`
has shown that low-lying strata are observable at vertical incidence at
2 Me/s, and he believes these to be closely related to or, in fact, to rep-
resent the same propagation mechanism as observed at VHF at oblique in-
cidence. He has studied diurnal and seasonal variation and solar cycle de-
pendence and finds striking correspondence with the VHF results.
- 17 -
ACENOWLEDGEMENTS
Acknowledgement is made to each of the organizations and individual
members (and former members) of the staff of the National Bureau of Stan-
dards listed below, for substantial help in furthering the research and/
or preparation of this report:
Contractors to the National Bureau of Standards
Collins Radio Company
Engineering Experiment Station
North Dakota Agricultural College
Page, Creutz, Garrison, and Waldschmitt
Consulting Radio Engineers
Other Organizations and Government Agencies
Lincoln Laboratories, Massachusetts Institute of Technology
U. S. Army Signal Corps
U. S. Navy (Pet. 4)
U. S. Air Force
Military District of Washington
Federal Communications Commission
Civil Aeronautics Administration
Personnel of the National Bureau of Standards
D. K. Bailey
R. Bateman
Catherine M. Bell
E. L. Berger
G. E. Boggs
H. G. Booker (consultant)
K. L. Bowles
W. H. Daniels
V. R. Eshleman (consultant)
J. Feinstein
R. S. Gillespie
Declassified in Part - Sanitized Copy Approved for Release ? 50 -Yr 2014/06/09 ? CIA-RDP - 4
W. S. Gofbaczewski
W. B. Harding
N. C. Hekimian
Adeline N. Kinchloe
Ann C. McCartney
G. F. Montgomery
W. I. Nodine
R. C. Peck
V. C. Pineo
Anita F. Reisig
C. C. Smith
E. K. Smith
J. L. Spindle
G. R. Sugar
K. W. Sullivan
P. G. Sulzer
Gloria W. Sutton
Joan E. Tveten
M. C. Weeg
A. D. Wheelon (consultant
D. C. Whittaker
Laura F. Williams
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
- 18 -
REFERENCES
1. D. K. Bailey, -R. Bateman, R. C. Kirby, ".Radio Transmission at VHF by
Scattering and Other Processes in the Lower Ionosphere," Proc. I.R.E.,
vol. 43, pp. 1181-1230, October, 1955. (Reprints are included as
Appendix XII in a limited number of copies of this report, furnished
to Contracting Officer, USASEL; copies were also furnished earlier
to the Contracting Officer as NBS Report 3563.)
2. D. K. Bailey, R. Bateman, R. C. Kirby, "First Report on Regular VHF
Ibnospheric Propagation Observable Over Long Distances," NBS Report
No. 8A111, June 30, 1952, (Interim Technical Report on MIPR No.
821-PHIBP-51-04).
3. D. K. Bailey, R. Bateman, R. C. Kirby, "Second Report on Regular VHF
Ionospheric Propagation Observable Over Long Distances," NBS Report
No. 8A117, June 30, 1953, (Interim Technical Report on EIPR No.
821-PHIBP-51-04).
4. R. C. Kirby, "VHF Propagation by Ionospheric Scattering - A Survey of
Experimental Results," Trans. I.R.E., CS-41 1955.
5. R. M. Davis and. E. K. Smith, "The Effect of Sporadic E on VHF Trans-
mission in the U. S.," NBS Report No. 5547, Jan. 28, 1958.
6.
7.
G. R. Sugar, "Some Fading Characteristics of Regular VHF Propagation,"
Proc. I.R.E., vol. 43, pp. 1432-1436, October, 1955.
R. C. Kirby, ?Extreme Useful Range of VHF Transmission by Scattering
from the Lower Ionosphere," I.R.E. National Conventional Record,
Part 1, pp. 112-115, March, 1958.
8. H. V. Cottony, A. C. Wilson, "High Gain Low Side-Lobe VHF Antennas fbr
Ionospheric Scatter Communication," NBS Report in preparation.
? ?
"Backscatter Multipath Elimination Program," Page Communication Engineers
Report PCE-R-4681, August, 1957.
10. R. M. Davis and J. W. Finney, "Interference Effects of F2 Propagation,"
NBS Report No. 6004,
11.
12.
A. D. Wbeelon, private communication, December, 1955.
A. D. Wbeelon, "Radio Frequency and Scattering Angle Dependence of
Ionbspheric Scatter Propagation at VBF,".Jour. Geophys. Res., vol 62,
pp. 93-112, March, 1957.
13. "Quarterly Progress Report on Engineering Studies in VHF Ionospheric
Scattering," NBS Report No. 5554, February, 1958.
14. R. G. Merrill, "Radiation Patterns in the Lower Ionosphere and. Fresnel
Zones for Elevated Antennas over Spherical Eaith," NBS Report in pre-
paration.
- 19 -
15. R. G. Merrill, "Optimum Antenna Height for Ionospheric Scatter Propa-
gation," NBS Report in preparation.
16. J. W. Koch, "Modulation Studies for VHF Ionospheric Scattering," NBS
Report in preparation.
17. K. L. Bowles, R. Cohen, "N.B.S. Equatorial Region VHF Scatter Research
Program for the I.G.Y.," QST, pp. 11-15, August, 1957.
18. G. A. Isted, "Analysis of Gibraltar - United Kingdom Ionospheric
Scatter Signal Recordings," Jour. I.E.E., vol. 104, Part B, R25231
Jan., 1958.
19. J. Gregory, "The Relation of Forward Scattering of VHF Radio Waves
to Partial Reflections of MF Waves at Vertical Incidence," Jour.
Geophys. Res., vol. 61, pp. 165-169, 1956.
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
60
50
40
30
20
bio
0
0
0
Iii
0
0
CEDAR RAPIDS TO STERLING ? 49.8 Mc/s
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN?CIRCUIT ANTENNA VOLTAGE ? 60011 ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40KW; ESTIMATED ANTENNA POWER 30 KW
SEPTEMBER, 1952
A
107.
= 50%
z
/'\
I I 1
OCTOBER, 1952
1070
00
30
20
10
0
04
08
12
16
20
.???????.,
9070
0000
-
I II II
NOVEMBER, 1952
:
_
-
_
:
-
:
-
:
_
2
:
_
_
:
107:
:
__
-
:
:
50%
:
-
_
_
:
i_o_?.0.-?
90%
Curves
50%
:
-
show
of days
Analysis
signal
of
data
tit!
level
heavy
all data,
otter
equalled
middle
Including
eliminating,
tit
_
_
or
cur'
I
-I
:
-
-
lit
:
-
-I
t
1
t
?Same
tlil
00
04
08 12
16
20
ex
),
va4
00 00
04
08
12
16
20
00
z
DECEMBER,
II
III
I
/952
-
.
-
_
:
-
_
.
-
:_
_
-
_
_
_
_
_
:
1 oe/0
-
_
_
_
.
-
_
-
:
NMIPF
_
:
_
_
po%
_
.::
ceeded
90%
ts affected
sterpolatian,
I-I I
10%
of
by
t I
1
of days
days
(upper
(lower
curve),
curve)
-
-
_
-
Es -type propagation
Es-type propagation
t Iii I I i
I
_
-
_
04
08
LOCAL TIME AT PATH MIDPOINT
FIG. I
12
16
20
00
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
60
50
40
30
20
-J 10
0
0
C.)
0
00
0
coo 60
co
LU
-J
m 50
Ui
? 40
CEDAR RAPIDS TO STERLING ? 49.8 Mas
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN?CIRCUIT ANTENNA VOLTAGE ? 600 4, ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30 KW
EMIJANUA.. k
); 1953 III
III am nom
imi immilme
r
imm........ E
iinia.. 10% IIit
11111101111 I Ill 111
Kew 5?77 moist
;
raritioNiNfilis
ow?.., 0 "(7/0 imPE
ANNIE
imisim Imot
miiiiminumimmii
I
FEBRUARY, 1953
10%
1\
12 16 20 00 00
I I
MARCH, 1953
1111111 ?1111
10%
2011 111111I1011& 1111111
IIIIIIIPM .50%_. L It
10 alleratillral%6b40471
iiiiillilll 913'Y? 'braille
Curves show signal level equalled or exceeded 10% of days (upper curve),
50% of days (heavy middle curve) 90% of days (lower curve)
?Analysis of all data, includmg values affected by Es?type propagation I
--Same data after eliminating, by interpolation, Es?type propagation
08
12
16
20
00
APRIL, 1953
10%
NBS
12 16 20 00 00 04 06
LOCAL TIME AT PATH MIDPOINT
FIG. 2
12
16
20
00
?
CEDAR RAPIDS TO STERLING ? 49.8 Mas
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN ?CIRCUIT ANTENNA VOLTAGE ? 600 a ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30 KW
?1
MAY,
III'
/953 I
M
signal
(heavy
leve
equalled
midd
:
.
.
.:
I:
1
Curves show
50% of days
:
.
or
e curv
-
_
:
-
-
.
IC%
-
_
--1---
-
-..........-
No
5074,
90%
-
.
:
:
.
1-
-
.
-
5C
2C
i
-J
0
0
0
Iii
>60
0
03
350
03
Ui
040
00
30
20
10
0
04
08
12
16
20
-ceded
90%
1
10%
of days
J
of days
(upper
(lower
I
JUNE,
I -
1953
III
curve),
curve)
-
I
I
In.
/
10%
l
,
50%
s\
91
LI
-swam
90%10415-,4
Iiilig
I
11/111t1/1
0000
I I 1
JUL); /953I
.
1
Il
II
II
III
10%
II
11111
I
111"/-5-0-L.,--1-4,
..11,
..1111111111111
?i
\ .
IrilliMralli"441111I
90%
''''"
iuuu?
?Analysis
of all data, including
data after eliminating,
1,Iiliq

?Same
?iiI
vol
04
08
12
16
20
IA
?1?
AUGUST, 1953
II
I.
I
I
II
IIII
11111.10%
II
II
11:11111kink
lUll
mg,
I
11
sonoinwurpi
war
515
50%
90%
am
1161,410
affected by
interpolation,
P. I .
Es -type
Es -type
I .
propagation
propagation
1 .
I
IllgJCS
.
.
.
i
i
00 04 08 12 16 20 0000 04 08
LOCAL TIME AT PATH MIDPOINT
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
FIG. 3
12
16
00
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
CEDAR RAPIDS TO STERLING ? 49.8 Mc/s
CEDAR RAPIDS TO STERLING ? 49.8 Mas
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN-CIRCUIT ANTENNA VOLTAGE - 600 a ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40KW; ESTIMATED ANTENNA POWER 30KW
SEPTEMBER. 1953
Clines show signal level equalled or ex
50% of days heavy middle curve),
.
-
_
IIIII
OCTOBER, 1953
.
:_
-
-
-
_
:
:
:-...
_
-
.
.
_
.
.
-
_
.
_
_
';erried
90%
)0%
of
of cloys
days
'Op94r
(lower
curve),
curve)
_
_
-
-
_
_
:
-
0%
_
-
_
_
_
_
_
-
-
-
_
-
es
da
es
en,
1
:
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN - CIRCUIT ANTENNA VOLTAGE - 600 a ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLFIER 40KW; ESTIMATED ANTENNA POWER 30KY/
_
-
I I 1 I
JANUARX 1954
_
_
_
_
1-_
_
_
_
I
_
_
_
Curves
50%
show
of days
signal
level
heavy
equalled
middle
_
_
or
curl
_
_
_
I
I-
_
10%
_
_
_
_
_
-
90%
-
_
-
_
):-
:
iIIII 1
FEBRUARY, 1954
r
_
_
ceeded
90*/
:
10%
of
1
of days
days
(uppercurve),
(lower
I1
curve)
1
'10%
i
I
I
1
I 50%
..
I
I
.r......--....
90%
A
?I
-1
-
---4 I ---.:
'
IIIII
NOVEMBER, /953
50:
_
_
40_7-
_
30:-.
?
20L.......,../50%
10:
10%
_.
...?
90%
,4*..0*
. I
of
data
. I
ail data,
attar
, I
mdudo,g.
silialna
, I
Ind, b
, -I
?Sams
I
:
-
1 1 1 1 1
DECEMBER, /953
1
_
_
^
.
-
1
_
:
-
-
_
_
:
_
_
_
_
-
_
-
_
-
_
-
-
-
-
10%
_.
-
-
-
_
:
_
_
:
-
50%
_
**4
?\ '1/4...
90%
-
o affeclad
4srpoletien,
1-.11111
by
Es-
Es-type
11111
pp. crosegotise
prevesterlee
1
I
I
I
i
:
-
I-
7
T1T -Y '
MARCH, 19541
,
i !
1 i
I
-
-
-
10%
50%
I
9 0%
,
00 04 08 12 16 20
I I
APRIL, /954
? Analysis of a I data, including values o (acted by Es ?type propagation
--Same data after eliminating, by interpolation, Es ?type propagation
10%
LOCAL TIME AT PATH MIDPOINT
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
CEDAR RAPIDS TO STERLING ? 49.8 Mc/s
PATH LENGTH: 1243 km
' DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN- CIRCUIT ANTENNA VOLTAGE - 600 a ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40KW1 ESTIMATED ANTENNA POWER 30KW
bt)--r-
:
,
MAX 1954
-
:
50:
:
40:
10%
la
.
30:
20:
10
....-
-k.
?
s.,/
e
50%d
I
I
v.
9
I
0 %
riri

'0(
na oft,
1
JULY, 1954 Ii..
iIII
I
10%
507,
III)
0
90%
show
of
signal
cloys
I
Curves
50%
level equalled a
heavy middle cur
i
--Analysis of all data, inclixfing
--Same data after eliminating, I
2C
JUNE, 1954
_
:
:
10%
_
-
,
:
-
00
vol
08 12 16 20 00 00 04 08
04
12
00
Ii 1 1
AUGUST 1954
.
:
_
-
7
7
-
-
7
-
-
-
7
:
-
7
7
-
-
-
:
7-
7
10%
-
-
_
7
-
-
:
-
-
_
-
_
_
-
-
;loaded
90%
10%
of
.
of days
days
.
(uppercurve),
(lower
.
.? curve)
90%
_
_
2
_
_
Ws of leasd by Es -type propagation
Interpolation, Es -type propagation .
,
,
,
,
_
_
-
-
LOCAL TIME AT PATH MIDPOINT
FIG. 6
12
16 20 00
6
5
4
3
2
CEDAR RAPIDS TO STERLING ? 49.8 Mc./s
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN-CIRCUIT ANTENNA VOLTAGE - 600 .0, ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30KW
.
I !III
SEPTEMBER, 1954
.
:
.
_
-
..
_
;-
_
Curves
r
show
of days
signal
level equalled
heavy middle
,
or
cun
-
50%
_
10%
I
7
-
7.
-
50%
4 -
90%
_
7
7
00
04
08
12
1I I
NOVEMBER, 1954
16
20
OCTOBER, 1954
.
_
_
1.---
_
eeded
_ 90%
_
10% of days (upper
of days (lower
curve),
curve)
I
_
10%
_
_
-
-
1
I
I -
? -
-
00 00
04
08
12
1
16
20
00
1
DECEMBER, /954
1954
30
20
10
0
,
Analysis of all data, including vol
?Same data after eliminating, by
00
04
08
12
16
20
ues affected by Es -type -propagation
terpolation, Es -type propagation
00 00
04
08
LOCAL TIME AT PATH MIDPOINT
FIG. 7
12
16
20
00
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
CEDAR RAPIDS TO STERLING ? 498 Mc/s
CEDAR RAPIDS TO STERLING ? 49.8 Mc/s
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN - CIRCUIT ANTENNA VOLTAGE - 600 ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30 KW
1.I
:
_
III
!III 'III,
JANUAR1' 1955
I
I
T.
-
_
.
:
_
-
.
-
:
:
?
-
.
-
.:-
Curves
show
of days
signal
(heavy
leve
equalled
middle
-
or
cu n
.
)"..
_
7
507
7.-
1:
-
_
10%
_
-
_
50%
\\.......?.
_
-
:
ill
90%
,1????
1
-
/"..4
-
_
:
.
-
-
II IIrrr
FEBRUARY, 1955
exceeded 10% of days (upper curve),
90% of days lower curve)
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN - CIRCUIT ANTENNA VOLTAGE - 600 SI, ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30 KW
1 1
1
1
I
it
MAI;
I
ET1 r"-
1955
I
I?
t----r-
-
-
-
_
Curves
show
of
I
signal
days (heavy
level
-1-
equalled
middle
or
cur
-
-
I
50%
_
1
7
-
10%
.
7.
: -
t---2.
I --'
:_.?
i
!
i
I
-
90%
1
:
: 1
1
VT-
, I ' 1-T
JUNE,
1955
T
"
:
1
:
ceeded
90%
10%
of
of days
days
(upper
lower
curve),
curve)
1.
10%
- ?ft.
-I-
..
,
.7
_
_
:
50%
90%
I
,
? " ? I ? I ?
MARCH, 1955
Analysis of a I data, including values a fected by Es ?type propagation
:--Sarne data after eliminating, by Interpolation, Es?type propagation
10%
50%
90%
90%
MINI
T?r- I
JUL) 1955
1
,
III
I
ii
1 -.....,.
'
170
,.
lit
..z--;
50%
-f
ty,
mAl;
=II
P
/IIIIIMI
90%
_l__
III
44
PI
1
data,
sneludag
eliminating,
DIPAnalysis
1 1
1
1
of all
-Seine data after
11111111111'i
00 04 08 12 16 20 00 00 04 08 12 16 20
LOCAL TIME AT PATH MIDPOINT
AUGUST 1955
10%
iyPd Contailatian
p. Prottotstion
t I I i t
12 16 20 00 00 04 08 12 16
LOCAL TIME AT PATH MIDPOINT
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-_RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
601_
50
40
CEDAR RAPIDS TO ' STERLING ? 49.8 Mc/s
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
. RECEIVED OPEN -CIRCUIT ANTENNA VOLTAGE - 600 SI, ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30KW
I I
SEPTEMBER, 1955
9
Curves show signal level equalled or ex
50% of days (heavy middle curve),
20
30
20
10
0
-
-
-I
OCTOBER, 1955
-i
-,
1-
[
1
_
_
-,
-
_
ceeded
90%
10%
of
-
of days
days
(upper
(lower
I_
1
curve),
curve)
1
i
1
- -
k
I
i
_
-
=
7I,
I
50%
7
-I
1
90% ,
I I !
.I
I,
-
_
-
08 12 16 20 00 00 04
I I I I
NOVEMBER, 1955
z
10%
08
12
16
20
00
DECEMBER, 1955
10%
50%
nolysis of all data, including
--Same data after eliminating,
It I tit -I
values affected by Es -type propagation
by interpolation, Es -type propagation
(-1 lo
00 04 08 12 16 20 00 00 04 - 08
LOCAL TIME AT PATH MIDPOINT
FIG. 10
12
16
20
00
?
60
50
40
30
20
FARGO TO CHURCHILL? 49.7 Mc/s
PATH LENGTH: 1326 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN -CIRCUIT ANTENNA VOLTAGE - 600 .0; ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40KW; ESTIMATED ANTENNA POWER 30 KW
SEPTEMBER, 1952
OCTOBER, 1952
10?/0
00
30
20
10
0
04
08
12
16 20 00 00
I
NOVEMBER,
1
I I
1952
I
1
1

?
Emu
gm
ti
50%
2111111111
90%
\ _IlliallimE
luipm
propiimr"
Curves
show
of days
i i
Analysis
1.1,11111,
signal
(heavy
I
of
data
!eve
all data,
after
equa
midd
includingNBS
eliminating,
led or
e curv
3
h
."1
50%
...
?Same
ex
alu
04
08
12
16
20
00
1 I I I 1
DECEMBER, 1952
r
10%
I
90%
Ai.
ceded
90%
r
es, affected
tterpolation,
III
10% of days (upper
of days lower
1 1
by Es -type propagation
Es -type propagation
II tit!
curve),
curve)
I
I
00 04 08 12 16 20 00 00
04
08
LOCAL TIME AT PATH MIDPOINT
FIG. II
12
16
20
00
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
FARGO TO CHURCH I LL ? 49.7 Mc/s
PATH LENGTH: 1326 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN-CIRCUIT ANTENNA VOLTAGE - 600 SI ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30 KW
6 ill
FEBRUARY 1953
_I
-
:
I
JANUARY, 1953
:
_
-
_
-
-
_
_
-
Curves
show
of
signal
days
level
heavy
equalled
middle
or
cur
-
1:
_
50%
_
_
1-
-
50%
1
_
-
:
:
_
...,
90%
:
.
-
:
_
exceeded 10% of days (upper curve),
I, 90% of days (lower cu ye)
ANCHORAGE TO BARROW ? 48.87 Mc/s
PATH LENGTH: 1156 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN-CIRCUIT ANTENNA VOLTAGE - 600 41, ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30 KW
1 1 1 1 1
SEPTEMBER, 1952
III I I
OCTOBER, 1952
1 0 eirS
50%
10%
50%
90%
90%
04 08 12 16 20 0000 04 08 12
08 12 16 20 00 00 04 08 12 16 20 00
1 III
MARCH, 1953
10%,
1111
V.
up
4
Ipir
50%
1111
r
Li
AIIIIM
90% .--
MI
--....
?
--Some
Analysis
of all
data after
data,
eliminating,
including values
by
iIiIIIIIIIiIiIr
affected
interpolation,
by Es
Es ?type
?type
propagatron
propagation
04 08 12 16 20
LOCAL TIME AT PATH-MIDPOINT
i_
_
I 1 I I
NOVEMBER, 1952
-
-
-
-
_
-
_
'-_
E
10%
_
:,../
-
_
-
-
_
_
:
_
50%
:
:
_
90%'
:
-
_
/
Curves show signor level equalled 01
50% of days heavy middle cu
?Analysis of all data, including
?Same dar afterr etiminating,
rrrrrr,r,
)
- 411-4:ii?
-
--t r?
r
I
t
r
DECEMBER, 1952
I 0 %
9Crio
exceeded 10% of days (upper curve),
), 90% of days (lower curve)
values offer:Sad by Es ?type propagation
Y terpoi?tior, Es-type .prapartlan
08 12 16 20 00 00 04 08 12 16 20 00
LOCAL TIME AT PATH MIDPOINT
FIG. 12
FIG. 13
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
ANCHORAGE TO BARROW ? 48.87Mas
PATH LENGTH. 1156 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN - CIRCUIT ANTENNA VOLTAGE - 600 41 ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30KW
JANUARY 1953
ANCHORAGE TO BARROW ? 48.87 Mc/s
PATH LENGTH: 1156 km
DISTRIBUTION or HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN-CIRCUIT ANTENNA VOLTAGE -600 O, ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30KW
MAX 1953
50%
Curves show signal leve equalled or ex
50% of days (heavy middle curve),
? Analysis of all dots, including vol
?Same data offer eliminating, by
-
-
I 1 I
JUNE, 1953
_
I
io%
_
_
_
_
_
_
_
?-
_
_
-
=
=
-
ceeded
90%
MS affected
interpolittion,
Fil,Iii,l,l,
10%
of
by
of days
days
Es ?type
Es ?type
(upper
(lower
propagation
curve),
curve)
propagation
=
_
-
-
_
-
_
-
_
04 08 12 16 20 00 00 04 08 12 16 20 00
MIDPOINT
-
lilt!
MARCH, 1953
=
-
=
_
-
:-
:
:
-
_
Z
-
_
_
10%
_
-
:
v....--
.-
/
1
:
:
...
'
Curves
show
of
signal
days (.heavy
level
all data,
after
equalled
middle
including
eliminating,
?
or
cun
I
- ?
50%
:
-,

0
2
Ui
ILl>
090
-J
w 80
(7)
Ui
70
ANCHORAGE TO BARROW ? 24.325 Mc/s
PATH LENGTH: 1156 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN-CIRCUIT ANTENNA VOLTAGE ?600 a ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30 KW
FIVE -ELEMENT YAGI TRANSMITTING AND RECEIVING ANTENNAS USED HAVING COHERENT-PLANE-WAVE GAIN
OF APPROXIMATELY 9 DECIBELS RELATIVE TO HALF-WAVE DIPOLE
s
I I
MARCH,
I
1953
Iii
1111
I
?. ?
, ,
,,
??
.. .
V.'
v
,... ?. ,
,?
00,0
,
,..,
500,0
90%
,
,
,
,
,
,
,
60
50
40
30
20
04
08
1
16
1 1
MA)' /953
I
11111
,,...
10%
All!
1
r y /
ill
50%
-
??? j
.
_e
Mill
90?k
.
Nag
-........-
.
yr
Curves show signal level equalled or
50% of days (heavy middle cur
ilot,
4114.0.4,
t 11111.1111.1111H
Analysis of all data, including
--Same data after eliminating, I
00
04
08
12
16
20
=
-
I I I
APRIL, 1953
-::\--N
-
=
=
=
t...,
i -
_.:,...?.%
1 o%
? 50%
1
_
=
=
90%
_
=
- .
,
,
,
.
,
.
.
?
:
00
en
e),
val
y I
00 00
04 06
12
16
20
00
= 1 1
= JUNE, 1953
1_2,
_
_
IL]
_
i
_
locy.
Ai iil
-
,..
.
...
-.....,...__
1
._
50%
i
Th--
7..--''
I
---?-.....-92
90%

=
_
ceeded
90%
es affected
nterpolation,
1-1
10%
of days
by
Es
II
of days
(lower
Es -type
-type
I tit
(upper
curve)
curve,
propagation
propagation
It
04 08
LOCAL TIME AT PATH MIDPOINT
FIG. 16
12
16
20
00
?
70
60
50
40
CEDAR RAPIDS TO STERLING ? 27775Mas
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN - CIRCUIT ANTENNA VOLTAGE - 600 41, ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER BO KW
I I
APRIL, 1954
30
10%
00
40
04
08
12
16
20
-
-
-
t 1 t i i 1
:
_
_
MAY, 1954V
-
.
_
-
_
:
-
10%
1
E
-
-
-
50%
--.,
-...,
-
=
-
_
_
..- _,
? p
, e .....
....
\ ?
'.%
%..
?
90%
_
.
___
=-
=
_
-
_
E
-.mi.
_
_

00 00
itititItititit
l
:
--
i
JUNE, 1954
_
-
\/\/\
10%
_
-
-
-
_
-
_
=---
-
-
".------
-
=
-
,
_..........
\
_
_
.
_
....._
,...
A
\.,
Illagw-
, 7-
:
e
_ .
.
=
Curves
show
of
I
signal
days
I
level equalled
heavy middle
I
01 date, melwies;
alter eliminating,
or
cur
I i
t
:
50%
?
1411
:1,'
0
,
,
t
A/whets el
?Semi dela
I I,I,I,1
,,1?1
20
1
00
eel
04 08 12 16 20 0000 04 08
LOCAL TIME AT PATH MIDPOINT
04
08
12
16
itititititItitl
. 1954
JULY
-
:
_
10%
:_.
-
_
:
_
-
_
p
?
5O%'"??-?
?:
-
-A
,...,
?
-.%,
_
_
-
-N,
.
A
;
\
,....
.....
1
-
iceedec
90%
t
ma at Israel
Interpolation,
I-.1,1,Iiiii,
10%
of
?
by
of days
days
1
.Es -typo
Es -type
(upper
(lower
1
popoption
curve),
curve)
1
preemption
1
:
-
:
_
,
t
,
,
-
-
_
t ^
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
FIG. 17
12 16 20 ' 00
60
50
40
30
_I
0 cv
0
2 10
0 00
070
w 60
50
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
CEDAR RAPIDS TO STERLING ? 27.775Mc/s
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN-CIRCUIT ANTENNA VOLTAGE - 600 41 ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40KW; ESTIMATED ANTENNA POWER 30KW
1 AUGUS7; 1954
1
1 1
i I
I
i.,i0%
1
1
i
1111.1
-
-
/
Alb. /
. ir
50%
..-....,
1
.
...
90%
NH\Pr"'
,
:
_
.:
, -
40
30
20
10
04
08
12
16
20
SEPTEMBER, 1954
00 00
,
IIIII
OCTOBER, 1954
:
:
_
:
:
:
_
_
:-
_
:_
:
10%
_
:
_
hill
I
:
---
I
90%
:
-
-
?
i
show
of days
signal !eve
(heavy
equa
middle
_
-
-
_
led or
cun
Curves
50%
C-
- ,

, i.
,
,
,
?Analysis 'of all data, including
,---Same data -after eliminating, b
o I u
y I
04
08
12
16
20
00
_
_
I I I 1
NOVEMBER, 1954
-
-
:
:
.
.
10%
i%
-.:
:
:
:
50%
_
:
........
I
90%
_
--?
s
_
_
I
la"
:
_
-
eeded 10%
90% of
of days
days
(upper curve),
(lower curve)
_
_
_
_
-
-
as affected by Es -type propagation
rderpolotion, . Es -type propagation
-
_
00 04 08 12 16 20 00 00 04
08
LOCAL TIME AT PATH MIDPOINT
FIG. 18
12
16
20
00
?
70
60
0
50
c.)
LLI
z 4
0
CEDAR RAPIDS TO STERLING ? 27775Mcis
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN -CIRCUIT ANTENNA VOLTAGE - 600 41 ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40KW; ESTIMATED ANTENNA POWER 30KW
III I I
DECEMBER, 1954 -
,
10%
-
/
-'
..????
e
50%
.0
-I
:
I%
Ailli.......
907o
..?
show
of days
signal
(heavy
of
data
leve
all data,
after
equalled
middle
including
eliminating,
-,
or
cun
b
1
44>
/ i
Curves
50%
I
i
i
?Analysis
---Same
111111111/1.
00
04
08
12
16
20
e)
vat
00 00
i
I I I 1 :
JANUARY, 1955
:
ni:
i,
0,--
0
_
._
_
_
A10
_
_
_
r
__.
-50%
A
If
_
....A&
115110111111r111111.P-
.
-::-
W
:
ceeded
90%
its affected
Interpolation,
1 '11111
10%
of
by
of days
days
Es -type
Es - type
(upper
(lower
propagation
I 11111
curve),
curve)
propagation
-
-
-
1
1 /
1
1
-
-
-
-
04
08
LOCAL TIME AT PATH MIDPOINT
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
FIG. 19
12
16
20
00
?
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
CEDAR RAPIDS TO STERLING ? 27.775 Mc/s
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN-CIRCUIT ANTENNA VOLTAGE? 600 A ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40KW; ESTIMATED ANTENNA POWER 30 KW
MAY. 1955
Curves show signal level equalled or exceeded 10% of d
90-0% of days (heavy middle curve), 90% of days
CEDAR RAPIDS TO STERLING ? 27775Mas
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN ?CIRCUIT ANTENNA VOLTAGE ? 600 .0. ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30KW
JUNE, 1955
3ys (upper
(lower
curve)
curve),
I
10%
/-
? 50%
--
-r
90%
\
,
..
..
.
?
00 04 08 12 16 20 00 00 04 08
I
SEPTEMBER, 1955
1 1 1 1 1
OCTOBER, 1955
50%
12 16 20 00 00 04
ULY,' 195
= AUGUST, 955
?Analysts of all data Incaudi
--Same data of sr ? ImInatin
,titititi
12 16 20 00 00 04
10%
pAl
jo AO
MIMI PIA
gukunor
IR
IN
-
,
=
=
II
_
g vo ues affected by Es-type propagation
, by Interpolation, Es-type propagation
rIIIIIII II 11111111
LOCAL TIME AT PATH MIDPOINT
-
7.
1 1 1 1 1
NOVEMBER, 1955
-
:
=
-
_
-
_
-
:
=
-
=
:
=
_
_
_
=
=
_
100/.
=
-
=
_
=
_
=
----..
=
_
=
_
--4:1
_
_
-
:_
=
_
Curves
50%
show
of days
signal
(heavy
level
data,
_
_
equalle
middle
_
:44:rir-
-ii i ii
t
I
i
1
Analysis of all mitt
?Same data after ellminati
1111t1t1t-1
-I
=
-
I 1
DECEMBER, 1955
-
-
_
-
7
-
-
-
_
-
-
-
_
-
_
_
-
-
=
10%
_
-
z_
.7.
_
7.
=
=.
I ??-?---N,
50%
_
?
?
?
_
-
_
-
,
_
-
_
..-
-
-
/
-
-
_
exceeded
ve), 90%
values
by interpolation,
I-i
affected
I i
10% of
of days
by
Es
I ill
days
(lower
Es -type
-type
(upper
propagation
propagation
ilifi
curve),
curve)
_
_
_
-
i I
i
I
_
-
_
I -
00 04 08 12 16 20 0000 04
LOCAL TIME AT PATH MIDPOINT
FIG. 20
FIG. 21
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
30
20
10
0
CEDAR RAPIDS TO STERLING ?107.8 Mc/s
PATH LENGTH: 1243 km
DISTRIBUTION OF HOURLY MEDIAN OBSERVATIONS OF
RECEIVED OPEN-CIRCUIT ANTENNA VOLTAGE? 600 4/ ANTENNA
REFERENCE TRANSMITTER POWER TO ANTENNA 30 KW
I 1 1 II
? NOVEMBER, 1952
OCTOBER, 1952
1 1
. r Curves show show signal level equalled o; exceeded 10% of days (upper curve), ,
507 of days (heavy middle curve), 90%
10 %
1 I
of days (lower curve)
1
10% 7
?150Yol
0 -20
0
Lir -30 -
Z 00
LU
>
0
30
9
w 20
cn
Lu
10
0
-10
-20
-30
Op 04 08 12 16 20
U.
/011111b.
90%
08 12
16 20 00 00
rat
DECEMBER,
I I
1952
_
_
_
10
?/0
-U
UUU
50%
UUU-
90%
_
_
Analysis of oll data, including
data after eliminating,
t .-i ,?1 . 1 ,
_
:_
I ,1
: 4riCirr?
-
?
--Some
.
1 I 1 1 1
JANUARY, 1953
1
,1_
1
1
r-i
_
_
:
-
:
:
Ir
%
50%
:
_
90?/0
_
_
'es affected
nterpolation,
r,
by
lit,iii.
Es ?type
Es ?type
propoga
propagation
/
ion
cc
cccli
LOCAL TIME AT PATH MIDPOINT
FIG. 22
12
16
20
00
49.8 Mc/s
STERLING ?
S1381330 'amrsinos-33NVISIO-3S113ANI 01 3 Al1V138 SSO1
fo) f2 c?, 51 8 2 2 8 f2 8 g ?.. 2 2 2
I i I I I 1 I I I I 1 I
S1381330 'sscri NOISSIVISNV81
0 0
a 2 2 0 0 0 0 0 0 0
I, CO 01 N .0 10 4D
OF SIGNAL INTENSITY
MEDIAN VALUES
CEDAR RAPIDS
DISTRIBUTION
CUMULATIVE
I 1
I
I
1
1
1
NE, JULY, Al
I I I
,
I i
i
,
,
/
1
i
I
MARCH, APRIL 1951
0
co
I II
I
I
I
I
i
R, DECEMBER
, FEBRUARY
1 1 I
is
is
,
I
I
FALL
'EMBER. OCTOBER
/
I
I
I
I
I
I
I ?
c, 0 ? 8 2 2 00
0)
(SPINO 009) 110A0HOW1 3140 3A08V S1381330 ?A.LISN31N1 1VN9IS
20 30 40 50 60 70 80 90 95
0
0
On
30 40 50 60 70 80
0
OJ
0
INTENSITY EQUALS OR EXCEEDS ORDINATE
C?l
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
60
.50
40
"cr,
2
30
20
10
0
Lu
> 10
co
< 60
Lu
03 50
C,
Lu
g40
1.0. 30
I? ?
a 20
10
10
0.1 Q2 05 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99995 998999 01 02 Q5 I 2 5 10 20 30 40 50 60 70 80 90 95 98 99 995998999
PER CENT OF TIME HOURLY MEDIAN SIGNAL INTENSITY EQUALS OR EXCEEDS ORDINATE
CEDAR RAPIDS TO STERLING ? 49.8 Mc/s
CUMULATIVE DISTRIBUTION OF HOURLY MEDIAN VALUES OF SIGNAL INTENSITY
?ALL VALUES, INCLUDING VALUES AFFECTED BY Es?TYPE PROPAGATION
--- INTERPOLATED VALUES USED DURING HOURS WITH Es?TYPE PROPAGATION
SPR/NG
MARCH, APRIL 1952
t_
FALL
SEPTEMBER, OCTOBER 1952
^
I I
SUMMER
I
MAY, JUNE, JULY, AUGUST 1952 ?
??????
140
150
160
II WINTER 1 I I
NOVEMBER, DECEMBER 1952
JANUARY, FEBRUARY 1953
^
170
180
190
200
140
150
ISO
170
180
190
200
FIG. 24
TRANSMISSION LOSS, DECIBELS
60
70
80
90
100
110
120
60
70
80
90
100
110
120
?130
LOSS RELATIVE TO INVERSE-DISTANCE-SQUARED, DECIBELS
(7,
2
60
50
40
30
20
10
X
w 0
6
?10
60
Lu
50
. 40
30
20
V)
10
0
10
01 Q2 0.5 1 2
S
CEDAR RAPIDS TO STERLING ? 49.8 Mc/s
CUMULATIVE DISTRIBUTION OF HOURLY MEDIAN VALUES OF SIGNAL INTENSITY
?ALL VALUES, INCLUDING VALUES AFFECTED BY Es?TYPE PROPAGATION
---INTERPOLATED VALUES USED DURING HOURS WITH Es?TYPE PROPAGATION
SPRING
MARCH, APRIL 1953
FALL
SEPTEMBER, OCTOBER 1953
SUMMER I I
MAY, JUNE, JULY, AUGUST 1953 =
????
I IWINTER
I
NOVEMBER, DECEMBER 1953
JANUARY, FEBRUARY 1954
140
150
160
170
180
190
200
140
150
160
170
180
190
200
5 10 20 30 40 50 60 70 80 90 95 98 99 995998999 0.102 0.5 I 2 5 10 20 30 40 50 60 70 80 90 95 98 99 945998999
PER CENT OF TIME HOURLY MEDIAN SIGNAL INTENSITY EQUALS OR EXCEEDS ORDINATE
FIG, 25
TRANSMISSION LOSS, DECIBELS
60
70
80
90
100
110
120
60
70
80
90
100
110
120
130
LOSS RELATIVE TO INVERSE-DISTANCE-SQUARED, DECIBELS
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
60
50
40
? 30
0
0
0
5?- 20
10
w 0
0
10
0
co
? 60
tn
Ui
co ? 50
Ui
CEDAR ?RAPIDS TO STERLING ? 49.8 MCA
CUMULATIVE DISTRIBUTION OF HOURLY MEDIAN VALUES OF SIGNAL INTENSITY
?ALL VALUES, INCLUDING VALUES AFFECTED BY Es ?TYPE PROPAGATION
--INTERPOLATED VALUES USED DURING HOURS WITH Es?TYPE PFt0PAGATION
II I
SPRING
MARCH, APRIL 1954
FALL
SEPTEMBER, OCTOBER 1954 -
c7i
10
0
-10
0.1 0205 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99995998999 0.1 02 0 5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 995 99899.9
PER CENT OF TIME HOURLY MEDIAN SIGNAL INTENSITY EQUALS OR EXCEEDS ORDINATE
SUMMER
MAY JUNE, JULY, AUGUST 1954
WINTER
NOVEMBER, DECEMBER 1954
JANUARY, FEBRUARY 1955
140
150
160
170
180
190
200
140
150
160
170
180
190
200
FIG. 26
?
TRANSMISSION LOSS, DECIBELS
60
70
80
90
100
110
120
60
70
80
90
_ 100
110
120
130
LOSS RELATIVE TO INVERSE-DISTANCE-SQUARED, DECIBELS
90
? 80
70
60
G 50
o ? 40
30
20
1
10
0
10
8 60
? 50
40
^ 30
20
10
0
10
0.1 0.2 05 1 2 5 10 20 30 40 50 60 70 80 90 95 98 9999.599899.9 01 02 05 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99995 99.8999
PER CENT OF TIME HOURLY MEDIAN SIGNAL INTENSITY EQUALS OR EXCEEDS ORDINATE
a
CEDAR RAPIDS TO STERLING ? 49.8 Mc/s
CUMULATIVE DISTRIBUTION OF HOURLY MEDIAN VALUES OF SIGNAL INTENSITY
--ALL VALUES, INCLUDING VALUES AFFECTED BY Es ?TYPE PROPAGATION
-- ?INTERPOLATED VALUES USED DURING HOURS VAIN Es?TYPE PROPAGATION
SPRING
MARCH APRIL 1955
FALL
SEPTEMBER, OCTOBER 1955
SUMMER
MAY JUNE, JULY, AUGUST 1955
?
140
150
160
170
180
_ 60
WINTER
NOVEMBER, DECEMBER 1955
160
170
180
190
200
FIG. 27
70
80
90 5
03
100 W
0
Ito
120 tla
60
7 II
0
80 1?
a
90 rui
cc
cr)
100 g
110
120
130
?
SIGNAL INTENSITY, DECIBELS ABOVE ONE MICROVOLT (600 OHMS)
60
50
40
30
20
10
-10
;70
60
50
40
30
20
10
-10
0.1 Q2 Q5 1 2 5 10 20 30 40 50 60 70 80 93 95 99 99 99.5 V.13 999 01
PER CENT OF TIME HOURLY MEDIAN SIGNAL INTENSITY EQUALS OR EXCEEDS ORDINATE
ANCHORAGE TO BARROW ? 48.87 Mc/s
CUMULATIVE DISTRIBUTION OF HOURLY MEDIAN VALUES OF SIGNAL INTENSITY
-ALL VALUES, INCLUDING VALUES AFFECTED BY Es-TYPE PROPAGAT1CN
---INTERPOLATED VALUES USED DURING HOURS WITH E.-TYPE PROPAGATION
I lIA4E4R1 I I
NOVEMBER, DECEMBER 1951
JANUARY, FEBRUARY 1952
_
MAY,
JUNE,
SUMMER
1952
?
JULY,
AUGUST
--- --..
-
N..
N. ss
, ????ss.
ss
?,.
'....".."?,...........................................?............
_
-
-
I I
SPRING
I I
MARCH, APRIL 1952
FALL
SEPTEMBER, OCTOBER 1952
140
150
160
170
FIG. 28
180
190
200
130
140
150
160
170
180
190
200
TRANSMISSION LOSS, DECIBELS
60
70
BO
90 vt
co
100
LU
1 1 0 ir
120 1
0
50LUI
tr
60 >
0
70
80 GI
cc
90 P3
-J
100
110
120
-1130
?
7.
SIGNAL INTENSITY, DECIBELS ABOVE ONE MICROVOLT (600 OHMS)
70
60
50
40
30
20
10
10
70
60
50
40
30
20
10
-10
0.1 02 05 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 995 996999
PER CENT OF TIME HOURLY MEDIAN SIGNAL INTENSITY EQUALS OR EXCEEDS ORDINATE
FIG. 29
ANCHORAGE TO BARROW ? 48.87 Mc/s
CUMULATIVE DISTRIBUTION OF HOURLY MEDIAN VALUES OF SIGNAL INTENSITY
-ALL VALUES, INCLUDING VALUES AFFECTED BY E.- TYPE PROPAGATION
---INTERPOLATED VAWES USED DURING HOURS WITH E5-TYPE PROPAGATION
; q / ER I I I
111111:11111 NOVEMBER, DECEMBER 1952
JANUARY, FEBRUARY 1953
IIu
El MEIN MIME!
11111111.1111411111111111
11111.11.1111111 1.1114111
SUMMER
MAY, JUNE 1953
1
SPRING
I
130
MARCH, APRIL 1953 140
150
160
161 INE11 1:1111111111
Mill MINIM
1111 11111nbillell?
111111 11111 IllIlili
170
180
190
200
0.1 02 05 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99995 998 999
130
140
150
160
170
180
190
203
TRANSMISSION LOSS, DECIBELS
70
80
110
RELA
130
I.
TRANSMISSION LOSS, DECIBELS
50 .71
60
LU
70 g
so 2
90
100>LU
Ui
110
1.1
120
Ui
cc
130 '4
0
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
FARGO TO CHURCHILL ? 49.7 Mc/s
? CUMULATIVE DISTRIBUTION OF HOURLY MEDIAN VALUES OF SIGNAL INTENSITY
? ALL VALUES, INCLUDING VALUES AFFECTED BY Es - TYPE PROPAGATION
-- INTERPOLATED VALUES USED DURING HOURS WITH Es - TYPE PROPAGATION
1 I 'whrER 1
NOVEMBER, DECEMBER 1951
JANUARY, FEBRUARY 1952
II Si!MitfER: l I
MAY JUNE, JULY, AUGUST 1952
10 20 30 40 50 60 70 80 90 95 98 99 9a5 99B999 0.102 0.5 I 2 5 10 20 30 40 50 60 70 80 90 95 98 99 9S599899
PER CENT OF TIME HOURLY MEDIAN SIGNAL INTENSITY EQUALS OR EXCEEDS ORDINATE
FARGO TO
CUMULATIVE DISTRIBUTION OF
?ALL VAUJES. INCLUDING
--- INTERPOLATED VALUES
CHURCHILL ? 49.7 Mc/s
HOURLY MEDIAN VALUES OF SIGNAL INTENSITY
VALUES AFFECTED BY Ey - TYPE PROPAGATION
USED DURING HOURS WITH E1-TYPE PROPAGATION
IiviA;reh I I
I I I
NOVEMBER, DECEMBER 1952
JANUARY, FEBRUARY 1953
010.2 05 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99995998999
5 10 20 30 40 50 60 70 80 90 95 98 9999.5998999
PER CENT OF TIME HOURLY MEDIAN SIGNAL INTENSITY EQUALS OR EXCEEDS ORDINATE
+30
+20
+10
Co
2
0
-10
20
cc
2
la -30
-J
?c( -10
2
c7/
?
-20
-30
-40
0 10.2 0.5 1 2 5 10 20 30 40 50 60 70 eo 90 95 98 9999599999.9 0102 05 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.593.8999
PER CENT OF TIME HOURLY MEDIAN SIGNAL INTENSITY EQUALS OR EXCEEDS ORDINATE
CEDAR RAPIDS TO STERLING ?107.8 Mc/s
CUMULATIVE DISTRIBUTION OF HOURLY MEDIAN VALUES OF SIGNAL INTENSITY
ALL VALUES, INCLUDING VALUES AFFECTED BY E.-TYPE PROPAGATION
INTERPOLATED VALUES USED DURING HOURS WITH Es-TYPE PROPAGATION
SPRING
MARCH, APRIL 1952
? FAIL
EPTEMBER, OCTOBER 1952
summEk
MAY, JUNE JULY, AUGUST 1952
_1
WINTER '
NOVEMBER, DECEMBER 1952
JANUARY 1953
180
190
200
210
220
230Ui
S
5
240 Ia0
Co
180
5
190 Co
cc
200
210
220
230
240
FIG. 32
a
100
110
120Ui
9
130
. cr
UI
140 cc
Cl)
150 ed
I
100 z
110
3
120 ca
cc
1309
140
150
a
80
70
60
50
G. 40
2
8 30
CD
0 - 20
2 ? 10
14 II 0
5
8 100
T., 90
41
CO
5
ao
0
rf; ,70
60
-1
?
CD
Co 50
40
30
20
10
?
CEDAR RAPIDS TO STERLING ? 27.775 Mas
CUMULATIVE DISTRIBUTION OF HOURLY MEDIAN VAWES OF SIGNAL INTENSITY
?ALL VALUES, INCLuDia* VALUES AFFECTED BY Es -TYPE PROPAGATION
---INTERPOLATED VAUJES USED DURING HOURS WITH Es-TYPE PROPAGATION
SUMMER
MAY JUNE JULY, AUGUST 1954
- ?
WINTER
NOVEMBER, DECEMBER 1954
JANUARY 1955
?
?
0.1 02 0.5
FALL
SEPTEMBER, OCTOBER 1954
SUMMER
MAY, JUNE, JULY, AUGUST 1955
?
120
130
140
150
160
170
180
100
110
120
130
140
150
160
170
2 5 10 20 30 40 50 60 70 80 90 95 98 9999.5938999 01 02 as 1 2 5 10 20 30 40 50 60 70 BO
PER CENT OF TIME HOURLY MEDIAN SIGNAL INTENSITY EQUALS OR EXCEEDS ORDINATE
FIG. 33
180
90 95 98 99 99.5 948999
TRANSMISSION LOSS, DECIBELS
Cl
- 60
70
80
90
100
110
20
30
40
50
60
70
80
90
100
-110
LOSS RELATIVE TO INVERSE-DISTANCE-SQUARER DECIBELS
Declassified in Part - Sanitized Copy Approved for Release a 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
1-?
UI
8 80
cc
? 70
Ui
w 60
co
50
9
Ui
UI
? 40
UI
^ 10
CEDAR RAPIDS TO STERLING ?27775 Mc/s
CUMULATIVE DISTRIBUTION OF HOURLY MEDIAN VALUES OF SIGNAL INTENSITY
?ALL VALUES, INCWOING VAWES AFFECTED BY Es-TYPE PROPAGATION
---INTERPOLATED VALUES USED DURING HOURS WITH E3-TYPE PROPAGATION
FALL
SEPTEMBER, OCTOBER 1955
0
WINTEfl? ?
NOVEMBER, DECEMBER 1955
100
?30 5
co
t7.5
?40 c)
UI
120Ui
cc
?50 <
cn
130
?60 w
140
?70 5
150 z
0. ?80 LT,
tr. >
5 z
160 u.
z ?90 0
a 1-
cr
I- w
170 >
?100 5
..J
w
180 cc
^110 u)
u.
0
..3
0.102 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 9999.599.899.9 010 2 05 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.5 99.899.9
PER CENT OF TIME HOURLY MEDIAN SIGNAL INTENSITY EQUALS OR EXCEEDS ORDINATE
FIG. 34
a
?
+100
+70
re
+60
Ui
w +50
UI
4
+40
Cd
UI
a +30
UI
4
0 0
ANCHORAGE TO BARROW ? 24.325 Mc/s
CUMULATIVE DISTRIBUTION OF HOURLY MEDIAN VALUES OF SIGNAL INTENSITY
?ALL VALUES, INCLUDING VALUES AFFECTED BY Es -TYPE PROPAGATION
--INTERPOLATED VALUES USED DURING HOURS WITH Ey -TYPE PROPAGATION
I
SPRING
MARCH, APRIL 1953
a.
I I
SUMMER
MAY, JUNE 1953
Qs.
????
?????
110
120
130
140
150
160
170
180
190
200
210
10W02 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 9999599.899.9 0.102 05 I 2 5 10 20 30 40 50 60 70 80 90 95 9e 9999699.899.9
PER CENT OF TIME HOURLY MEDIAN SIGNAL INTENSITY EQUALS OR EXCEEDS ORDINATE
FIG. 35
TRANSMISSION LOSS, DECIBELS
20
30
40
50
60
70
80
90
100
10
20
LOSS RELATIVE TO INVERSE- DISTANCE-SQUARED, DECIBELS
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
SEASONAL VARIATION OF SIGNAL INTENSITY
MONTHLY MEDIAN VALUES OF SIGNAL INTENSITY FOR THREE-HOUR
PERIODS CENTERED AT 00, 06, 12, AND 18 HOURS, LOCAL TIME AT
PATH MIDPOINT, AND LEVEL EXCEEDED 95% OF HOURS DURING MONTH
40 CEDAR RAPIDS -STERLING 1243 km
,
(a) 00 h
3 0-
2 0
1 0
0J
40
(b) 06h
-30-
C,)
i20
8 IO-
W
0
8 40
> (c) 12h
o '2 ri
CC
0
2 0-
ILl
oz 10-
Lii 0
m 40
_u339
ci3 20
010
1\--Z\Z?SEE SCALE ON RIGHT
omit
til1
I s1 1:1MJSNIJMMJSNIJAIMJSNJ
1 1 1 1 1 1 I 1 1 1 1 1 1 I 1 1 I I III 1 I I 1 1
--10
NJA4MJSNJMNJ SNJMMJ SNJMMJ SN
0
? 30
20
10
0
MJSNJMMJSN
JMMJSN
JMMJSNJMMJSN
20
-10
20
MMJS.NIJ hiMJ S NJ MMJ S NJ
' ,; " 20
m SN
?10 t
jittit_tt(trit
MMJSNJMMJSNJMMJS
1951 1952 1953
FIG. 36
It !lit
JMMJ SN
1955
-10
-20
111
SEASONAL VARIATION OF SIGNAL INTENSITY
MONTHLY MEDIAN VALUES OF SIGNAL INTENSITY FOR THREE-HOUR
PERIODS CENTERED AT 00, 06, 12, AND 18 HOURS, LOCAL TIME AT
PATH MIDPOINT, AND LEVEL EXCEEDED 95% OF HOURS DURING MONTH
FARGO-CHURCHILL 1326 km ANCHORAGE-BARROW 1156 km
50 1 1 1 1 1 1 I 1
(a) 00h
4 0 -
3 0
20
IOJ S;,1
49.7 Mc/s
t III!
11i
I I I I
I I 1
1 I I
JMMJSNJMM JSNJMMJSN1JMM
50
(b) 06h
-40-
C/)
2
30-
o
8 20-
w
I 0 "
SN
50
(c) 12 h
cc? 40-
o
30-
w
z 20-
o
11-1> I 0
m 50 ,
(d) 18h
cn 40-
-J
(23- 30-
o
LLI
0 20-
1 I I
I it I II
JMMJSN
I I
JMM
1 1 1 1 1 1 1 I
I
SNJMMJ SNJMM
10
I
I
1
SNJMMJ SN
40 ,,,, I I
(e) WEAKEST SIGNAL:
30_ LEVEL EXCEEDED
95% OF HOURS
20-
10-
42>
n
- s N
1951
111111
J MMJ SN
1952
I t
SN
1 1 1
I 1111t11
MMJSNIJ
I 1
SN
I
JMMJSN
^
1 1
JMM
I I I
I 1 1
^
JMM JSNJMMJSNJMM
I
JMM JSN
1953 1951
FIG. 37
I I I
111
at
JMMJSNJMM
1952 1953
^
?
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
COMPARISON OF FIVE YEARS OF OBSERVATIONS OF SIGNAL INTENSITY
. CEDAR RAPIDS TO STERLING - 49.8 Mc/s
30
20
10
30
20
10
0
MONTHLY MEDIANS OF HOURLY MEDIAN VALUES OF
RECEIVED OPEN-CIRCUIT ANTENNA VOLTAGE -600 SI ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 kW; ESTIMATED ANTENNA POWER 30 KW
OBSERVATIONS AFFECTED BY Es ?TYPE PROPAGATION OMITTED;
INTERPOLATED VALUES USED WHEN AVAIL ABLE
_ I
-
i
i
I
i
1
I
i
r
I i _
MARCH -
?? 1951
1952
? 1954
_
_
---1953
_
_
-
-
_
1955
/._
-
_
_
..2........
_
-
_
/-..-P----.
- ....
::/
'? ??? .
,
...-,
--?
-? ...
...,,....-???....=
'_N \
,\
N
_
_
_
_
_ ..
.
..----
1
....
\-..
.......,.........________.----4
_
i 1 -
_...??
--
_
_
_
- 1
I
I
I
I
I
I
I
1
- 1
_
_
-
I
I
1
1
.
.
I
......
1
?.......
I.
1
1
I
1
_
_
_

JUNE
_
-
?
?
__
.? .....
--?------
..
, -
......?
'.
.? ?-? S.
??
,??????,
N
?
6 .6
N N..6
..
_
-
?
?
-el---
-4.
\
._
.-..\-
....\
- !...
. -.?.
?
.?
...
...
%

.. .-
?
._;?.,"
_.
7
, -
-
?
_
_
-/'
-
----??
-
_
??????????
-------
??
--1954
1951
1952
? 1953
_
-
-
_
?
-
_
__
1955
- ?
- 1
1
1
1
I
1
1
1
1
1
1
1
-
-
00 02 04 06 08 10 12 14 16
LOCAL TIME AT PATH MIDPOINT
FIG. 38
18
20
22
00
1.
DECIBELS ABOVE ONE MICROVOLT
COMPARISON OF FIVE YEARS OF OBSERVATIONS OF SIGNAL INTENSITY
CEDAR RAPIDS TO STERLING- 49.8 Mc/s
30
20
10
0
30
20
10
0
MONTHLY MEDIANS OF HOURLY MEDIAN VALUES OF
RECEIVED OPEN-CIRCUIT ANTENNA VOLTAGE -600 ANTENNA
REFERENCE POWER INPUT TO TRANSMITTER FINAL AMPLIFIER 40 KW; ESTIMATED ANTENNA POWER 30 KW
OBSERVATIONS AFFECTED BY Es ?TYPE PROPAGATION OMITTED;
INTERPOLATED VALUES USED WHEN AVAIL ABLE
_
_
-
_
I
1
I
i
I
I
I
i
I I I
-
=
-
SEPTEMBER'
??? ... 1951
- 1952
-
_
_
-
'
-
-
_
-
? ? 1953
--1954
-
_
_
. ? ? '
. ? ? *.---
.....
?
:::.*-'4?.,....
...
.??
-
-
-
1955
_
_
_
-
.
..? . .... ?
*-- .....4....
..... .? ??
.,,4?
?
? .
,
.
.
.
?.....-?
...-
? _/
..?.'4.'
?
.--
-..,.
, \. ..?.,
- .. s
...,
?N ' -'::.-
*S.'S.
N....,
? ._
-
-- '
--s.
NC....
............>:,.........
...,...
........
'.......
.
..i.?01.?L oo
we '
? ?????????? .../...."
.../..
...
woo.
?
?????
..
.......
.../".
oo.
I
I
1
I
1
1
I
1
I
r
1
1
_ 1
-
-
1
1
I
1
1
1
I
1
_
DECEMBER'
-
-
_
.,
???-?'"
....
../
?'??
.
. ..
? ...
? . .
_
_
_
_
_
...???'
i .?
i .?
i ?
*
?
N
-.\
_
_
_ ??
,...?
1-:?:-. .......___
>.. ----...4,
'??? --....
...
r?
__ .
/
? ?
?
-
.
\ .. \
?
-

---
- ---=_
_
- 1952
? ? ?1953
? ? 1954
-
_
------?
?
N .
/..
?-
_
*----*
-
_
-
- I
1
1
1
1
1
1
..... .....-?
..?-?
_
_
1955
1
1
00
02
04 06 08 10 12 14 16
LOCAL TIME AT PATH MIDPOINT
FIG. 39
18
20
22
00
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
LONG?TERM VARIATION OF SIGNAL INTENSITY
I2-MONTH RUNNING MEANS OF MONTHLY MEDIANS OF RECEIVED
SIGNAL INTENSITY FOR THREE-HOUR PERIODS CENTERED AT 00, 06, 12,
AND 18 HOURS, AND OF LEVEL EXCEEDED 95% OF HOURS DURING MONTH
CEDAR RAPIDS?STERLING 49.8 Mcis 1243 km
(6)00h
14-
12-
10-
I
11
8 m111111111111111 11111
MJSNJ MMJSNJMMJS NJMMJSN
16
Io
0
>
22
I" 20
8
111111
111111
111
.114MJSN
JMMJ
1111111111
-(b) 06h
1111111
1111
(I) I2-MONTH SMOOTHED RELATIVE SUNSPOT NUMBER
' ' 1 " I .(2) I2-MONTH SMOOTHED INDEX OF MAGNETIC ACTIVITY, Kp s
(C) 12 h .---,-. /
--..,.. /
4\ ...,
\,........ (V.'s's\
orN...,....... N- _
1111
140
120 inW
X >
0Z F-
10
1?
6800
Q--26 F.:
cn
-24 zIc:j
M -22 04
-40
w 2
> -20
20 tt -I8?
-J
0 IT 18 '6
0
(d) 18h
11111111111111
1
111
(e) WEAKEST SIGNAL?
10 - LEVEL EXCEEDED
95% OF HOURS
6 -
4
1951
1952
111
I
M MJSNJPAI 11111 11
MJSNJMMJSN
11111 111
1953 1954 1955 1956
JMMJ
FIG. 40
SMOOT
?
44,
40
?
?
?
?
APPENDICES
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
?
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
APPENDIX I
MEASUREMENTS OF FREQUENCY DEPENDENCE
PART I - REGULAR PROGRAM OF VOLTAGE RECORDING
R. M. Davis, Jr.
R. C. Kirby
1. INTRODUCTION
The various theories of the scattering of VBF radio waves give ex-
pressions for transmission loss in which received power is inversely pro-
portional to some exponent of the transmission frequency. 1,2 The previous
reports35 and general summery' have given earlier experimental data on
the value of this exponent n under particular circumstances. This report
gives the results of further measurements and indicates the possible sources
of error and limitations of the experiment.
The earlier studies showed that the exponent n, representing the ratio
of received powers at the two frequencies 49.8 Mc/s and 107.8 Mc/s, varied
from about 4 to 12. The median value of the exponent over an extended
period was 7. Under conditions of strong Signal intensity at 49.8 Mc/s
the frequency exponent was higher, in the vicinity of 8. While several
factors, such as sporadic-E and tropospheric propagation, could account
for many values in the wings of the distribution of n, it was concluded
that there was some genuine variation of the exponent with changing con-
ditions in the propagation medium.
2. EXPERIMENTAL PROGRAM
The experimental details of the program have been described in the
previous reports1,2 During the observing period to be reported here the
sane methods of frequency dependence observation were followed. Trans-
mitting and receiving antennas were scaled in proportion to their wave-
lengths so that the geometrical factors were identical for each pair of
frequencies. The aperture of the receiving antenna at the higher fre-
quency is smaller than that of the lower by a factor equal to the squnre
of the frequency ratio. An adjustment for "constant aperture" was made;
the ratio of the antenna apertures(in decibels) was subtracted from the
Observed ratio of signal intensities for scaled antennas.
Each ratio of signal intensity for a pair of frequencies was found
by comparing the median intensities recorded during simultaneous periods
of an hour. The values for such periods were adjusted to insure that
they applied to equal radiated power, equal transmission line losses, and
equal receiving antenna impedances. Thus adjusted, the decibel ratio of
two signal intensities can be interpreted as a ratio of available powers,
provided the signal voltages measured are effective root mean square values.
A test of this last assumption is made in Part II of this appendix.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
3. RESULTS
- 3.1 Observed Frequency Exponents
Figure I-1 incorporates in summary form all of the observations of
frequency dependence previously unreported! In each panel the histogram
shows the distribution of signal intensity-ratios and frequency exponents
found when signal intensities at the indicated frequencies were compared.
In the top row all of the observations for each pairof frequencies are
included. In the middle row of panels only cases of true scatter propa-
gation are retained, unadulterated_ by sporadic E or other recognizable
anomalous conditions. The influence of sporadic E is eliminated by dis-
carding both the hours of recognized Es occurrence and the adjacent hours.
Hours of abnormally high absorption/ affecting different frequencies by
different amounts, are also discarded, as are hours when auroral, or
'sputter," propagation was present. Finally in the bottom row of panels
attention is confined to the five hours centered approximately on local
noon at the path midpoint. These are the hours of typically strongest
signals and. presumably little influence of aspect-sensitive meteor re-
flections; only hours uncontaminated by sporadic E or abnormal absorption
are used in these distributions. The special circumstances of the fre-
quency-dependence recording of each circuit will now be mentioned.
a. Cedar Rapids to Sterling, 107.8 Mc/s vs. 49.8 Mc/s, rhombic to rhombic
This is a continuation for the four months, October, 1952 through
January, 1953, of the study-reported. previously, which covered. January
through September, 1952. It will be noted that the shape of the distri-
bution of all observations (top panel), the median frequency exponent of
6.9, and the range of exponents is quite similar to the earlier data.
The character of the distribution is affected very little where sporadic-E
hours are excluded in the middle panel. At the bottom, however, the dis-
tribution is shifted to the right and a higher -median exponent, 7.8, re-
sults when the smaller sample of data for the five hours around noon is
considered.
Operation of the scatter transmissidh equipment at 107.814c/6 pre-
sented special difficulties; the frequency stability and recording band-
width were inadequate for reliable recordings at 10w signal levels. For
the period reported in Figure :-.a, however, the experimental data. are felt
to be of comparable quality with those at 49.8 Mc/a.
b. Cedar Rapids to Sterling, 49.8 Mc/s vs. 27.775 Mc/s, rhombic to rhombic
The middle column of histograms of Figure I-1 presents the frequency
?
dependence found by comparing the signals received-at these two frequencies.
Again all system parameters have been adjusted to equal values for the two
frequencies. Signal intensity ratios for scaled:amtennas at the two fre-
quencies are shown an the bottom scale, and. :the corresponding frequencyex-
ponents, after adjustment to unit aperture of the receiving antenna, are
shown along the top scale.
App I page 2 -
The histograms show that systematically lower values of frequency
exponent result from the comparison of intensities 27.775/49.8 than from
49.8/107.8. Where a median of 6.9 is applied to the higher pair of fre-
quencies, the median exponent is 4.7 for the lower pair. This is true
whether all observations are considered, as in the top panel, or only
those free of sporadic E and high absorption hours, as in the middle
panel. When midday hours with their generally strong signals are the
basis of comparison, a slightly higher value of exponent, 5.1, is ob-
served. If allowance is made for the relatively high absorption at
28 Mc/s, a value of n as high as 6 to 6.5 for the 27.775/49.8 comparison
may be in order for the scattering process.
No definite source of error in the experimental arrangements has
yet been found sufficient to prove that the observed variation is not
genuine. Two factors, however, tend in this direction. First, the
available power from background galactic noise at 27.775 Mc/s was as
much as 3 db less than expected on the basis of other observations.
Second, the antenna siting was not ideal, as is presupposed in the com-
parison of signal intensities.
It is felt on the whole that caution should be used in accepting
the variation of exponent with frequency. Recent preliminary results
from a more elaborate experiment indicate that a single value of fre-
quency exponent may actually hold over a wide range of frequencies.
c. Anchorage to Barrow, 48.87 Mc/s vs. 24.325 Mc/s, Yagi to Yagi
The last column of histograms in Figure I-1 shows the frequency
dependence results of four months of simultaneous recording of Yagi-
to-Yagi transmissions over an arctic path at 48.87 Mc/s and 24.325 Mc/s.
For this recording period wide dynamic range recorders were in operation,
so that all the sporadic-E signal intensities fell within the limits of
the recorder. The wide range of ratios that resulted from complete Es
recording are shown in the top panel. They will be discussed later in
connection with Figure 1-9.
The shaded portion of the Anchorage to Barrow histograms indicates
ratios for hours when high absorption was in effect, as deduced from a
comparison of cosmic noise levels at the two frequencies and from ionos-
pheric sounder records. It was believed that non-deviative absorption,
affecting the lower frequency more than the higher, would reduce the fre-
quency exponent. The figure does indeed indicate that hours of unusually
high absorption are associated with relatively low ratios. The middle -
panel shows, however, that when the low ratios for absorption hours are
discarded, as well as the ratios for hours of Es and sputter, the resulting
median ratio is the sane as for all observations, 5.3.
In the bottom panel is shown the distribution of ratios from the five
hours centered on local noon. In this case the median ratio is again 5.3.
Unlike the noontime signals on the temperate zone path, Anchorage to Bar-
row signals do not in general reach a maximum around noon. Thus noontime
ratios are not above the average for the 24 hours.
- App I page 3-
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
3.2 Correlation of Signal Intensities at Various Frequencies
In the next three figures, I-2, 1-3, and. I-4, the diurnal variation
of the median signal intensity is compared for the two frequencies in use
on a given path.. Each comparison is based on a single month's data, the
month of June in each case, though the data were obtained in three dif-
ferent years. The lower panel of each figure is a histogram of ratios of
signal intensity for the two frequencies in question. Values in the histo-
gram represent all hours where both frequencies were in operation, except
those when sporadic E was in control and no interpolated value for the
scatter level could be obtained.
It will be noted in Figures 1-2 through 1-4 that frequency exponents
varying over a range of 4 or more were recorded in substantial numbers
of cases. This does not mean that signal intensities occurring at the
two frequencies were unrelated. The medians of these signal intensities
follow the same general diurnal patterns in each of the three comparisons.
The Cedar Rapids to Sterling comparison of 49.8 Mc/s with 27.775 Mc/s
(Figure 1-3) shows the closest agreement of diurnal patterns of signal in-
tensity. For this case the median value of n was 4.3 with a range of 5.9
to 3.1 from the 10% to 90% value. Anchorage to Barrow, 48.87 vs. 24.325 Ws
(Figure 1-4) with a weaker diurnal agreement has a median exponent of 5.0
and a 10% to 90% range of 6.3 to 3.3. Finally the comparison of 107.8 with
49.8 Mc/s intensities on Cedar Rapids to Sterling (Figure 1-2) gives a me-
dian exponent of 7.2 and a 10% to 90% range of 8.4 to 6.0.
3.3 Diurnal and Seasonal Variation of Frequency- Exponents
Figures 1-5 through 1-8 portray the dependence on local time of the
frequency exponent under selected conditions. Data from the 107.8 vs.
49.8 Mc/s Cedar Rapids to Sterling experiment are shown in Figure 1-5.
The diurnal variation of signal intensity ratio and frequency exponent n
are given for two summer months, June and July, 1952, and two winter months,
December, 1952, and January, 1953. It will be seen that the diurnal vari-
ation of n bears some resemblance to the variation of the median signal in-
tensity itself. There is the same maximum around midday and the same ten-
dency-toward low values around 2000. An already mentioned, generally high
values of signal intensity-tend to result in high values of frequency ex-
ponent. Another possible tendency that seems to be borne out in Figure 1-5
is.the occurrence of higher values of exponent in winter than in summer
months. In this small- amount of data the tendency is not striking. But
more hours are marked by exponents of 8 and above in December and. January
than in June and. July. The range of observed median exponents is greater
in winter than in summer, 7 db in ratio as opposed to 5.
' Figures 1-6 and 1-7 present curves of diurnal variation of the median
frequency exponent for each month from June, 1954, through January, 1955,
Cedar Rapids to Sterling, 49.8 vs. 27.775 Mc/s.' All possible effects of
Es propagation are eliminated from the data. The exponent undergoes an
increase about 0800 local time. In June throne' September it decreases
App I page 4 -
1??
during the afternoon or early evening. In subsequent months the exponent
maintains a fairly high level, or may even increase, during the evening.
There is some evidence of a seasonal variation in the exponent. From a
median value of 4.2 in June the exponent increases to 5.2 in October. It
remains high in November and December and slips back to a medium value of
4.7 in January. While the data in this sample is scarcely sufficient for
strong conclusions, it does support the possibility of a seasonal depen-
dence of frequency exponent, with a minimum in the summer months. In the
49.8/27.775 Mc/s results the effect is undoubtedly partly due to differ-
ential absorption effects.
Figure 1-8 shows the diurnal variation of the frequency exponent for
the Anchorage to Barrow path in the equinox period of March-April and the
summer months May-June. Only data uncontaminated by sporadic E and unaf-
fected. by absorption or sputter are included in the curves. The diurnal
pattern fOr this arctic path is basically unlike the temperate zone curves
of Figures 1-5 through 1-7. The midday maximum is lacking and there is
evidence of a midnight and early morning peak. This distribution may be
accounted for by two factors. First, the intense daytime absorption in
the Arctic with its greater effect at lower frequencies would tend to
suppress a midday maximum in frequency exponent. Observations associated
with high absorption are omitted from the curves of Figure I-81 but the
same tendency may be residual in the remaining Observations. Second)
occurrences of sporadic E too weak to be identified and removed would
principally affect the nighttime hours T800 to 0600. .This would raise
the level of the 24 Mc/s signals relative to those at 49 Mc/s and would
tend to produce a nighttime maximum in the frequency exponent.
An additional tendency shown in Figure 1-8 is the low general level
of frequency exponent in May-June. The median ratio of signal.intensity
ratios is 23 in March-April compared to 22 in May-June.
3.4 Frequency Dependence During Periods of Sporadic E and Sputter
The histograms of Figure 1-9 provide an indication of the dependence
of signal intensity on frequency during the recognized contamination of
the signal by sporadic-E and sputter propagation. The two displays serve
as an enlargement of the corresponding histograms of Figure I-1.
In the upper panel ratios of signal intensity and corresponding fre-
quency exponents are displayed for the four months of simultaneous re-
cording at 48.87 and 24.325 Mc/s on the Anchorage to Barrow path. At the
top is shown the distribution of exponents during the reflection from au-
roras referred to as "sputter." The distribution is broader than that for
simple scatter propagation in Figure I-11 and the median exponent is 6.3
compared to 5.3 for scatter. Still, exponents during sputter are not sys-
tematically different from those observed during normal scatter conditions.
The upper panel of Figure 1-9 also Shows the frequency exponents de-
rived during known sporadic E. These are shown in solid. black. For the
period of the recording program a wide dynamic range recorder was in use
at Barrow. It permitted the registering of the entire range of intensities
- App I page 5 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
I
in effect during Es. Thus the black columns represent the array of actual
ratios that prevailed when one or both frequencies were affected. by spora-
dic E. Many of the values are for hours when Es was present on 24 }leis
but not on 49 Ma/s. For this reason, the Anchorage to Barrow histogram
does not strictly represent the frequency dependence of sporadic-E signal
intensities, although it is a good approximation.
The distribution of Es frequency exponents for the arctic path re-
veals a large, systematic increase over the values measured for normal
scatter. With sporadic E present on one or both of the frequencies the
median exponent is 13.2 compared. to the scatter value of 5.3. Another
study of the Anchorage-Barrow data at 48.87 and 24.325 Mc/s confirmed
the Es frequency exponent of roughly 13. Simultaneous five-minute periods
when sporadic E was recognizable at both frequencies were compared. A
median frequency exponent of 12.9 was calculated, based on all available
periods from March through June, 1953.
Figure 1-9 also contains a histogram of signal intensity-ratios for
sporadic-E and adjacent hours for the Cedar Rapids to Sterling compari-
sons of June, 1954, through January, 1955. Values shown are generally
too low because the recorders were incapable of registering the full Es
signal intensity-in many cases. Such saturation occurred much more often
at 28'Mais than at 50 lib's, so that the true ratios were equal or greater
than those shown. Under these conditions the median frequency-exponent
per unit aperture of the receiving antenna is found to be 11.8 during spor-
adic E. (Amore comprehensive study of frequency dependence during spora-
dic-E propagation observed at 49.? and 27.775 MCA on the Cedar Rapids-
Sterling path has been published. ?)
4. CONCLUSIONS -
The frequency dependence of received. power appears to depart from a
constant exponential relationship on the basis of the data reported. A
median frequency:exponent of 6.9 was found for the 107.8 to 49.8 'Mais com-
parison in the U. S., against a value of 4.7 for the 49.8 to 27.775 Ma's
comparison. In the Arctic the corresponding exponent for 48.87 vs.
24.325 Nbis was 5.3. Despite this evidence, it is felt that the indicated
variation in the frequency-exponent should not 'be accepted without further
confirmation. Certain possible sources of error in the experiment may be
responsible for the observed curvature of the frequency exponent.
The frequency-exponent undergoes a diurnal variation and appears to
be lower in summer than at other seasons. A Comparison of signal inten-
sities during sporadic-E propagation reveals 'exponentvalues very much
higher than those pertaining to normal scatter. These exponents cover a
wide range of values.
App I page 6 -
?
APPENDIX I
PART II - DEMMINATION OF FREQUENCY DEPENDENCE
EXPONENT BY TRUE POWER MEASUREMENT
G. E. Boggs and N. C. Hekimian
1. INTRODUCTION
In experimental measurements of frequency-dependence, as reported
in Part I and in the previous reports13/4 the question has arisen as to
the validity of interpretation of average-voltage measurements in terms
of ratios of received. power at the two observed frequencies. In an
effort to determine the experimentally correct value, the tests described
in this report were initiated.
To understand the basic differences between the usual averaged volt-
age recording and a true power average recording, it is well to digress
at this point to illustrate the differences encountered in a simple ideal-
ized ensemble of three numbers. Consider the set (1, 2) 3). The average
is 2 and the root average square is 2.16. Now consider a second set (2,
1, 2). The average is 1.67 and the root average square is 1.73. It should
be observed that in both cases the root average square is greater than the
average. indeed this is always the case with any sequence of values
averaged in these fashions where all of the values averaged are not iden-
tical. The order of the inequality can never reverse and is only an iden-
tity when all values are the same in the set being analyzed. The greater
the range of values, in general, the greater is the discrepancy. Applied
to field intensity recording this shows that power and voltage recording
will differ greatest with widely fluctuating signals or noisy signals so
long as the fluctuations considered occur within the averaging interval of
the recorder.
In a practical sense average power recording can be accomplished by
squaring the recorded signal prior to averaging. Since in general a
rather large range of signal level is anticipated, some compressed system
of recording is preferable, one popular method being a logarithmic scale.
Where AGC systems are used to obtain this type of characteristic, it must
be controlled by the averaged output which, as stated:before, must follow
the squaring device. This determines the sequence of devices in the block
diagram. The averaging circuit is conventionally a simple RC low-pass
filter and the averaging interval is usually considered to be RC seconds.
Thus, the interval during which variations in signal level will cause dif-
ferences between root average square recordings and averaged voltage record-
ing is RC seconds and variations of slower rate should cause no discrepancy.
The basis of the experiment was the measurement signals at 49.8 and
107.8 Ma/s on a time-sharing multiplexing system into a true power recorder
while simultaneously- making continuous voltage recordings of the signals
- App I page 7 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
0
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
in separate charnels. The recording intervals were taken as 20 minutes
on each channel. The higher frequency-transmitter was run with an off
period of several .minutes every half-hour Included. for noise level deter..
mination. Because of the consistent high level of the lower frequency
signal, noise level determination was unnecessary.
The experiment was conducted during the month of January, 1953. The
recording site was Sterling, Virginia, and the transmitter was located
at Cedar Rapids, Iowa. This experiment was part of the "Variations with
Frequency" portion of the program outlined in NMS Report 8A111, pp. 4-5,
and was mentioned in Appendix I of that report.
In addition to the basic purpose of the experiment, it was felt
that the tests would serve to check the 'validity of simple voltage record-
ing as a. basis for 'power measurement at either frequency, and comparisons
were made toward this end.
2. RESULTS
The recorded data was scaled for 20-minute interval median values in
conventional fashion. Data for periods of unscalable level or prevalent
interference or equipment trouble at either site were discarded. Insofar
as possible, meteor bursts were disregarded in the scaling. Due to the
time multiplexing, comparisons had to be made for periods where one or the
other frequency was recorded but not both. To improve the comparison, when
a value was needed for an interval when the signal was not recorded., the
averaged decibel level for periods immediately prior and following was used.
Intervals where either the prior or following periods were not recorded or
usable were 'discarded- For example, if the 49.81.1c/s signal was recorded
from 1100 to 1120 and 1140 to 1200 and the 107.8)4c/s signal from 1120 to
1140, the averaged values of the first and third intervals were used for
the 49.8 1.1c/s reading and. comparison was made to the second interval, 107.8
signal. If the first interval was subject to error, it was discarded
and a comparison centered on the third interval would be used if good data
were available for the fourth period, and so on. Several voltage record-
ings of both signals were scaled in the -same fashion. Calibration for 107.8
1.1qs was made with 20 db IF attenuation added to mask converter noise in the
signal range? Correction was made to this reading. Noise levels during the
transmitter off 'periods were scaled for 'the 107.8 :Mcis signal and the averag-
ing process applied to them as with the other signals. True signal level
for the 107.8 Ne/s signal was obtained by converting the decibel level to
equivalent microvolts squared and subtracting the noise voltage squared
from the recorded signal and noise. The net equivalent microvolts squared
were then reconverted to signal level in decibels. Correction for actual
radiated, power, transmission line impedances, line losses, etc., totaling
18 decibels was subtracted from the 49.8 Mc/s signal relative to the 107.8
signal, and the difference between the corrected 49.8 Mc/s signal and the
true 107.8 M[c/s level was divided by 10 log10 (107.8/49.8) to convert to
exponent of the frequency ratio, 2.16. The accuracy of the method is con-
tingent upon the signal and: noiseof the 107.8 ',leis signal adding in the
App I page 8 -
square as well as validity of the averages used for the signal missing
during the comparison periods. Further, it is assumed in this signal
extraction technique that the noise is adequately-masked in calibration
and that the converter is linear from noise level to the upper calibra-
tion level.
The effect of power recording is evident in the pen records by the
unusually large range of signal swing. Althouel this causes some extra
difficulty in arriving at a satisfactory median in scaling, the records
are still readily scaled. There were no other significant differences
noted between voltage and power recordings.
The statistics of the experiment are summarized as follows:
Number of exponent determinations:
Over-all average of exponent ?
Average deviation from average
Median exponent
Greatest recorded exponent
Least recorded exponent
115
7.51(3)
0.46(3) or 1.56 db
7.58
10.15 (date: January 20, 1953)
4.74 (date: January 20, 1953)
Figure I-10 presents a bar graph showing the distribution of record-
ed, exponent values for the entire test. Comparison should be made to
Figure 53 of NBS Report 8A111 (or Figure I-1 of this appendix) which was
based on conventional voltage recording. Figure I-11 presents the averaged
daily exponent and the range of values together with the number of obser-
vations for each day of the experiment. Figure 1-12 shows 20-minute median
values for a typical day.
Comparison between voltage and power recording for the strong 49.8
Mc/5 signal showed less average difference than the estimated accuracy of
the data, and it is assumed that the two systems are essentially in agree-
ment for this type of signal. However, comparisons of the 107.8 Mc/s data
almost always showed about 2 db difference in favor of the power recording
system. Since the compared data was scaled in the same fashion by same
personnel and calibrations were made on both systems by the same operators,
it is assumed that there may be an inherent difference in the response of
the average voltage vs power recording systems for signals of this type.
Since the 107.8 Mbis signal is well within the noise level, it is subject
to much greater "small detail" fluctuations and as seen in the introduction,
the root average square will tend to exceed the average as the spread and
fine structure of the test specimen is increased. The 107.8.1.1c/s true sig-
nal level is shown in Figure 1-13 for both voltage and power recording dur-
ing a typical day. Other usable days were substantially-the same.
3. CONCLUSIONS
A median value of 7i for the frequency dependency exponent was ob-
tained for the period of the observations, all taken near-midday. Further,
the validity of voltage recording as a basis for power calculations seems
App I page 9 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
. . =
-verified., especially where the signal recorded does not have high level
fine structure during the averaging interval of the recorder. In the case
of noisy or wide signal level excursions during the averaging interval,
there maybe first order errors introduced.
4. DESCRIPTION OF EQUIPMENT
The block diagram for the recording systems is shown in Figure 1-14
and is conventional except in the squaring amplifier in the second IF
chain of .the power recorder and in the time multiplexer. The squaring
amplifier circuit is based on a design of J. R. Johler of the Central Radio
Propagation Laboratory. It employs a push-push plate, push-pull grid ar-
rangement wherein the plate circuit is tuned to the second harmonic of the
input. Squaring is excellent over at least 40 db input range.
The power recorder bandwidth was 6 Kcis compared to about 2.5 Kc/s
for the voltage recording channels. This caused the apparent recorded
107.8 Mcis signal in noise and noise levels to differ considerably in
actual use; however, signal extraction eliminates this effect.
The cathode follower was included to prevent switching effects in mul-
tiplexing from disturbing the voltage recording channels. Sufficient pad-
ding was left in the 0-70 db attenuators to cause no disturbance in the
voltage recorder chains.
The GS-3 receiver and power supply were highly gain stabilized units
developed by the authors at the Central Radio Propagation Laboratory.
REFERENCES
1. D. K. Bailey, R. Bateman, R. C. Kirby, "Radio Transmission at VHF by
Scattering and Other Processes in the Lower Ionosphere," Proc. I.R.E.,
vol. 43, pp. 1181-12301 October, 1955. (Reprints are included as
Appendix XII in a limited number of copies of this report, furnished
to Contracting Officer, USASEL; copies were also furnished earlier to
the Contracting Officer as NBS Report 3563.)
2, A. D. Wheelon, *Radio Frequency and Scattering Angle Dependence of
Ionospheric Scatter Propagation at VHF," Jour. Geophys. Res., vol. 62,
pp. 93-112; March, 1957.
3. D. K. Bailey, R. Bateman, R. C. Kirby, "First Report on Regular VHF
Ionospheric Propagation Observable Over Long Distances," NBS Report
No. 8A111, June 30, 1952, (Interim Technical Report on MIPR No.
821-PHIBP-51-04).
4. D. K. Bailey, R. Bateman, R. C. Kirby, "Second Report on Regular VHF
Ionospheric Propagation Observable Over Long Distances," NBS Report
No. .8A1171 June 30, 1953, (Interim Technical Report on MaPR No.
821-PHIBP-51-04).
App I pagel0
?11,
?
?
5. R. C. Kirby, "VHF Propagation by Ionospheric Scattering - A Survey of
Experimental Results," Trans. I.R.E., CS-4, 1955.
6. R. M. Davis and E. K. Smith, "The Effect of Sporadic E on VHF Trans-
mission in the U. S.," NBS Report No. 5547, Jan. 28, 1958.
- App I page 11 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
?
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
a
?
?
I.
?
NUMBER OF HOURLY OBSERVATIONS
O 2 4 6 8 10 1?
1111111111111
EFFECTIVE EXPONENT n (per unit aperture of receiving antenna)
O 2 4 6 8 10 12 0 2 4 6 8 10 12
240-
220: ALL OBSERVATIONS 660
200 - 1 600
180- r MEDIAN n ? 49 540
160- 480
140 f 420
120 360-
100-300-
80-240-
60 180-
40: F 120 :
20 60
L 0-
o 10 2- 0 30 40 50 60 70 80 0 to
O 2 4 6 8 to 12
200
180
160
140 -
120 -
100 -
-
BO -
60
40 -
2 0 -
-
1111111111111
OBSERVATIONS DURING HOURS
UNAFFECTED BY Es PROPAGATION
OR ABNORMAL ABSORPTION
MEDIAN A.69
I
I I
to 20 30 40 50 60 70 80
ALL OBSERVATIONS
/-MEDIAN n.47
VALUES AFFECTED
BY Es
?
jihum. SEE FIG 1-9
0 30 40 50 60 70 80
O 2 4 6 8 10 2
1111111111111
720 -
III
llll 11111
ALL OBSERVATIONS
330 FSECIAN n.5.3
300
270
240
210
180
150
120
90
60
30
0 0
a
SHADED AREAS RF_FRESENT OB-
SERVATIONS DAM COPCITIONS OF
MOOERATE OR INTENSE ABSORPTION
AS INDCATED BY ATTENUATION OF
FACKGROLAID GALACTIC NOISE OR
BY1.31 FREQUENCY BLACKOUT
VALUES AFFECTED BY Es
PROPAGATION OR SPUTTER
SEE FIG 1-9
10 20 30 40 50 60 70 80
0 2 4 6 8 10 12
1111111111111
OBSERVATIONS DURING HOURS
UNAFFECTED BY Es PROPAGATION.
660 - AURORAL 'SPUTTER OR ABNORMAL
/MEDIAN n ? 4.7
600 ADSORPTION
:
540- OBSERVATIONS WRING HOURS
UNAFFECTED BY E. FROPAGA-
480- TION OR ABNORMAL ABSORP?
TION c?MEDIAN?5.3
? -
420- 140
360 - 120
300 - 100
240- 80-
180 - 60
120- 40
60 - 20
1.-? I LI I I I I I I
10 0 30 40 50 60 70 80 0 to ? 20 30 - 40 50 60 70 80
O 2 4 6 8 10 12 0 2 4 6 8 12
120
110
100
90
80
70
60
1111111111111
45 MEDIAN n.7.8
40 -
35
30
25 -
20 -
1 5 -
10 ^
5 -
MCDAY OBSERVATIONS DURING
FIVE-HCUR PERICO (10-4511 LT)
CENTERED APPROX. ON LOCAL
NOON AT PATH MIDPOINT
In I
to 20 30 40 50 60 70 80
so
40
30
20
10
MEDIAN n ? 5.1
MIDDAY OBSERVATIONS WRING
FIVE-HOUR PERIOD (I0-15 h LT)
CENTERED APPROXIMATELY ON
LOCAL NOON AT PATH MID-
POINT
?0 J ? 2 30 40 50 60 70 80
40
36 -
32 -
28 -
24 -
20 -
16 -
12 -
4
4 6 8 tz
IIIIIIIII
rMEDIAN n ? 5.3
MIDDAY OBSERVATIONS DURING
FIVE-HOUR PERIOD DO- Mb LT)
CENTERED APPROXIMATELY ON
LOCAL NOON AT PATH MID-
POINT
I I%I I I I
10 20 30 40 50 60 70 BO
RATIO OF RECEIVED SIGNAL INTENSITIES (SCALED ANTENNAS) DECIBELS
107.80 Mc/s vs 49.80 Mc/s
CEDAR RAPIDS-STERLING
OCT 52 THROUGH JAN.53
27 775 Mc/s vs 49.80 Mc/s
CEDAR RAPIDS-STERLING
JUNE 54 THROUGH WAN. 55
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
FIGURE I-I
24.325 Mc/s vs 48.87' Mc/s
ANCHORAGE-BARROW
MARCH 53 THROUGH JUNE 53
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
0
0
CD
25-
cri
--5
-10
--15
00 02 04 0'6 d8 10 12 114 18 18 20 22 00
LOCALTI ME AT PATH MIDPOINT
FREQUENCY EXPONENT ( Per unit aperture
4 5 6 7 8 9 10
(/)
zI00-
(LI 90-
80-
0 ? 70-
>- 60-
_1
cc 50-
m
=0 40-
?-3 30-
cc 20-
w
2 10-
D
SCALED RHOMBIC ANTENNAS
-1Th
10 115 20 25 30 35 40 45 50
RATIO OF RECEIVED INTENSITIES, 49.8 TO 107.8 Mc/s, DB
COMPARISON OF SIGNAL INTENSITIES, 107.8 vs. 49.8 Mc/s
Cedar Rapids to Sterling
FIGURE I ?2
4
40
? 35-
.4?3 25-
CD
0
27.775 Mc/s
(SCALE ON LEFT)
49.800 MCA
(SCALE ON RIGHT)
JUNE 1954
-20
15
-10
00 02 04 0'6 08 10 12 14 16 16 20 22 00
LOCAL TIME AT PATH MIDPOINT
(c)
z100-
o.
(2) 90-
x
80-
m
0 70-
c)
>- 60-
- 50-
0
40-
u_
o30-
wx 2 0-
FREQUENCY EXPONENT ( per unit aperture)
0123456789101112
I I I
SCALED RHOMBIC ANTENNAS
?Th
0 5 10 5 20 25 30 35
RATIO OF RECEIVED INTENSITIES, 27.775 TO 49.8 Mch, DB
COMPARISON OF SIGNAL INTENSITIES, 49.8 vs. 27.775 Mc/s
? Cedar Rapids to Sterling
FIGURE 1-3
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
I 45-
(c)
5 40-
co
24.325 Mc/s
( SCALE ON LEFT)
48.870 Mc/s
(SCALE ON RIGHT)
JUNE 1953
-25
-20
-15
00 02 014 06 018 101 12 14 1 16 18 20 22 00
LOCAL TIME AT PATH MIDPOINT
zI 00
90
g 80
2
o70
>.60
S 50
40
FREQUENCY EXPONENT ( per unit aperture)
I 2 3 4 5 6 7 8 9 10
SCALED YAGI ANTENNAS
0 5 10 15 20 25 _ 30 35 40
RATIO OF RECEIVED INTENSITIES, 24.325 TO 48.87 Mc/s, DB
COMPARISON OF SIGNAL INTENSITIES, 48.87 vs. 24.325 Mc/s
Anchorage to Barrow
FIGURE I ? 4
SCALED RHOMBIC ANTENNAS
34 ?
JUNE 1952
03
3 34
Ui ? 32
Ui
1? 30
-J
? 28
Ui
>-
- 34
cc
u_ 32
o?
4
? 30
w ? 28
X
34
32
30
28
8
7
JULY 1952
DEC. 1952
I I
00 02 04 06 08 10 12 14 16 18 20 22 00
LOCAL TIME AT PATH MIDPOINT
DIURNAL VARIATION OF FREQUENCY DEPENDENCE 107.8 vs 49.8 Mc,
CEDAR RAPIDS TO STERLING, JUNE 1952 ? JAN 1953, INCL.
FIGURE 1-5
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
6
8
7
6
8
7
6
8
MEDIAN EFFECTIVE FREQUENCY EXPONENT ( PER UNIT APERTURE OF RECEIVING ANTENNA)
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
SCALED RHOMBIC ANTENNAS
JUNE 1954
OCT. 1954
22
20
18
16
14
JULY 1954
NOV. 1954
AUGUST 1954
DEC. 1954
SEPT. 1954
JAN. 1955
00 02 04 06 08 10 12 14 16 18 20 22 00
LOCAL TIME AT PATH MIDPOINTS
DIURNAL VARIATION OF FREQUENCY DEPENDENCE 49.8 vs 27.775 Mc,
CEDAR RAPIDS TO STERLING, JUNE ?SEPT. 1954 INCL.
14 ^II I I 1 1 1 1
00 02 04 06 08 10 12 14 16 18 20 22 00
LOCAL TIME AT PATH MIDPOINTS
DIURNAL VARIATION OF FREQUENCY DEPENDENCE 49.8 vs 27.775 Mc,
CEDAR RAPIDS TO STERLING, OCT. 1954 ? JAN. 1955, INCL.
FIGURE 1-6 FIGURE 1-7
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
EFFECTIVE EXPONENT n (PER UNIT APERTURE OF RECEIVING ANTENNA)
2 4 6 a 10 12 14 16 18 20 22
I I I I 1 I I I 1 I I 1 I I I I I 1 I
MARCH?APRIL 1953
SCALED YAGI ANTENNAS
90
80
70
60
50
40
SPUTTER OBSERVED ON
ONE OR BOTH FREQUENCIES
MARCH 1953 THROUGH JUNE 1953
ilONE OR BOTH VALUES
ADJACENT TO AN HOUR
STRONGLY INFLUENCED BY Es
IONE OR BOTH SIGNALS
STRONGLY INFLUENCED
BY SPORADIC E
MAY?JUNE 1953
,-0 10 20 30 40 50 60
-I
m RATIO OF RECEIVED SIGNAL_ INTENSITIES (SCALED ANTENNAS) DECIBELS
g ANCHORAGE TO BARROW ?48.87 VS. 24.325 Mc/s
IL
0
cc EFFECTIVE EXPONENT 11 (PER UNIT APERTURE OF RECEIVING ANTENNA)
to 0 2 4 6 8 10 12 14 16 18 20 22 24
m
m I 1 I 1 1 I I I I I I I I I I I I I I I I I I I 1
m
z
18
00 02 04 06 08 10 12 14 16 18 20 22 00
_
LOCAL TIME AT PATH MIDPOINT
DIURNAL VARIATION OF FREQUENCY DEPENDENCE 48.87 vs 24.325 Mc, ANCHORAGE TO BARROW, MARCH?JUNE 1953, INCL.
FIGURE 1-8
70
60
50
40
30
20
10
JUNE 1954 THROUGH JANUARY 1955
10 20 30 40 50 60
RATIO OF RECEIVED SIGNAL INTENSITIES (SCALED ANTENNAS) DECIBELS
CEDAR RAPIDS TO STERLING ? 49.8 VS. 27775 Mc/s
FIGURE I - 9
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
-
NUMBER OF OBSERVED VALUES OF EXPONENT WITHIN GIVEN RANGE
24
22
20
18
16.
14
12
10
8
6
4
2
1
1 1 1
AVERAGE-4 1.-MEDIAN
2
5
17
8
19
13 13
3
2
4
5
6
7 8
EXPONENT VALUE
9
10
DISPERSION OF RECORDED MEDIAN VALUES OF EXPONENT OF FREQUENCY RATIO (1/2:16)
FIGURE 1-10
a
1 It-1
1 1
1 1 1
co
2
Lucr
cs
co
LENGTH OF BLOCK INDICATES RANGE OF EXPONENT.
NUMBER OF EXPONENT OBSERVATIONS INDICATED IN OR ABOVE BLOCK.
INNER LINE INDICATES AVERAGE EXPONENT FOR EACH DAY
' 0 0)
0
N- (I) to ro C?I 0
011V2:1 1VNOIS 03.1.03N803 3(11VA NVIO3V4 AO 1N3N0dX3
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R603000180001-4
LI ?1 3101913
AVG AO 31411.
ST'
8
8
8
0
17-1-00081-000?001?1701-0-1-8dCl-V10 60/90/171-0Z -1A-09 ? eSeeiei .104 panaiddv Ado paz!PeS u! PeWsseloeCI
RECORDED LEVEL IN DB CORRECTED FOR AMBIENT NOISE
I I i I i
-
-A N o 9) iii
CA 9/01 =
9NKD1033H 83MOd 301:11
ZI-1 381191A
-
VARIATION OF MEDIAN EXPONENT DURING A TYPICAL DAY, JANUARY 12, 1953
MEDIAN EXPONENT
AVG AO 3
0
1 I 1
1
17-1-00081-000?001?1701-0-1-8dCl-V10 60/90/171-0Z -1A-09 ? eSeeiei .104 panaiddv Ado paz!PeS u! PeWsseloeCI
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
0
AMPLIFIER'
SUPER-PRO
0
0
?
AMPLIFIER
SUPER- PRO
RECEIVER
to
cr.
?
100
tr)
U,
w 5
0
0
=0
1? ?I
- 0-1
Ij
0
It)
CONVERTER
0
CONVERTER
0
CONVERTER
BLOCK DIAGRAM OF RECORDING SYSTEM
?
FIGURE 1-14
APPENDIX II
EXPERIMENTAL INVESTIGATIONS OF IONOSPHERIC FORWARD
SCATTERING AT VHF USING PULSE TECHNIQUES
V. C. Pineo
This report is in four parts. Part I describes the results of
wit experiment performed during the winter of 1954. In this experiment
heights of ionospheric forward scattering were deduced from one-way
measurements of the relative transit times of the tropospheric compo-
nent and the ionospheric component of a pulsed signal received at 49.7
Mc/s at a distance of 793 Km from the transmitter. These deduced heights
were compared with virtual heights observed simultaneously at 46OCKcis
near the midpoint of the path. Part II describes the results of one-way
transit time measurements made during June 1954 at 49.7 Mc/s over a path
length of 810 Km. Part III gives results of oblique incidence pulse mea-
surements as a function of path length to determine dependence on scat-
tering angle. Part IV gives experimental results on heights of ionos-
pheric and tropospheric scattering determined from round-trip pulse delay
measurements.
PART I - CORRELATION WITH OBSERVED HEIGHTS AT LF
1. INTRODUCTION
The purpose of this experiment was to determine if a correlation
exists between the heights of ionospheric forward scattering observed
at VHF and the ionospheric heights observed, at vertical incidence at IF.
VHF ionospheric scatter heights ranging from about 70 Km to about
90 Km were obtained in earlier experiments by this laboratory)-2 Heights
within this ,same range have been observed consistently at LF.3/". Thus
the possibility is suggested that the propagation of LF radio waves by
ionospheric refraction and the propagation of VHF radio waves by ionos-
pheric forward scattering are both due to the same ionization mechanisms.
If so, both phenomena night be expected to have similar diurnal varia-
tions in height.
In order to test for such a correlation, heights of ionospheric for-
ward scattering observed at 49.7 Mcis over a 793 lon path length were com-
pared with virtual heights observed simultaneously at 46o Kc/s at vertical
incidence at Sterling, Virginia. The VHF path was cbpsen to place the NES
Sterling, Virginia, laboratory.near the midpoint.
The VHF scatter heights were deduced from measurements of the dif-
ferences in transit time between pulses propagated by tropospheric for-
ward scattering and. pulses propagated by ionospheric forward scattering.
For convenience, this difference in transit time is called the sky-wave
delay.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
????????
2. _EXPERIMENTAL DETAILS
The height observations at 460 Kc/s were made with the NBS vertical-
incidence LF ionosphere sounder at Sterling.
The .NBS mobile high-powered VHF pulse transmitter5 was set up at
Montgomery, New York, on an unused airfield belonging to the U.S. Air
Force. The receiving equipment for the sky-wave delay measurements was
located near Dobson, North Carolina.
The width of the transmitted pulse was 30 microseconds. The pulse
repetition frequency was 50-pulses per second. The peak power input to
the transmitting antenna was about 500 Kw. Rhombic antennas were used
for transmitting and receiving.
The VHF receiving equipment at Dobson Was housed in a box-body
truck. The receiving equipment consisted of a receiver, an oscilloscope-
camera recorder, a monitor oscilloscope and associated tilling and synchro-
nizing equipment.
The VHF sky-wave delay measurements and the LF virtual height mea-
surements were both obtained with the use of automatically recording
cameras. In each case an intensity modulated range-time display on the
screen of a cathode-ray oscilloscope was recorded on continuously moving
35 mm film. The rates of film travel in the recording cameras were 3
inches per hour and 8 inches per hour respectively in the VHF and the LF
recorders.
Data from which sky-wave delays could be obtained were recorded at
49.7 Mc/s at Dobson during the period January 11, 1954, to February 4,
1954. During this period the 49.7 Mo/s pulse transmitter at Montgomery
was operated daily, except for Saturdays and Sundays, during the daylight
hours. There were also several 24-hour runs duiing the period and opera-
tions were extended as often as possible to include the hours around sun-
rise and sunset.
o.
The IF 'equipment was operated continuously during most of the period
without regard. to the operating schedule of. the VHF equipment. However,
some time was lost from the IF height recording_program when it was learned
that harmonics of the 460 Kc/s operating frequency were interfering with
other radio services. The transmitter was shut down for one day while
filters were installed to reduce the radiation of harmonics.
The tropospheric component of the pulse signal at Dobson was usu.-
ally very weak and often could not be received at all. This weakness of
the tropospheric pulse limited the amount of sky-wave delay data avail-
able for comparison with the virtunl heights observed at., Sterling.
App II page 2 -
Because of these and other difficulties, simultaneous recordings
at 49.7 Mc/s and at 460 Kc/s were not obtained until January 20.
Typical examples of the photographic records obtained simultane-
ously at Dobson and at Sterling are shown in Figure II-1.
3. ANALYSIS OF 'alb DATA
For the purposes of this report the following data were reduced
from the records: (1) the median hourly values of the virtual heights
of the lowest ionospheric stratum observed at 460 Kc/s at Sterling and,
(2) the median hourly values of the sky-wave delays observed at 49.7
Mc/s at Dobson.
The values of the virtual heights obtained at 460 Kc/s are thought
to be close to those of the actual heights since there is very little
group retardation at frequencies well below the critical frequency of the
layer.
The sky-wave delays obtained at 49.7 M/s were converted into
equivalent midpoint heights using the following equation:
At
[2 CR sin-T-)2 + (H + R [1 - cos
1
where At = sky-wave delay in seconds
c = 3 x 105 km/sec.
O = D/R
D = surface arc distance in kilometers
R = mean radius of the earth = 6368 Km
H = midpoint height in kilometers
Q 2
2. 1) - 2R tan (1)
2R tan 0/2 = tropospheric ray-path length determined by the intersection
of the horizon planes of the transmitter and. the receiver on
the Great Circle path.
The arguments for the two tangents model of Equation 1 were discussed in
a previous report.2
In Figure 11-2 the diurnal variations of the hourly median values of
midpoint scatter heights obtained from the sky-wave delay observations at
49.7 Mc/s at Dobson are compared with the virtunl heights observed simul-
taneously at 460 Kc/s at Sterling during the period January 20, 1954,
through February 3, 1954.
- App II page 3 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Figure 11-3 shows the diurnal variations in hourly median values
of heights obtained from simultaneous observations at 460 Kc/s and at
49.7 Mc/s during two typical January days.
4. RESULTS
Statistically significant amounts of data were obtained for only
the hours 0600 to 1900 inclusive, 75 degrees West time. During these
hours there was a pronounced similarity between the diurnal variations
of the midpoint heights calculated from the sky-wave delays observed
at 49.7 Mc/s and the diurnal variations of the virtual heights observed
simultaneously at 460 Kc/s near the midpoint of the VHF path. The
virtual heights at 460 Kc/s varied from about 80 Km around midday to
about 100 Km around sunrise and sunset. The calculated scatter heights
at 49.7 Mc/s varied from about 75 Km during the midday hours to about
90 Km around sunrise and sunset. Thus the midpoint scatter heights
obtained from the sky-wave delays at 49.7 Mc/s were from 5 Km to 10 Km
lower than the virtual heights observed at 460 Ke/s. It should be noted,
however, that these calculated scatter heights will be lower than the
actual heights if the transit time of the tropospheric pulse is greater
than that of the two tangents model of Equation 1.
In a later experiment (described in Part IV), round-trip pulse
techniques were used to measure the transit times over a surface path
length of 624 Km of both tropospherically propagated pulses and ionos-
pherically propagated pulses. The equivalent tropospheric scatter heights
obtained from these round-trip measurements were between 20 Km and 30 Km
above the earth at the midpoint of the path instead of a height of 7.6 Km
predicted by the two tangents model. This result was attributed to the
radiation patterns of the antennas used in the experiment. These antennas
were designed to illuminate a region about 90 Km above the earth at the
midpoint of the path. Thus the intensity of illumination over the mid-
point of the path increased very rapidly with height, thereby partly off-
setting the loss in scattering efficiency with increasing height in the
troposphere.
The radiation patterns of the antennas used in the experiment de-
scribed in this report were similar to those used in the round-trip ex-
periment (Part IV) in that the main lobes were directed at a region in
the ionosphere over the midpoint of the 793 Km path. The height of in-
tersection of the tangent rays of the transmitter and receiver over the
.midpoint of the 793 Km surface path is about 12 Km. However, the actual
height of maximum tropospheric scattering was probably greater than 12 Km
and could have been.between,20 Km and 30 Km. A. tropospheric height of
25 Km used in calculating the reference transit time results in calculated
ionospheric heights about 5 Km greater than given by Equation 1. Thus,
in this experiment, the actual heights of ionospheric scattering at 49.7
Mo/s may have been in better agreement with the virtual heights observed
at 460 Kc/s than are indicated in Figures 11-2 and 11-3.
App II page 4 -
5. ACCURACY OF q21:11.. MEASUREMENTS
The virtual height records from the low frequency ionosphere sounder
had height markers at intervals of 20 KM. Virtual heights were read to
the nearest kilometer. The estimated reading error was + 3 Km.
The sky-wave delay records from the VHF equipment had time-delay
markers at intervals of 100 microseconds. Sky-wave delays were obtained
by measuring the distance on the record from the leading edge of the tro-
pospheric trace to the leading edge of the ionospheric trace. This dis-
tance was then converted into microseconds of delay. The estimated
reading error was ? 5 microseconds of delay or approximately+ 4 Km in
the calculated midpoint height.
There were no known instrumental errors in the low frequency mea-
surements of virtual heights.
The only significant instrumental error in the VHF sky-wave delay
measurements was due to the rise time of the transmitted pulse. This
rise-time effect caused the received pulse to broaden in the direction
of the leading edge with increasing signal intensity. The maximum error
from this effect was about 5 microseconds, and occurred at times when the
intensity of one component of the signal was very weak and that of the
other component was near the saturation level of the receiver. During
this experiment the intensity of the tropospheric component of the re-
ceived signal was always very weak and the intensity of the ionospheric
component was usually well below-the saturation level of the receiver.
Errors in the hourly median values of sky-wave delay due to rise-time
effects were estimated to be not more than minus 3 microseconds and the
resulting error in calculated height to be no more than minus 2 Km. The
maximum errors from this rise-time effect probably occurred most often
during the midday hours for this was the time during which the intensity
of the ionospheric pulse signal reached its diurnal maximum,
6. CONCLUSIONS
The differences between the values of the virtual heights obtained
at 460 Kc/s at Sterling and the calculated height obtained from the sky-
wave delays observed at 49.7 Mc/s at Dobson are probably spaller than
indicated in Figures 11-2 and 11-3 because the actual tropospheric delay
is almost certain to have been greater than that of the two tangents model
of Equation 1. Also, any correction of the measured sky-wave delays for
errors due to pulse rise-time effects would result in. slightly greater
values for the calculated heights. Furthermore, the values of the virtual
heights obtained at 460 Kc/s must be slightly greater than the real heights
despite the small effect of group retardation at this frequency.
Thus, during the daylight hours, the ionospheric forward scattering
of VHF radio waves and the refraction of vertically incident LF radio
waves seem to occur at about the same height.
- App II page 5 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
The most significant result of the experiment, however, is the simi-
larity in the diurnal variations of the heights obtained at the two exper-
imental frequencies. This result is tentatively interpreted as indicating
that the ionospheric refraction of LF radio waves and the ionospheric scat-
tering of IMF radio waves are both due to the same region of ionization.
PART II - SUMMER HEIGHiS OF VET SCATTERING
1. INTRODUCTION
The purpose of this experiment was to obtain, summertime values of sky-
wave delays for the 810 Km 'path from Sterling, Virginia, to Bluffton, South
Carolina, to supplement similar data obtained for the same path during the
fall of 1952 and during the winter of 1953.2
2. EXPERIMENTAL EgTAILS
The equipment and the techniques used in this experiment were the sane
as those that were used. for the VHF time-delaSr measurements described in
Part I of this Appendix. The pulse transmitter was located at Sterling,
Virginia, and operated- at a radio frequency of 49.7 Mc/s. The pulse width
was 40 microseconds and the pulse repetition frequency was 50 pulses per
second. The receiving and recording equipment was located near Bluffton,
South Carolina. Sky-wave de1ay-6 were obtained during the period from
June 18 throne)) June 27, 1954.
The 49.7 Mc/s transmitter at Sterling was normally operated from about
0500 hours to 1900 hours, 75 degrees West time and, on one occasion, it was
operated continuously for 48 hours. Unfortunately, intermittent failures
of the electric power generator and the recording camera at Bluffton during
the 48-hour run caused an aggregate loss of 932-- hours from the recording pro-
gram. The durations of these interruptions ranged from less than hour to
5-17bnurs. Equipment failures also occurred occasionally on other days during
the experiment. These failures, oddly, occurred most often during the hours
around sunset.
3. ANALYSIS AND RESULTS
For purposes of analysis the photographic records were divided into
half-hour periods beginning On the hour and the half hour. The median
value of 'the sky-wave delay in microseconds for each half-hour period was
read from the records to the nearest 5 microseconds with an estimated read-
ing error of + 5 microseconds.
These individual half-hour median values of sky-wave delays and of the
calculated equivalent midpoint heights are shown in the mass plot of Figure
II-4-.. The diurnal variation of the median value of these half-hourly medians
is also shown in Figure 4.
App II page 6 -
Equivalent midpoint heights calculated from the observed sky-wave
delays varied from 67 Km during the midday to 87 Kmat night. The half-
hourly median values of calculated equivalent midpoint heights for the
period of the experiment varied from 71 Km during midday to between 80
Km. and 85 Km during the night.
The values of equivalent midpoint height shown in Figure II-4 were
calculated using Equation 1 which is discussed in Part I of this Appen-
dix. There it is pointed out that use of Equation 1 leads to values of
midpoint heights which are likely to be lower than the actual heights.
Thus, for reasons given previously, the calculated heights shown in
Figure II-4 are probably about 5 Km lower than the actual hdights.
4. DISCUSSION OF THE RESULTS
The median values of the sky-wave delays obtained in this experiment
were approximately the same values as those that were obtained for the
same path in November, 1952, and in January, 1953.2 There is, however,
one notable difference between the records obtained in this experiment
and the records obtained in the earlier experiment. The records from the
earlier experiments frequently showed a split sky-wave trace during the
daylight hours which seemed to indicate the existence of two scattering
stratum, one at a height of about 70 Km and the other at a height of about
85 Km.
Tentative conclukons drawn from these earlier results were that the
higher stratum accounted for most of the nighttime scatter and tended to
endure weakly throughout the day, and that the lower stratum, which was
detected during the daylight hours, accounted for the comparatively high
daytime signal intensities.
The signal intensities of pulses received via the lower stratum. dur-
ing the midday hours were roughly 10 db to 25 db greater thsn either the
daytime or nighttime signal intensities of pulses received via the higher
stratum.
The separation between the leading edges of the dual sky-wave traces
was approximately the sane as the width of the transmitted pulse and could
only be resolved at times when the signal intensities of the pulses received
via the lower stratum were relatively weak. During periods of high signal
intensities the two traces overlapped to produce a single trace roughly
twice the width of the transmitted pulse.
This splitting of the sky-wave trace is not evident in June, 1954,
records, but it is suggested. by a pronounced widening of the sky-wave
trace during the midday hours.
The difference, with respect to split traces, between these records
and the records from the earlier experiments is attributed to one impor-
tant difference between the experiments in the manner in which the photo-
graphic records were made. This difference was in the rate of film travel
- App II page 7 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
in the recording camera. A film travel rate-of 3 inches per hour was
used during the June, 1954, experiment. But a rate of about 1 inch per
minute was used. in the earlier experiments.
The intensities of two groups of received pulses propagated. by dif-
ferent modes have independent fading rates. .Under certain conditions
this independent fading characteristic makes it -possible to distinguish
between separate pulse groups even if the difference in transit time is
less than the duration of the transmitted. pulse. The transit-time dif-
ference can be resolved if the intensities at the camera of pulses propa-
gated. by the earlier mode are well below the saturation level of the film
and if the film is moving at a fast enough rate. Resolution is impossible
if the intensities of the pulses due to the earlier mode saturate the film
or if the film travel rate is very slow relative-to the fading rate of the
pulses. In such cases the pulses for the two modes overlap and produce a
widened trace having a time-width w = 2d - s
where d = duration of transmitted pulse
s % difference in transit time between the trailing edge of
the pulse due to the earlier mode and the leading edge
of the pulse due to the later mode.
Widened sky-wave traces were obtained frequently during the June,
1954, observations. The effect of this widening was to decrease the sep-
aration between the leading edge of the reference tropospheric trace and
the leading edge of the sky-wave trace. Widening of the trace usuplly be-
came noticeable some time after sunrise and disappeared by sunset.
The width of the received tropospheric pulse was approximately the
same as that of the transmitted pulse.
A good example of the difference in width between the sky-wave traces
obtained in the morning and those obtained around. midday is shown in Figure
11-5. In the morning record the sky-wave trace has about the same width as
that of the tropospheric trace, and the relative delay between the leading
edges of the two traces is about 75 microseconds. In the midday record the
,sky-wave trace is about 25 microseconds wider.thah it was in the morning,
and. its delay relativ* to the tropospheric trace is about 55 microseconds.
But in both the morning and the evening records, the relative delays be-
tween the leading edges of the tropospheric traces and the trailing edges
of the sky-wave traces are between 130 microseconds and 135 microseconds.
5. CONCLUSIONS
The widening of the sky-wave trace during the daytime can be consid-
ered to be the result of simultaneous scattering from two strata at dif-
ferent heights above the earth. The upper stratum is probably an enduring
phenomenon. The lower stratum seems to be a phenomenon of the daylight
hours only and its scattering efficieacy Is usually much better than that
of the upper stratum.
App II page 8 -
With the foregoing considerations in mind, the results of the June,
1954, sky-wave delay observations discussed herein are interpreted as sup-
porting evidence obtained in other experiments that the propagation of VHF
radio waves by ionospheric scattering is the result of at least two prin-
cipal propagation modes. The mode associated with the lower scattering
stratum is probably the result of solar influence.
PART III 1 SCATTERING ANGLE DEPENDENCE TESTS
1. INTRODUCTION
This experiment was performed in July, 1953, at a radio frequency of
49.8 Mc/s for the purpose of determining the dependence of the received
signal intensity on the forward scatter angle of a VHF radio wave propaga-
ted by ionospheric scattering.
A preliminary transmission equation for propagation by ionospheric
scattering is given in a paper on long-distance VHF propagation by Bailey,
et al.8 This equation was based on a simplified model for which all the
transmitted power is passed through a scattering volume in which the scat-
tering parameters are assumed to be constant and equivalent to those at
the path midpoint. Since these conditions will not be realized in practice,
this formulation should be regarded only as providing a useful estimate of
received signal levels and as a guide for future theoretical and experimen-
tal work. A simplified form of this equation is:
Pr -
f? 2 Kg
(1)
where K includes all the terms relating to ionospheric structure, wave-
length, antenna gains, etc., q is a purely geometric factor and Pr/Pt is
the ratio of received power to transmitted power. If the scattering vol-
ume is located above the midpoint of the Great Circle path between the
transmitter and receiver, q can be expressed as follows:
1
q "
2 n7 (2)
/ sin -
where: / = length of the ray path
7 = the angle at which the transmitted ray passes through the
2 scattering layer.
n is the angle-dependence exponent.
App II page 9 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
In the paper by Bailey, et al,6 n is given a value of 5. Theoretical
work by and Salzberg7 yields a value of n = 7 and by
Eckersley8 a value of n = 9. The values of n attributed to Feinstein
and Eckersley were not developed in the papers by these authors but re-
sult from extending their analyses to the transmission equation.
Experimental values for the angle-dependence exponent, n; were
obtained by comparing signal intensities received nearly simultaneously
during the daytime at distances of 491 Kt, 592 Km and 811 Km from the
transmitter which was located at Sterling, Virginia.
2. DESIGN OF 'AM EXPERIMENTAL SYSTEM
An ideal system for making an experimental determination of the
angle-dependence factor would consist of at least three identical trans-
mitter receiver combinations having a common midpoint but different path
lengths along the same great-circle path. Signal intensity measurements
could then be made simultaneously at each of the receiving points. The
differences in received power could be attributed to differences in dis-
tances and scattering angles. Such an ideal system would be very expen.L
sive. For this reason it was decided to set up an experiment using a
single transmitter and three receiving sites along the sane great-circle
path appropriately spaced to provide the greatest sensitivity to differ-
ences in the scatter angle. The effect of the angle dependence in the
geometric factor is to cause the received signal power to increase with
distance. This effect is greatest for path lengths under 1000 Km.
There were two other requirements governing the choice of the re-
ceiving sites. These were (1) low noise levels and uniformity with
respect to flatness of the ground and freedom from objects capable of
affecting the radiation patterns of the antennas, and (2) location as
nearly as possible on the meridian of Sterling so that the differences
in local time at the midpoints of the paths would be small. Considera-
tion of these requirements led to the choice of a path running in the
approximate direction of Savannah, Georgia, from Sterling. The choice
of the shortest path length for the experiment was based on a compromise
between desired angle sensitivity and required transmitter power.
Pulse techniques were required to isolate the ionospheric component
of the signal from the tropospheric component. The minimum difference in
time of arrival between tropospheric and ionospherib pulses anticipated
at the most distant site established the maximum pulse width and conse-
quently the bandwidth of the system. A pulse transmitter with useful peak
power output in excess of 500 Kw was available at Sterling.
The existence of a large amount of signal-intensity data taken at
49.8 Me/s over the 1243 'KM Cedar Rapids - Sterling path provided a means
of estimating signal intensities expected for other paths. As a result
of these considerations, approximate path lengths of 500 Km, 600 Km and
800 Km were regarded as suitable. Three receiving sites were located
App II page 10 -
?
along this path at distances of 491 Kt, 592 KM and 811 Km from the
Sterling, Virginia, transmitter. The locations and geographic coordi-
nates are given in Table I.
Function
Transmitting
Receiving
Receiving
Receiving
TABLE I
GEOGRAPHIC DETAILS OF TILL SITES
Distance to
Location Coordinates Transmitter
Sterling, Va.
Hamlet, N. C.
Florence, S. C.
Bluffton, S. C.
38?59% 77?28'w
34?53N, 79?34'w
34?12'N, 79?54'W
32?17'N, 80?51'W
491 Km
592 Km
811 Km
RhoMbic antennas were used for transmitting and receiving. Each
antenna system was designed to align the main lobes to the vertical angle
of departure corresponding to a midpoint scatter height of 70 Kt. The
assumption of a scatter height of 70 KM was based on oblique-incidence
measurements2 made in November 1952 and January 1953. During these exper-
iments the daytime heights observed were usilally less than 80 Km and the
strongest signal strengths tended to occur when the height was around 70
Kt. In no case was the height of the scattering stratum observed to be
greater than 90 Kt.
3. THE tYrtCT OF ANTENNA GAINS ON THE DATA
In the analysis of the data obtained in an experiment such as this
one, it is necessary to take account of differences between the antenna
gains of the three systems.
It was assumed that the gain due to the antennas for each one of the
experimental systems was directly proportional to the product of the plane-
wave gains of its transmitting and receivingzantennas. This assumption was
originally justified on the grounds that the use of peak-reading signal
voltage recorders and the operation of the q factor (Equation 1) would
result in recorded values of received signal intensities closely related
to the field intensities prevailing at the exact midpoints of the paths.
Consequently the received signal intensity for any one of the paths would
change approximately in direct proportion to changes in illumination of
this point resulting from changes in directivity of the antennas. Further
support for this assumption is found in the discussion of realized antenna
gains in the paper by Bailey, Bateman and Kirby,9 which shows that the
gain realized from a practical antenna system used in propagation by ionos-
pheric forward scatter is directly proportional to the product of the plane-
wave gains of its transmitting and receiving antennas, when the beamwidths
of the antennas are larger than the beamwidth of the effective scattering
volume.
- App II page 11 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
It should be noted that the neglecting of differences in antenna
gains in the analysis of the signal intensity data will result in de-
duced values for the angle dependence exponent somewhat larger than
they would be if the data were normalized to equal antenna-systems gain.
Since it is necessary to take into account any differences between
the gains of the three antenna systems it is desirable that these dif-
ferences be kept to a minimum.
4. DESIGN OF la ANTENNA
The effect on the gain of a rhombic antenna resulting from adjust-
ing the tilt angle and leg length to obtain the desired lobe alignment
is such that the gain increases as the required vertical angle of depart-
ure for which the antenna is designed decreases. This effect on the
relative gains of several rhombic antennas with main lobes designed to
have different vertical angles of departure can be minimized by using the
same leg lengths for all of the antennas and achieving the required lobe
alignments by adjusting only the tilt angles and the heights above the 12
ground. This is called the reduced-leg-length compromise design method
and was the method used in the design of the rhombic antennas employed
in this experiment.
The antennas for the shortest path length (491 100 were designed
to have the maximum gain consistent with correct alignment of the main
lobe. The antennas for the longer paths were designed with the same leg
lengths as those for the 491 Kin path and the heights and tilt angles were
adjusted to obtain the desired alignments of the main lobes. Even with
this antenna design there are differences in gain which cannot be neg-
lected. Furthermore, t4e differences in gain between installed antenna
systems such as these do not remain constant with changing midpoint
scatter height. This sensitivity to variations of the midpoint height
affects the relationship between the signal intensities observed at the
different path lengths. For this reason provisions for making height mea-
surements were included in the plans for the experiment.
5. DETAILS OF Ilit EXPERIMENT
The pulse transmitter located at Sterling, Virginia, was operated
at a peak pulse power of approximately 500 kilowatts. The pulse dura-
tion was approximately 50 microseconds and the repetitionfrequency was
50 pulses per second. The transmitter was provided with three antennas,
each electrically identical to that at the receiving site toward which
it was directed.
The equipments at two of the receiving sites were housed in small
transportable buildings. A, box-body truck was used at the other site.
,Air conditioners with thermostatic controls were installed at each site.
Each receiving site was provided with a pulse-type signal intensity
recorder consisting of a gated receiver and a-peak-reading recording
App II page 12 -
voltmeter. The response, of the recorder was essentially independent of
pulse duration for pulse durations greater than 20 microseconds. A var-
iation in pulse width from 20 to 200 microseconds produced a change of
less than * db in the record. The recording time-constant was 5 seconds.
The gated feature in the receiving equipment was required to prevent con-
tamination of the signal intensity record by the tropospheric component
of the signal. The time duration of the receiver gate was approximately
400 microseconds. Synchronization of the receiver gating pulses with the
transmitter repetition frequency was accomplished through the use of iden-
tical 100 Kcis crystal-controlled oscillators at the transmitter and re-
ceivers. Each receiving site was also provided with equipment for making
automatic range-time recordings from which the difference in transit time
between the ionospheric and the tropospheric components of the signal
could be obtained. Approximate values of midpoint ionospheric heights
could. be deduced from these range-time measurements.
The range-time recording equipment consisted of a receiver, a cathode-
ray oscilloscope and a recording camera.
A block diagram of a complete receiver-recorder installation is shown
in Figure 11-6.
The output from the pulse transmitter was alternately switched be-
tween the three transmitting antennas in accordance with a prearranged
schedule. At the outset of the experiment the schedule called for 15
minutes operation on each antenna during the hour with a 5-minute silent
period at the end of each transmission. The first transmission of the day
started at 0855 hours EST and the last one at 1435 hours EST. This sched-
ule was maintained from July 2 through July 16, Saturdays and Sundays ex-
cluded. The level of cosmic and other noise was recorded at the receiving
sites during the silent periods. Interference from the Cedar Rapids trans-
mitter, which was operating on the same frequency, was encountered occa-
sionally. In order to avoid loss of records from this source, the Cedar
Rapids transmitter was put on a time-sharing schedule with the Sterling
pulse transmitter. Under this arrangement the Cedar Rapids transmitter
operated for 15 minutes during the hour and the Sterling transmitter was
operated during the remaining 45 minutes in three 12-minute intervals, each
of which was followed by a 3-minute silent period. This schedule was main-
tained from 0900 hours to 1357 hours each day from July 17 until the con-
clusion of the experiment on July 31 at 1112 hours. Each of the signal
intensity recorders was calibrated at least twice daily. Identical pulsed
signal generators were used at each site for calibration. The signal gen-
erators were cross-checked against each other at the beginning and end of
the experiment. The maximum difference in output between any two of the
signal generators at any attenuator setting was less than half a decibel.
The receivers were first calibrated in terns of open-circuit antenna volt-
age. Immediately following this, another calibratien was made with the
antenna connected to the receiver input and the signal generator loosely
coupled to the antenna. The signal generator output attenuator was then
adjusted to produce full-scale deflection of the recorder pen and the
App II page 13 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
attenuator reading noted on .the record. The signal generator output was
then reduced in convenient steps until the antenna noise level was reached.
The attenuator setting for each step was noted on the record. This second
calibration was called "signal plus noise calibration" since it provided a
means of obtaining a quantitative evaluation of the effect of noise on the
recorded value of weak signals. The output from the gated receiver was
displayed on a monitor oscilloscope. The operator could adjust the re-
ceiver gate delay at will with a continuous 360 phase shifter provided
for the purpose. At the start of a recording period, the operator could
quickly bring the pulse signal into position on the oscilloscope screen
by putting the receiver in the ungated condition and at the same time hold-
ing down a "drift" switch which allowed the timing oscillator to run free.
The pulse signal would drift onto the screen immediately and could be
locked in at once by releasing the "drift" switch. The receiver gate was
then turned on and the gate delay adjusted to reject the tropospheric com-
ponent of the received. pulse signal. The operator watched the monitor os-
cilloscope continuously and adjusted the gate delay whenever necessary to
compensate for any frequency drift in the timing oscillators.
The power input to the final amplifier during each transmission was
recorded in the transmitter log book. A further check on the stability of
the transmitter power was provided. by a small dipole using a crystal rec-
tifier and a dc milliammeter. The dipole was kept in a fixed position with
respect to the transmitting antennas. The dipole current was read and re-
corded in the log book Once each hour.
Log books were also kept at each of the receiving sites. In these
logs, the operators recorded information which would be of use in evaluat-
ing the signal strength records.
An example of the pulse photographs is shown in Figure II-7;
6. ANALYSIS OF TIO, DATA
The signal intensity data were corrected for the effects of noise and
variations in transmitter power. Differences in-antenna gains were taken
' into account approximately by normalizing the data to the gain of the anten-
na system for the 811 KM. path. The calculated gain in decibels of each
antenna system relative to that for the 811 Km path for ionospheric scatter
heights of 70 Km.l 80 Km and 90 Km are shown in Table II.
/ TABLE II ,
RELATIVE GKLaa OF THE ANTENNA SYSTEMS
? "
Path Length
In Kilometers
Relative Gain in Db
Height of Ionospheric Scattering
70 Km
80 Km
90 Km
811
592
491
0
2
5
-
0 -
2
6
0
3
3
10
- App II page 14 -
It is to be noted that the antenna system gain at 592 Km i.elative to that
at 811 Km is not appreciably affected by changes in scatter heights within
the range of 70 Km to 90 Km.
Two noise correction curves were prepared. One of these curves uti-
lized the noise calibrations made at the receiver sites. The other curve
was calculated using the equation:
Es2 = 2 Es+n- En
2
where Es = actual signal voltage
E lc combined signal and noise voltage, observed
s+n
En = random noise voltage.
(3)
In each of these curves the ratios in decibels of apparent (observed) sig-
nal intensities to noise intensities are plotted as abscissae against the
correction factors in decibels as ordinates. The apparent signal intensi-
ties are the unadjusted values scaled from the signal intensity records and
include the effects of noise. Noise intensities were recorded during the
scheduled interruptions in transmission. The correction factor is the amount
by which the apparent signal intensity exceeds the actual signal intensity.
Both correction curves are shown in Figure 11-8. The lack of agreement at
small differences between signal and noise levels is attributed in part to
the difference between the condition of a continuous wave signal mixed with
purely random noise assumed in Equation 3 and the. conditions of the experi-
mental curve, i.e., a 50-microsecond pulse signal in a 400-microsecond re-
ceiver gate with some contamination by nonrandom noise, and in part to var-
iation of the detector characteristic with signal strength. The signal-to-
noise ratios encountered during the times that signals were recorded at more
than one site were such that the differences in results between the two noise
correction curves were small. At 811 Km the difference ranged from 0 to 1 db
with a median difference of db. At 592 Km the range was 0 to 2 db with a
median of db. At 491 Km the range was 0 to 2 db with a median of 1 db.
The difference between the two noise correction curves is even smaller when
applied to the differences between the signal intensities observed at 592
Km and 491 Km and those observed at 811 Km. As a practical matter, in order
to present a clear picture of the results without undue elaboration, noise
corrections lying halfway between those on the observed and cal:colated curves
were chosen. Midday median scatter heights of about 80 Km were obtained in
a later experiment performed in June 1954 at the same frequency over the 811
Km path. For this reason the data were normalized to the antenna systems
gains for an 80 Km scatter height, except for the hours 09 to 11, July 15,
when simultaneous 'time-delay measurements indicated a height of 90 Km.
- App II page 15 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
The records from each site were scaled and the signal intensity and
noise level in db above 1 microvolt open-circuit antenna voltage, 600 ohms
reference impedance, tabulated for each period during which the pulse trans-
mission was directed towards the site. Graphs of signal intensity vs time,
using linear interpolation between consecutive coordinate points, were drawn
for each site. The median signal intensity for each hour which presumably
would have been recorded at the site if the transmission had been continuous
was then obtained from these graphs.
7. RESULTS
Nearly simultaneous records of sky-wave signal intensities were ob-
tained for 10 hours at all three receiving sites and for 25 hours at the
811 Km site and at the 592 Kin site. Signal intensity records were obtained
at both 592 Km and 491 Km on four occasions during which the records at 811
DI were obliterated. by interference.
The automatic time-delay recording part of the program could not be
carried out because of corona troubles which developed in the recording os-
cilloscopes after their installation at the receiving sites. However, some
measurements of sky-wave vs tropospheric-wave delay time were obtained man-
ually with the monitor oscilloscope at the 592 Km site between 10 hours and
11 hours, 75QW time on July 15. A midpoint ionospheric height of about 90
Km was calculated. from these measurements. This was the only occasion on
which height data were obtained while signal intensities were being recorded
at all three sites. In addition to these measurements, a series of single-
frame photographs of a cathode-ray oscilloscope "10-type presentation were
made on the same day at the 491 Km site between 0900 and 1000 Est. A dual-
beam cathode-ray oscilloscope was used. (A sample of the record obtained is
shown in Figure 11-7.) The output from the ungated IF amplifier is presented
on the lower trace. The output from the gated IF amplifier containing only
the sky-wave pulse is shown on the upper trace. Both tropospheric and sky..
wave pulses together with 100-microsecond markers are seen on the lower trace.
The tropospheric pulse is first from the left followed after a delay of 110
microseconds by the sky-wave pulse. The equivalent midpoint ionospheric
height is 88 Km. On the basis of these measurements it was assumed that a
scatter height of approximately 90'Km prevailed from 0900 hours to 1100 hours
on this particular day.
The signal intensities obtained at 811 Km were compared with those ob-
tained at 592 Km and at 491 Km. The results of these comparisons are shown
in Figure 11-9. Curve (a) in Figure 11-9 shows the comparisons for the hour
10 to 11 an July'15. Curve (b) shows the comparisons of the median values
of all of the other interpolated hourly signal intensities obtained in the
experiment. Two curves showing q as a function of distance calculated for
exponents of 5 and 10 for a midpoint height of 80 Km are shown for reference.
These comparisons indicate values for the angle-dependence exponent between
approximately 8 and 11.
App II page 16 -
?
Reference is again made to Table II which shows the effects of varia-
tions in midpoint height on the relative gains of the antenna systems. It
is seen that within the height range of 70 Km to 90 Km the change in rela-
tive signal intensity resulting from such height-related variations in an-
tenna gains is 1 decibel at 592 Km, and is 5 decibels at 491 Km, with 4
decibels of this change occurring in the height range of 80 Km to 90 Km.
Consequently the median relative signal intensity at 592 Km will be within
decibel of the value shown in Figure II-9(b) regardless of which height
within the range of 70 Km to 90 Km is assumed in normalizing the data to
equal antenna systems gain. The corresponding results for the 491 Km path
are considerably different. Here the use of antenna gains for a 90 Km mid-
point height in normalizing the signal-intensity data used in Figure II-9(b)
yields an apparent signal intensity at 491 Km slightly greater than that at
592 Km. This is inconsistent with the other results. The smooth curve of
Figure II-9(b) resulting from assuming a midpoint height of 80 Km in nor-
malizing the signal intensity data to equal antenna systems gain indicates
that the prevailing ionospheric scatter height was about 80 Km during the
experiment. This is in agreement with the median midday value of 80 Km
obtained a year later over the 811 Km path from measurements of the differ-
ences between the transit times of ionospherically-propagated pulses and
tropospherically-propagated pulses.
The possibility that the experimentally-determined values of n might
be a function of general propagation conditions was investigated by com-
paring the ratio of the signal intensities observed at 811 Km to those ob-
served at 592 Km with the signal intensities observed at 811 Ent; no correl-
ation was apparent.
8. SUMMARY AND DISCUSSION OF THE RESULTS
Received signal intensities increase rapidly as the forward scatter
angle at the ionosphere becomes smaller, or, stated differently, the sig-
nal intensity increases with increasing distance over the range of distances
used in this experiment. A daytime value between 8 and 11 for the angle..
dependence exponent, n, in the q factor was indicated.
The values obtained for n did not appear to be correlated with general
propagation conditions.
These values for n were deduced by comparing signal intensities which
had been normalized to equal antenna systems gains taken to be proportional
to the product of the computed plane-wave gains of the transmitting and re-
ceiving antenna for each of the test paths.
A qualitative consideration of the relationship of realized antenna
systems gains to the beamwidths of the antennas and the beamwidths of the
common volumes would indicate the use of normalizing factors slightly
smaller than the ones that were used but not as small as those that would
result from considering only the gains of the receiving antennas.
App II page 17 -
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Any errors due to the use of too large .normalizing factors for the
antenna gains are partially offset by the tendency of the effective com-
mon volume to decrease with distance.1?
It should be noted. that the use of smaller normalizing factors
results in larger deduced values for n. Consequently the values obtained
for n are considered to be about the minimum values actually existing for
the experimental paths during the observation periods.
PART IV - HEIGHTS OF IONOSPHERIC AND TROPOSPHERIC SCATTERING
DETERICENED.FROM ROUND-TRIP PULSE DELAY MEASUREMENTS
1. INTRODUCTION
? Round-trip pulse techniques were used to obtain the heights at
which ionospheric and tropospheric scattering of VHF radio waves occurs.
Round-trip delay observations were made during the periods December 1
through December 31, 1954, and September 28 through October 14, 1955, at
a radio frequency of 49.8 Mcis over a 624 Km path.
The results given by this part represent an extension of work des-
cribed in Parts I and II, and earlier?-L in which height of ionospheric
scattering was deduced from measurements of transit time of ionosphere
pulses relative to troposphere pulses. The experiment was conducted
jointly by the National Bureau of Standards and Lincoln Laboratories of
the Massachusetts InstiWy,of Technology. Results have been given in
separate presentations,"'" but are included here for the sake of com-
pleteness of the present appendix.
2. EXPERIMENTAL ARRANGEMENT
Two pulse transmitting and receiving systems were used. One was
located at Lincoln Laboratory's Round-Hill, Massachusetts, field station,
and the other was located at the NBS Sterling, Virginia, laboratory.
Synchronization of the two systems was achieved by using identical 100
Kcis crystal-controlled timing oscillators at each site. The pulse repe-
tition frequency was 50 pulses per second. Pulse widths of about 20 mic-
roseconds were used during the December, 19514., observations. During the
1955 fall observations, the pulse widths were 40 microseconds and 50 mic-
roseconds respectively for the Round Hill and the Sterling transmissions.
The peak power output of the Round Hill tranamitter was about 300 Kw. The
peak power output of the Sterling transmitter was 300 Kw during the Dec.
1954; observations and about 900 Kw during the 1955 fall observations.
The full rise time of the pulse transmitted from Round Hill was 4 micro-
seconds. The rise time of the pulse transmitted from Sterling was 8
microseconds during the December, 1954, experiment and 14 microseconds
during the 1955 fall experiment. Failure of the 'pulse modulator in the
Sterling transmitter at the start of the fall experiment forced a last
minute substitution of a pulsing unit having a slower rise time. Sep-
arate rhombic antennas were used. for transmitting and receiving at each
App II page 18 -
site. These antennas were electrically identical. The elevation
angle of the main lobes was 12 degrees. The bandwidth of the Sterling
receiver was 100 Kc/s; that of the Round Hill receiver was 300 Kc/s.
The noise figure of each receiver was about 3.
The output of the receiver at each site, containing the local
transmitter pulse and the pulses received from the other site, were
displayed on a monitor cathode-ray oscilloscope. A fixed delay of
about 8 milliseconds was inserted at Sterling between the local
Sterling pulse and the first pulse received from Round Hill. This
delay was inserted in order to separate the backscattered echoes of
each local transmitter from the forward scattered. pulses received
from the other site. An expanded portion of an intensity-modulated
sweep containing only the received forward scattered pulses was dis-
played on the screen of a recording cathode-ray oscilloscope. Reference
markers of known delay relative to the locally transmitted pulse were
included in the display. This display was photographed on 35 mm. film
moving continuously at about 6 inches per hour in the recording camera.
During the December, 1954, observations, the transmitters were
usually operated during alternate half-hour periods from 0915 hours
to 1645 hours. Whenever nighttime observations were made the trans-
mitter operated continuously. During the 1955 fall observations the
transmitters were operated continuously from 0900 through 1600 hours.
Photographic records were made simultaneously at each station.
These records indicate two principal modes of propagation, both of
which were continuous over long periods of time. The first mode in
time of arrival is attributed to tropospheric scattering and. the sec-
ond mode to ionospheric scattering. The pulses received. by these two
modes were very weak. In the "A" display of the monitor oscilloscope
their apparent amplitudes were seldom =eh above that of the cosmic
noise. There were also nunerous random bursts lasting up to a minute
or so having various time delays from zero to several hundred micro-
seconds relative to the second mode. These random bursts were probably
due to reflections from ionized meteor trails.
Each record was scaled to obtain the delays of the received. pulses
relative to the locally transmitted. pulse. The round-trip propagation
time is the sum of the delays measured simultaneously at each of the
two sites minus the inserted delay. The delays of the ionospheric com-
ponent of the signal relative to that of the tropospheric component were
also reduced from the records at each site.
Equivalent midpoint heights were calculated from the round-trip
data using the following equation:
H
.1(
2 2
HR sin Ek- - R 1 R cos 4)
App II page 19 -
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
(1)
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
where H = midpoint height in Km
t = round-trip propagation time in seconds
c = 3 x 105 Km/sec.
D = surface path length in Km
R = radius of curvature of the earth for the path
3. RESULTS
Mass plots of all the ionospheric and tropospheric round-trip delays
as read from the photographic records at approximately 15-minute intervals
and the ,equivalent heights are shown in Figures II-10 and II-11 for Dec.,
1954, and in Figures 11-12 and 11-13 for September - October 1955. Each
dot in these mass plots represents a single observation at about the indi-
cated time. A slight displacement timewise waS occasionally necessary to
avoid superposition of the dots.
The hourly median values of equivalent ionospheric heights calculated
from the December round-trip data varied between 931m at 1000 hours and
1700 hours and 89 Km from 1200 hours to 1400 hours. There were not enouel
data to calculate the medians for the other hours of the day. The median
value of all the measured round-trip ionospheric delays was 14.358 microseconds.
The equivalent midpoint height is 90 Km.
The equivalent hourly median values of midpoint ionospheric heights
calculated from the September - October, 1955, observations ranged from
87 Km at 1230 hours to 97 Km at 1430 hours. The median value of all the
heights was 921m.
In December the hourly median values of equivalent midpoint tropos-
pheric heights varied from about 301m to about 35 Km. The lower heights
were observed most frequently during the daylight hours. However, there
was no pronounced diurnal variation in the tropospheric height. The med-
ian value of all the round-trip tropospheric delays measured during Dec.
was 14.193 microseconds. The equivalent midpoint height is 32 Km.
? The median value of all the tropospheric round-trip delays measured
during September - October 1955 was 4206 microseconds. The equivalent
midpoint height was 39 Km.
Equivalent midpoint iono4heric scatter heights were also calcula-
ted from the relative delays between the ionospheric component and the
'tropospheric component of the signal measured at each of the two sites.
This relative delay is related to the midpoint ionospheric scatter height
by the following equation:
-,..m,legowtomeretnaw-,Ams-rrr*"...,",.
- App II page 20 -
???
1
(2)
c
where At = relative delay in seconds
La = tropospheric ray-path length in Km.
The smallest value for La consistent with the forward scatter mode of
propagation is 2R tan .4. This two-tangent model is also in accordance
with Gordon's analysis2 which shows that the optimum height for tropos-
pheric forward scattering is at the intersection of the horizon planes
of the transmitter and the receiver on the great-circle path.
The hourly median values of ionospheric heights derived from the
relative delays observed at Sterling during December, 1954, and the
heights derived from simultaneous round-trip observations are compared
in Figure 11-14. The heights obtained by the round-trip method averaged
about 8 Kin greater than thoe obtained by the relative-delay method. The
median value of all the heights obtained from the relative-delay measure-
ments at Sterling during December, 1954, was 85 Km. The relative delay
data obtained in September - October, 1955/ have not been examined.
4. DISCUSSION OF RESULTS AND SOURCES OF ERROR
There are three sources of error which affect the results obtained
with either one of these two methods of height determination. They are:
(1) errors in computing the surface distance between the two sites, (2)
reading errors by persons reducing the data from the recordings, and (3)
instrumental errors.
The surface distance for this experiment was computed by the Coast
and Geodetic Survey Office of the Department of Commerce.
The approximate magnitude of the reading errors was obtained by
comparing the individual numbers read from the Sterling record by NBS
personnel with the corresponding numbers read from the same record. by
Lincoln personnel. This comparison indicated reading errors of about
plus or minus 6 microseconds for the tropospheric delays and about plus
or minus 13 microseconds for ionospheric delays.
A check for possible instrumental errors was made by taking a low.
powered pulse transmitter and a receiver-recorder tO each of the two
sites in turn and measuring the round-trip delay time over a 2-mile path.
With full output from the low-powered transmitter the median value of the
measured delays was within 2 microseconds of the calculated delay. De-
partures from the median were less than A- 5 microseconds. However, when
the output of the low-powered transmitter was reduced to the point where
the amplitude of the signal received by the opposite receiver was just
above the threshold of recordability, the measured delay increased by
an amount almost exactly equal to the full rise time of the transmitter
App II page 21 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
pulse. This-weak-signal test simulated the actual condition which pre-
vailed at both sites throughout the round-trip-delay experiments.
It is evident from the above discussion that the greatest errors
in the measured round-trip delays are due to this pulse rise time effect.
The relative delay measurements were less sensitive to this effect be-
cause the tropospheric and ionospheric components of the received sig-
nal were about the same amplitude. The round-trip delays and correspond-
ing equivalent heights shown in the figures do not take account of this
effect and therefore are larger than the actual delays. The probable
lower limit for the actual round-trip delays can be obtained by subtract-
ing the rise times of both of the transmitted pulses from the measured
delays. The total time to,be subtracted from the round-trip delays mea-
sured in December is .12 microseconds and is 18 microseconds from those
measured in October. The hourly median values of equivalent ionospheric
heights corresponding to these lower-limit values of the round-trip delays
obtained from the December, 1954, observations were 86 KM. from 1200 hours
to 1400 hours and 90 Km at 1000 hours and 1700 hours. The lower-limit
median value of all the equivalent ionospheric heights for December was
87 KM. Similarly, the lower limit of the hourly median values of equiv-
alent tropospheric height for December varied between 20 Km and 28 Km.
The lower limit of the median value of all the equivalent tropospheric
heights for December was 24 EM. A similar adjustment of the September -
October, 1955, data gives lower-limit values of 87 I'm for the equivalent
ionospheric height and 29 'Km for the equivalent tropospheric height.
The tropospheric heights obtained in this experiment were several
times greater than the. optimum height predicted by Gordon's mode1.12
This result could be due to the radiation patterns of the antennas which
were designed to illuminate a region about 90 Km above the earth at the
midpoint of the path. Thus the intensity of illumination over the mid-
point of the path increased very rapidly with height. The loss in scat-
tering efficiency with height in the troposphere could have been partly
offset by antenna gains at the greater heights.
The diurnal variations in the observed raand-trip delays could have
been due to variations in pulse rise-tine effects with changes in signal
levels.
The ionospheric heights obtained in this experiment are valid only
for propagation paths similar to the experimental path in length and
geographic position. An average height of about, 90 Km is indicated.
Tropospheric heights between 20 Km and 30 Km are consistent with
tropospheric propagation over distances of 600 miles or so reported by
other observers .-"3
More round-trip pulse measurements should be made over longer paths.
Shorter pulses with better rise-time characteristics should be used. The
- App II page 22 -
Co
experiments should be designed to discover possible seasonal and geo-
graphic effects on ionospheric scatter heights.*
*Note: Since completion of this report and work under the Signal Corps
MIPR, an experimental set-up has been established for such measurements
between 4avaaa, Illinois, and Boulder, Colorado, and a preliminary report
issued.14
REFERENCES
1. V. C. Pineo, "Oblique-Incidence Pulse Experiments at 49.8 Mc," NBS
Report 4627, April 10, 1956.
2. V. C. Pineo, "Oblique-Incidence Measurements of the Heights at which
Ionospheric Scattering of VHF Radio Waves Occurs," Jour. of Geophys.
Res., 61, No. 2, June 1956.
3. J. N..Mrown.and J. M. Watts, "Ionospheric Observations at 50 KC,"
Jour. of Geophys. Res., 55, No. 2, June 1950.
4. J. M. Watts, "Oblique-Incidence Propagation at 300 Kc using the Pulse
Technique," NTS Report 1674, May 19, 1952.
5. W. H. Martin and P. G. Sulzer, "50 Megacycle Pulse Transmitter," NES
Report 8A1171 Appendix XVI.
6. D. K. Bailey, R. Bateman, L. V. Berkner, R. G. Booker, G. F. Montgomery,
E. M. Purcell, W. W. Salisbury, J. B. Wiesner, "A New Kind of Radio
Propagation at Very High Frequencies Observable over Long Distances,"
Phys. Rev., 86, pp. 141-145, 1952.
7. J. Feinstein, C. Salzberg, "Ionospheric Reflection of VHF Radio Waves,"
NBS Report 1812, July 22, 1952.
8. T. L. Eckersley, "Studies in Radio Transmission," J.I.E.E., 71,
405-459, 1932 (see pp. 439-443).
9. D. K. Bailey, R. Bateman, R. C. Kirby, 'Radio Transmission at VHF by
Scattering in the Lower Ionosphere," Proc. I.R.E., 43, pp. 1181-1230,
October 1955. (See pp. 1217-1218.)
10. W. G. Abel and V. C. Pineo, "Scatter Heights Determined from. VHF
Pulse &periments," given orally December 15, 1955, at URSI Fall
Meeting, Gainesville, Florida.
PP.
11. V. C. Pineo, "Heights of Ionospheric and Tropospheric Scattering
Determined from VHF Pulse Measurements," NBS Report 3570, February
10, 1956.
App II page 23 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
12. W. E. Gordon, 'Radio Scattering in the Troposphere," Proc. I.R.E.,
43, pp. 23-28, January 1955.
13. K. A. Norton, F. L. Rice and L. E. Vogler, "Use of Angular Distance
in Estimating Transmission Loss," Subtitle No. IV, Proc.
p. 1494, October 1955.
14. V. C. Pineal "An Exploratory Investigation, using Pulse Techniques,
of Ionospheric Forward Scatter Propagation at VHF over the 1295 Km
Path between Havana, Illinois, and Boulder, Colorado," NTS Report
5532, October 29, 1957.
- App II page 24 -
497Mc/s
460 Kc/s
AMP!
vAtritlw,P.1; sky wove
i....A144,1-440416 trop wove
0615 to 0715
?o
49 7 Mc Is Of Orlin
Eft 11
4 111... NI Melia
460 Kc/s
49 7 Mc/s
460 Kc/s
sky wove
trop. wove
1130 to 1230
1730 to 1830
EXAMPLES OF OSCILLOGRAMS OBTAINED SIMULTANEOUSLY AT
STERLING, VA. AND DOBSON, N.C. ON FEBRUARY 2, 1954.
FIGURE I I -I
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part- Sanitized Cop Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
100
90
80
70
DIURNAL VARIATION OF HOURLY MEDIAN VALUES OF (a) VIRTUAL
HEIGHTS OBSERVED AT 460 Kc/s AT STERLING, VA. AND ( b )
EQUIVALENT MIDPOINT HEIGHTS CALCULATED FROM SKY-WAVE
DELAYS OBSERVED SIMULTANEOUSLY AT 49.7 Mc/s AT DOBSON
N. C., 793 KM FROM THE MONTGOMERY, N.Y. TRANSMITTER.
FIGURE 11-2
100'?
(a)
90 ?
80 ? (b)
JAN. 25, 1954
70 ?
60 1 1 1 I
00 04 08 12 16 20 00
100
(a)
90
(b)
80
70 JAN.30, 1954
(a) VIRTUAL HEIGHTS OBSERVED DURING TWO TYPICAL DAYS IN
JAN., 1954 AT 460 Kc/s AT STERLING, VA. AND (b) EQUIVALENT,
MIDPOINT HEIGHTS CALCULATED FROM SKY-WAVE DELAYS
OBSERVED SIMULTANEOUSLY AT 49.7 Mc/s AT DOBSON, N.C.
FIGURE 11-3 ?
Declassified in Part - Sanitized Cop Approved for Release ? 50-Yr 2014/06/09 ? CIA-RDP81 01043R003000180001-4
SKY-WAVE DELAY IN MICROSECONDS
85
80
75
70
65
60
55
50
45
00 02 04 06 08 10 12 14
750 WEST TIME
?
? ? ? ? 00 00 00 ? 00 00 ? ?
?
? ?
? ? 00 ? ? ?
?
? ?
? ?
? ?
? ? ?
?? a ?
00 ? ?? ? ? ?? ?? ?? ?? ? ?? ?
? ? ? ? ID OM ? ?? ?? ????? ?? ?
EACH DOT REPRESENTS THE MEDIAN VALUE
FOR ONE CONTINUOUS OBSERVATION OF ONE
HALF HOURS DURATION. MEDIAN VALUES OF
THE HALF-HOUR MEDIANS ARE OUTLINED.
?
16
18
20
22
SKY-WAVE DELAYS AND EQUIVALENT MIDPOINT HEIGHTS OBSERVED DURING THE PERIOD JUNE 18
TO JUNE 27, 1954 AT 497 Mc/s FOR THE STERLING, VA. TO BLUFFTON, S C PATH, 810 KM.
FIGURE 11-4
?
?
?
?
85
(f)
cc
2
80 .0..1
c'e
cp
17)
75
I-
a.
?
TO
65
00
If
if
?
?
T IS TROPOSPHERIC SIGNAL AND I IS IONOSPHERIC SIGNAL WITH DELAYS RELATIVE TO THE TROPO-
SPHERIC PULSES CORRESPONDING TO HEIGHTS OF 85 KM and 75 KM RESPECTIVELY AT 0630 HRS
AND 1200 HRS. TIME-DELAY MARKERS ARE AT 100 MICROSECOND INTERVALS. DURATION OF TRANS-
MITTED PULSE IS ABOUT 40 MICROSECONDS. THE PRF IS 50 PULSES PER SECOND. DRIFTING
OF TRACES RELATIVE TO TIME-DELAY MARKERS CAUSED BY DIFFERENCE IN FREQUENCY BETWEEN
THE TIMING OSCILLATORS FOR TBE TRANSMITTER AND RECEIVER. UPPER OSCILLOGRAM SHOWS
COMPENSATING ADJUSTMENT AT 0622 HRS.
BLUFFTON, S. C. 0600-0700 20 JUNE 1954 75? WEST TIME
; NIIRPO'"
...4.*.wwere41)110/ MIA?
rtg grUllZOCKOTA:71?"'Fr"Arciaittriiiir-lettr
$ ...A.s.etworrXr I .
: a!' egfrilggi 14 ?
1111:f.tliail.".'1:14170Pratailrtni 1" ""in
atAti ,1 X u
i-:4-fragEgr.41,11.748141RIMIT/KM001Nic4
BLUFFTON, S. C. 1130-1230 20 JUNE 1954
grts-inemommairmummir
MIRIA111101.111A
027.,...11121VP.111111ii., 111111Weifferitlir
wriprwItrprryir
r i "r M'
1.111/11191MENTRIMetilittiNtRillicl
pT' riIrI(rg1; 1r1frIYArif Tir tirmx, ttp,,r.811. r 11 r. ? '"
o'
40.4,7 ...I 4,- .4 q,..1 +4 ,n# I inIKAIX?rtinirrftlAterdrr?str;?-? T
gym/nu iti.14510 rano
reCilinilfi OM If 111111NOMMIMIENNIIIII
It11(1117111INIMOP7.11:0PEEMPREEVErfalif,rgiffai
75? WEST TIME
I...1.-.)., eil 4 r 41.44Ibi tr'r" "" ' " - ' 4 4* 4 !"101,1ft"t1"Ittrt *reltwi*Iri-
. - ---- ?'"...... IWO VLCOLI .
4ITIOVIONCRIMER INEXIMECOr rA
lanlif S MIIMENIMANTISMIZEOF
Arjaanufafflizoilym ex unterivfiCergrIENDI3 f VIZIR
ammouptowr
113114,1171=1,11111fr
PEW;
IP Pr' ilfri.1.01111or1 41;0,4 "ril flt,. I Wt." "
4?
..L..H."11,1E21:tarimuir.trufrf Frivrrrry fravilluvfmr^of
L
two ,,C0111121fintq I 14- T
lagratigHPIPINMANIRIEWITINPre
. 1 . ill 14 Cil ? I a ,
VAINITATITNEEMERIIIIIMINNITEM ;
hErvanymazoarranoramacmowthrmenlyirittv;iira ,
EXAMPLES OF OSCILLOGRAMS OF RECEIVED PULSE SIGNALS AT 49.7 Mc/s ON THE STERLING
TO BLUFFTON PATH, 810 KM. SHOWING CHARACTERISTIC DIFFERENCES BETWEEN MORNING AND
MIDDAY IN WIDTH OF IONOSPHERIC TRACES.
FIGURE 11-5
50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
;4,
0
In
0 ?
Q.
0
4-
0
0
C
?
t
E ?
a
0
?
&
Imo. recording
05
0
pulse gen.
0
frog. divider
0
C.)
?4-
0
V
V
a.
100kc to 1 kc

8
Jr
BLOCK DIAGRAM OF RECORDING SYSTEM
FIGURE 11-6
JUL. 15 1953 BETWEEN 09 HRS AND 10 MIS, 75? W
TIME. TOP TRACES SHOW THE OUTPUT FROM THE GATED
RECEIVER CONTAINING ONLY THE IONOSPHERIC PULSE.
THE LOWER TRACES SHOW THE OUTPUT FROM UNGATED
RECEIVER CONTAINING BOTH THE TROPOSPHERIC PULSE,
FERST FROM LFPT, AND THE IONOSPHERIC PULSE.
TIME MARKERS IN LOWER TRACES ARE AT INTERVALS OF
100 MICROSECONDS. EXPOSURE APPROXIMATELY ONE
SECOND PER FRAME AT INTERVALS OF ABOUT 5 SECONDS.
SAMPLES FROM A PHOTOGRAPHIC RECORD OF PULSES
RECEIVED AT THE 491 KM SITE
FIGURE 11-7
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09 CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
777 .77"`--,f2tIn
/ (Sti,o)
A = Calculated
= Observed
COMPUTED FOR A HEIGHT OF 80 Km
DASHED LINES SHOW EXPERIMENTAL RESULTS,
(a ) FOR THE HOUR 10-11 75? WEST TIME, 15
JULY 1953 (b) MEDIAN OF ALL VALUES OB?
TAINED FROM 9 JULY TO 31 JULY, 1953.
- bo I 2 3 4 5 6 7 1 9 10
RATIO IN db OF APPARENT SIGNAL INTENSITY TO NOISE INTENSITY
-24
400 500 600 700 800 900
PATH LENGTH IN KILOMETERS
CURVES SHOWING THE CORRECTION IN DECIBELS TO BE APPLIED TO THE
RETURNED SIGNAL INTENSITIES TO COMPENSATE FOR NOISE EFFECTS.
FIGURE 11-8
RESULTS OF ANGLE-DEPENDENCE OBSERVATIONS
FIGURE 11-9
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09 ? CIA-RDP81-01043R003000180001-4
1316...M.????????
4440
4420
4 400
0
z 4 380
0
3 4360
4340
4 320
4300
00 02 04 06 08 10 12 14
75? WEST TIME
?
?
?
?
MEDIAN
LESS
?
? ?
?
1
= 4358
lOps
=
1 1 1
OF ALL OBSERVATIONS
ps ( 90 km)
FOR RISE TIME
4348 ps (87 km)
?
. ?
? ?
?
?
?
?
?
? ? ?
?
?
?
?
?
? ?
??
?
?
?
?
?
?
?
? ?
?
1
?
?
? ?
?
?
??
?
?
? ?
?
? ?
?
?
? ?
1
?
?
?
?? ?
?
? ?
??
?
: i
?
?
?
?
? li
llo ?
? ?
?
? ?
? ?
:
? ?
?
? ?
?
?
?? ?

: :
?
?
?
?
?
?
?
?
?
?
??
?
? ?
?
?
1
?
?
?
?
?
?
?
?
????
?-.
1
....
?
?
?
?
?
?
?
?
?
?
1 I 1 I 1 1 1 1 1 1 1
???1
?
16
18
20
ROUND?TRIP IONOSPHERIC DELAYS OBSERVED DURING DECEMBER, 1954
FIGURE 11-10
?
22
105
95
LT!
90
a.
85
00
80
75
4260
4240
z 4220
8
t)Di
3 4200
4180
4160
00 02 04 06 08 10 12 14
75? WEST TIME
?
1 1 1 I 1 1 1 11
MEDIAN OR ALL OBSERVATIONS
= 4I93ps (32 km)
LESS 10ps FOR RISE TIME
= 4183 ps (25 km)
?
?
?
? ?
?
?
?
?
?
?
?
small
???
?
?
?
?
1 1 1 1 I 1 I 1
?
16
18
20
ROUND?TRIP TROPOSPHERIC DELAYS OBSERVED DURING DECEMBER, 1954
FIGURE Il?l1
22
60
50
40 =
00
0
(D
0
CD
='!
(D
0_
-0
CD
CD
(D
0_
0
0
(7)
(D
0_
(D
(T)
CD
(D
50-Yr 2014/06/09 : CIA-RDP81-01043R003000180001-4
4440
4420
4400
z 4380
0
0
4360
4340
4320
4300
MEDIAN OF ALL OBSERVATIONS
= 4365 ps (92 km)
LESS 18 ps FOR RISE TIME
= 4347ps (87 km)
4-PERIOD OF OBSERVATIONS-0
1
1
00 02 04 06 08 10 12 14
75? WEST TIME
16
18
20
22
105
80
75
00
ROUND?TRIP IONOSPHERIC DELAYS OBSERVED DURING THE PERIOD SEPT. 28 TO OCT. 14, 1955
FIGURE 11-12
?
4260
4240
z 4220
Iii0
c.) 4200
4180
4160
00 02 04 06 08 10 12 14
750 WEST TIME
(D
-h
(D
(D
CD
()
(D
MEDIAN OF ALL OBSERVATIONS
= 4206 ps (39 km)
LESS 18ps FOR RISE TIME
= 4188 is (29 km)
*-PERIOD OF OBSERVATIONS-4
.? ?
? ? ?
.. ? . ?
? ? t?-? ix? ? .? ?
?
. ? ?
? ? ? ?
? ? ?
1
1
16
18
20
22
60
50
30 0
00
ROUND?TRIP TROPOSPHERIC DELAYS OBSERVED DURING THE PERIOD SEPT. 28 TO OCT.14,1955
FIGURE 11-13
20
10
2
ir 1-0008 1-000?0a1?170 1-8dC1I-V10 60/90/17 L O A-09
Declassified in Part - Sanitized Cop Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
0
N ui H. I. Leighton
>--
--- .ri
UJ
0 In Appendix II of NBS Report 8A117, it was shown that a correlation
u.
0 o exists between VHF signal intensities on the Anchorage-Barrow path at
u) LE 148.87 Mcis and. Magnet ic K-indices at Fairbanks, Alaska. The possibility
I- = of correlation between VHF signal intensities on the Cedar Rapids-Ster-
z a.
w CI) ling path at 49.8 Mc/s and. planetary magnetic K-indices is here con-
2 0
UJ Cl- sidered.
ce 0
n cc As before, the scale of K runs from 0-9 'according to intensity of
0 disturbance. Those K.-figures of seven, eight and nine are grouped in
<
w >
I- one category in ma3sing the analysis. The individual VHF signal inten-
- m o o sity departures are determined by taking the decibel difference between
z.,
0 Lt- the numerically central hourly median, for the particular three-hour
0 w
w' period, and. the monthly median of all the hourly values observed during
Z Cl- the month in the three-hour period.
N a co
I-- 0
-J Z
0 0 It will be observed. from the eight histograms of Figure .W.-1 and.
2 ? the graph of Figure III-2 that there is no very evident over-all corre-
O Zii 23 lation between 'VHF signal strength on the path observed in the temperate
0 1?
N-zone and. either planetary or Cheltenham -magnetic K-indices for the sane
02 c:
cc z three-hour periods. There is, however, a tendency toward greater signal
LL. c:t strength with the decidedly higher K-figures.
u) 0)
i-- >-
c0 Appendix X and. the graphs of Figure 140 in the main text of this re-
0 o _i
a5 w port show long-term relationship between signal strengths and. planetary
z 0 K-indices, Kp. The month-to-month correspondence shows a /general de-
o crease in both signal strength and. magnetic index during the years of
La- decreasing sunspot activity and. an increase during the rise in sunspot
= 1 activity. The correspondence is especially marked for the noontime
0 wt.., .
0 z period..
a.
cn ?
z
00
o _
o
0 - 0 o o o 0 o?
0 0) to t- to, to
Lull L_LH913H INIOdCIIIN
Appendix iij - page 1
Declassified in Part - Sanitized Cop Approved for Release ? 50-Yr 2014/06/09 ? CIA-RDP81 01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
APPENDIX III
PART II
CORRELATION OF VHF SCATTER INTENSITIES WITH H'F RECORDS
R. M. Davis, Jr.
?
1. INTRODUCTION
This appendix continues and concludes for the present the studies
described in previous reports,1,2 in which a connection was demonstrated
between intensities of VHF signals received over the Anchorage to Barrow
path and concurrent conditions of the ionosphere as deduced from h'f
ionosonde records. VHF signal intensities were shown to exhibit a sta-
tistical correspondence with the degree of absorption indicated by the
h'f records, both those made at the end points of the transmission path
and those made at Fairbanks, Alaska, in the general area of the path mid-
point. The period covered by the earlier studies was January through
August 1952. The data previously published for the period March through
August 1952 are reproduced for the sake of continuity in Figures 111-3
through III-10. This appendix presents subsequent data for the autumnal
equinox period of 1952, the winter season, 1952-53, and the spring equi-
nox months of 1953.
2. METHOD OF ANALYSIS
The procedure used in analyzing the observations reported in this
appendix is the same as that followed in the-case of the two earlier
reports. It is briefly reviewed. below.
Each hour for which VHF signal intensity recordings were available
-was classified under two headings:
a. The amount of departure of the median signal intensity for the
given hour from the.monthly median of such intensities.
b. The degree of simultaneous disturbance in the ionosphere indi-
cated on h'f records. The disturbance was evalpated in two ways; first,
from the h'f records made at the end points of the transmission path con-
sidered together; second, from a single record made at Fairbanks, Alaska.
To facilitate the study the signal intensity departures were grouped
in intervals 5 db wide, above and below the median. A, positive departure
denoted a value of signal intensity above the median, a negative departure
a value below the median.
Ionospheric conditions were evaluated by use of a special classifi-
cation system presented earlier by D. K. Bailey. The system was designed
to group the h'f records according to degree of disttrbance. Each h'f
Appendix III - page 2
record was placed in one of the classes a, p, s, or el in order of
decreasing absorption and increasing regularity and stability of the
ionospheric layers. In this system a represented complete absorption
of the h'f trace, corresponding to blackout of BF sky-wave communica-
tion. The symbol E stood for stable ionospheric layers and very low
D-region absorption. Besides these classifications, there was another,
7, denoting the extremely scattered and fragmented records, with mul-
tiple oblique reflections and unusual stratifications, considered to be
characteristic of auroral activity. The 7 classification was taken to
indicate an ionosphere more favorable to BF communication than p, but
less favorable than ep. The addition of a star (*) indicated that the
minimum frequency returned by the ionosphere was below 1.0 Mc/s. Such
records were considered to denote fairly quiet conditions.
In assessing ionospheric conditions for the pair of end-points,
the two Greek symbols assigned were replaced by a single figure, 1, 2,
3 or 4. The figure "1" represented the most disturbed average h'f con-
ditions; "4" represented the Quietest. Each possible combination of
the five Greek symbols was placed in one of the four numbered groups.
For Fairbanks, of course, one of the five Greek symbols alone sufficed
to describe the ionosphere.
The practice of grouping the observations together for presentation
by season is continued in this appendix. Autumnal equinox data are com-
prised of September and October, winter data of November, December,
January and. February, and spring equinox of March and April.
3. VARIATION OF VHF SIGNAL INTENSITY WITH HT OBSERVATIONS
Figure III-11 consists of four normalized histograms depicting for
September-October, 1952, distributions of departures of signal intensi-
ties from their monthly. medians. The histogram at the top refers to
category 1, representing extremely disturbed ionospheric conditions at
the path end-points. The other histograms present data for increasingly
quiet conditions, until the bottom figure illustrates the departures
during stable ionospheric conditions when no more than weak absorption
was ekperienced. Figure 111-12 presents comparable data for November
through February, 1952-53, and Figure 111-13 the data for March-April,
1953. Corresponding distributions of departure, referred to the ionos-
pheric conditions recorded at Fairbanks, appear in Figures 111-17, 111/18
and 111-19.
The interesting feature of these figures is the progressive shift
in the peaks of the histograms from positive values of departure to
negative values as the associated h'f records change from disturbed to
quiet. All of the figures, except 111-19, exhibit the shift in the peaks
of the histograms. It is most clearly shown in the winter figures, 111-12
and 111-18, in which twice as many observations are incorporated as in
other figures. The poor correspondence of signal intensity with h'f con-
dition in Figure III-19 may possibly be explained by the difficulty that
Appendii. III - Page 3
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09 ? CIA-RDP81-01043R003000180001-4
.4.0544,41a.
Declassified in Part - Sanitized Co .y Ap roved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
was experienced in classifying the April records at Fairbanks. Of the
many h'f records classified as Q, almoet all were doubtful.
4. DIURNAL AND SEASONAL VARIATION OF IONOSPHERIC CHARACTERISTICS
In connection with the above analysis, it is of interest to con-
sider the diurnal and seasonal variations of the occurrence of the four
types of ionospheric conditions in the Arctic.
Figures III-14, 111-15 and. 111-16 illustrate the diurnal distribu-
tion of the corabined Anchorage-Barrow conditions in the fall, winter
and.. spring seasons, respectively. Figures 111-20, 111-21 and 111-22
show how the four ionospheric conditions were distributed at Fairbanks
in the same three seasons. Finally there follow two figures describing
the ionosphere in the first half of the summer season 1953, for which
no stfldy of the signal intensity departures has been attempted. Figure
111-23 portrays the ionosphere for Anchorage-Barrow during May-June 1953
and Figure III-24 does the same for Fairbanks.
In general the h'f histograms follow the pattern of the ionospheric
characteristics in the late winter, spring and summer of 1952 presented
in previous reports. Conditions of total absorption (a) occur mostly
in the half-day Centered at 1200 hours. The tendency to peak at local
noon seems more pronounced at equinox thsr in winter or summer.
At the other end of the absorption range, the quietest conditions
occur predominantly between 1000 hours and midnight, with the afternoon
and early evening hours most favored. In winter the quietest conditions
are confined to the hours before 2000, and the maximum incidence is about
1400. In summer, on the other hand, the b-condition occurs most fre-
quently in the evening hours and has a peak about 2000 hours.
The complicated h'f records believed to be associated with auroral
activity are about equally frequent at all seasons of the year, accord-
ing to the Anchorage-Barrow analysis. But at Fairbanks no records of
this type were reported in the spring and summer of 1953, in contrast to
previous seasons. Fairbanks records.also indicate a substantial decrease
Ln the number of 7 and 5 records in the spring and summer of 1953.
Their place was taken-for the most part by increased numbers of a's.
Some question exists if the augmmt;E:dnunber of a's in the spring and
summer of 1953 is genuine." As mentioned above most of the a-classifi-
catiods in April, of which there were many, were considered to be doubtful.
Category 2 is comprised of generally poor conditions. In the
Anchorage-Barrow analysis, it is made up of cases of partial absorption
at both ends of the path, together with cases of complete absorption at
only one end. At Fairbsnks the corresponding designation, pl stands for
partial absorption of the h'i' trace. Category 2 appears to occur with
somewhat greater frequency in summer than in winter.
Appendix III - page 4
411-
?
?
When the distributions of all the characteristics are taken into
account, it seems probable that HF conditions in the Arctic are better
from 0600 to 1800 in winter than in summer, whereas nighttime conditions
(1800 to 0600) are likely to be better in summer than in winter.
Certain of the diurnal variations of the h'f characteristics may be
associated with the diurnal pattern of VHF signal intensity. The concen-
tration of both complete and partial absorption in the hours around mid-
day coincides with the high values of signal intensity recorded around
local noon on the Anchorage to Barrow path. Furthermore, it is probable
that the diurnal minimum of absorption in the early evening hours is as-
sociated with the diurnal minimum in VHF signal intensities about 2000
hours.
5. SUMMARY AND CONCLUSION
The relationship between VHF signal intensity and BF propagation
conditions, established in the first two reports, has been demonstrated
to exist in auroral regions in the autumn and winter seasons as well as
in the spring and summer. It has been shown that intensity of VHF scat-
tered signals observed at 50 Mcis is correlated with degree of ionospheric
disturbance, higher signal intensities being dbserved during conditions
of greater absorption.
REFERENCES
1. D. K. Bailey, R. Bateman and R. C. Kirby, ":Regular VHF Ionospheric
Propagation Observable over Long Distances," NBS Report 8A111, June
30. 1952.
2. D. K. Bailey, R. Bateman and. R. C. Kirby, "Second Report on Regular
VHF Ionospheric Propagation Observable over Long Distances," NBS
Report 8A1171 June 30, 1953.
Appendix III - page 5
Declassified in Part - Sanitized Co.y Ap?roved for Release ? 50-Yr 2014/06/09 ? CIA RDP81 01043R00300018florm4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
50
40
? 30
20
CLASS BLOCK )
10
0
50
40
30
20
10
0
50
40
30
20
I0
0
5
4
3
2
III
- Kp=0
?
I I
32-
I
35
I
12
I I
I
II
Kp =1
I
I 1 I
220
I
176
II
32
29
?
1-
- P= 2
264
249
44
Kp= 3
?
38
303
257
58
36
?
?
- Kp = 4
23
254
249
43
Kp= 5
163
145
42
- Kpz6
-
-
-
1 -
-
r---
4
69
78
?
Kp= 7
20
1
25
13
1
1 1
1
1?
1
1
t 1
1
-20 -10 0 +10 +20 -20 -10 0 +10 +
DEPARTURE OF MEDIAN SIGNAL INTENSITY FOR THREE-HOUR PERIOD FROM
MONTHLY MEDIAN SIGNAL INTENSITY FOR SAME THREE HOURS, DECIBELS
CEDAR RAPIDS-STERLING PATH (49.8 Mc/s)
COMPARISON OF SIGNAL INTENSITY OBSERVATIONS
WITH PLANETARY MAGNETIC K-INDEX
SEPTEMBER 1951 THROUGH SEPTEMBER 1952
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
FIGURE 111- I
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
I
+12?'
ANCHORAGE- BARROVI
+8
(48.87 Mc/S)
- 2902 OBSERVATIONS
CEDAR RAPIDS -STERLING
(49.8 Mc/S)
2978 OBSERVATIONS
I II I I I
CEDAR RAPIDS- STERLING
(49.8 MOS)
2978 OBSERVATIONS
[PLOTTED POINT IS THE MEDIAN VALUE OF THE DEPARTURES OF INDIVIDUAL
OBSERVATIONS OF SIGNAL INTENSITY FOR THE THREE HOUR PERIOD DURING
^ WHICH K IS MEASURED, FROM THE MONTHLY MEDIAN VALUE OF SIGNAL
INTENSITY FOR THE SAME, THREE HOURS. UPPER AND LOWER ENDS OF
THE LINES THROUGH EACH POINT REPRESENT DEPARTURES EXCEEDED 10%
AND 90% OF THE TIME RESPECTIVELY]
-12 ?
1 1 1 1 1 1 1 1 1 1 1 1 I I 1 1
0 2 4 6 >7 0 2 4 6 >7
,.
?
K - INDEX.
0 2 4 6 >7
COMPARISON OF SIGNAL INTENSITY' OBSERVATIONS
WITH MAGNETIC 1 0
-z
< 0
This report describes the development of a fixed-frequency
narrow-bandwidth receiver for recording the intensity of continuous-
APPENDIX
A NARROW-BAND RECORDING RECEIVER
G. Franklin Montgomery
1. INTRODUCTION
0 cc m cp
0 0 z wave VHF-scattered. signals in the 25 to 60-,Mcis range. The N
labora-
_
1 i
\
tory e o n s t r
u e te d receivers were designed to operate at 36.0 and i . 8
17i\ me/
s.
_ / //1%:\::I// :\\ w intheeourseoferinta2work,ithasbec0nneeessary:to
1 ,/ \\ w2 masurethentensttyo:scaersigasthatarec0nsierablyweale r
L hanthosedealtwithpioly.Tomakeaceuratemasurementsthe
It'
/ \ _J Z
m cz CO Z
/ \ cn o received. signal should. be as large as possible compared with the back-
/ \ w ..,(9
cr .... cn ground noise, and. this requirement is met by using a narrow receiver
/ \ bandwidth. The ultimate limit in usable pre-detection selectivity for
-' / \ 11 i
/ \ i- 0 (r) a monochromatic signal depends upon the frequency stability of the
w a z
2/
\ transRitter and receiver. Long-term frequency stability of one part n. 2 a
P / \ I 0 cc . in 10? is obtainable practically, and. for an operating frequency of
1-- / v-,:-; Li. z 0 1_-
_ N o, u
- w? 50 Mc/s and opposing. transmitter and receiver frequency drifts, the
o
/ 5: =
. maximum frequency difference with this stability is one cycle per
0 / i-
r- w 0 0
\ < ? second. This figure suggests that a receiver bandvidth of several
\ 0
a. cycles per second should be usable if the selectivity curve of the
(7) o I
0\
o (7) w u- receiver is flat-topped. Previous experience has shown, however, that
u_
/Ag 2 ? o scatter signals. are not monochromatic and. that at 50 Mcis the signal
0 u) w 0 bandwidth occasionally may be several tens of cycles per second.. Accord-
\
CD
('?..\ /71 u- a o (r) ingly, it seemed desirable to specify a receiver whose bandwidth could
-
% s \ // i 0 cc m w be changed. at will from about twenty cps to several hundred. cps.
1 \ ? I- LI
/ i
\ \,\N / /// z _.1
i-
i , / u)1- LI z 2. DESIGN
:///
0 \ \ Et. a_,
\\\ II,.1 I 0 >
eL u_ .cr o
M u- cc a
o 0 a. w
0
-w
0 Cil
0 W Z
_, N <
Z C.9
0 > cf) cx
i 0 l?.1
Cn .1 cr X
o
0
?
Selectivity of the desired. order might be attained. at the signal
frequency by the use of crystal filters in a tuned. radio-frequency
receiver, but a more practical arrangement is the usual heterodyne
technique wherein the signal frequency is reduced. before filtering. A
superheterodyne receiver 'would. require at least two frequency conver-
sions in order to provide sufficient selectivity and. image rejection,
and. at least two stable frequency heterodyning voltages are needed. in
such a system. A third. possibility is to heterodyne the signal, using
one stable heterodyning voltage, directly to the audio-frequency range,
where filtering can be accomplished. by simple means, and. to make spe-
cial provisions for eliminating the single image response. The receiver
to be described. uses this principle.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
A block diagram of the receiver is shown in Figure VII-1. The antenna
output is first amplified at signal frequency. It is then fed separately
through two radio-frequency tuned circuits whose outputs differ in phase
by 90 degrees. The tuned-circuit outputs are applied to two mixers which
receive a common heterodyning voltage from the local oscillator. The
intermediate-frequency outputs of the mixers, in the 1000-cps range, are
modified in phase by two audio-frequency-phasing networks and are then
added in a common' amplifier. The purpose of the twin nixing and the RV
and IF phase shifts is to eliminate the receiver image response that lies
approximately 2000 cps from the signal frequency in the direction of the
local oscillator frequency. The sense of the 'phasedifference produced
by the RF and IF networks determines whether the observed signal frequen-
cy is above or below the local oscillator frequency. Reversing the sense
of the phase change produced by either set of networks reverses this'
relationship
The combined intermediate-frequency output is filtered by a bandpass
filter whose frequency characteristic is centered at 1000 cps. This fil-
ter essentially determines the selectivity of the receiver, and since the
control of selectivity-is concentrated in one unit, it is convenient to
expend as much care in its design as is necessary. The filter units are
constructed in plug-in form, so that the receiver bandwidth may changed
easily-by substituting a different unit.
The filter output is amplified in a relatively-broad resistance-
coupled amplifier. The amplifier output is rectified, filtered and fed
to a D.-C amplifier which operates a recording milliammeter. The filtered
DC output is also available as AVC bias for the first three IF amplifier
stages so that the meter indication will be an approximately logarithmic
function of average signal amplitude.
Complete circuit diagrnms of the receiver RF, IF, and power-supply
sections and of the filter units are given in Figures 11II-2 through
11II-5. Photographs of the receiver unit, except for the frequency con-
trol'unit, are shown in Figures VII-6 (A) and (B). The stable-frequency
local oscillator units were developed separately and are described in
Appendix IX.
3. PERFORMANCE
The over-all gain of the receiver, that is, the ratio of D-C detec-
tor voltage to open-circuit 50-ohm source voltage, is about 2 x 108 for
maximum IF gain in the manual gap position; The maximum gain of the IF
amplifier itself is about 3 x 104 at 1000 cps and is 3 db below the 1000-
cps gain at 500 and 2000 cps. The open-circuit source voltage for full-
scale meter deflection is 0.1 microvolts at maximum gain. With the IF
gain reduced, meter deflection remains substantially linear for inputs up
to 100,microvolts. In the AVC position, and with maximum meter-circuit
gain, full-scale deflection is obtained with an input of 7 microvolts.
App VII page 2 -
With the meter-circuit gain reduced, meter deflection remains substan-
tially logarithmic for inputs up to 200 microvolts. With the 60-cps
bandwidth filter, in either the manual or AVC positions, an input of
.01 microvolts gives a measurable meter deflection over that due to
receiver noise. This figure represents a detectable available power of
5 x 10-19 watts. The receiver noise figure at 50 Mc/s is about 4 db.
Image rejection is about 26 db.
4. DISCUSSION
Two problems encountered during the development have not been com-
pletely solved. The first is microphonism in the IF amplifier; the
second is local oscillator noise.
In operation at maximum receiver gain, the signal level at the IF
input is on the order of 100 microvolts, and the signal level at the
mixer input is on the order of 3 microvolts. Since the output voltage
from the mixer circuit onward is at audio frequency, tubes used for the
mixers and succeeding stages should have stable physical structures if
microphonics are not to be troublesome. In practice, the worst sources
of microphonics have been the first two 6SG7 tubes in the IF amplifier;
the 66 tubes used in the mixer and combining amplifier stages have been
relatively quiet. It is probable that pentodes especially designed for
low microphonism would be useful substitutes for the 6SG7 in this appli-
cation, but none of the available low-noise pentodes has the remote-
cutoff grid characteristic that is desirable when AVC is employed. Suit-
able choice of 6SG7 tubes results in a nicrophonic level that is not par-
ticularly objectionable in normal operation, althonch vibration from
auxiliary equipment must be prevented from reaching the receiver.
Microphonics must also be avoided in the local oscillator. In prac-
tice, microphonics have not been a problem in the local oscillator designed
for the receiver, but oscillator random noise is of sufficient intensity
to require two RF stages ahead of the mixer. The effects of microphonism
and local oscillator noise could be reduced still further by using more
RF gain and less IF gain. If this modification were made, it would be
necessary to apply AVC and. =mini gain control to-one or more RF stages
to prevent overloading in the receiver front end.
The oscillator input to the receiver mixers is about one volt.
Since the signal level at this same point is about 3 microvolts, modu-
lation of the local oscillator voltage that falls within the IF passband
must be considerably less than 10-4 percent. Both amplitude- .and. phase
modulation of the local oscillator voltage are objectionable, since either
results in spurious noise output from the mixers. In general, the random
modulation present in the heterodyning voltage obtained by direct fre-
quency-multiplication of a stable, low-frequency source is too large to
be tolerated. One workable solution is to use a crystal-controlled oscil-
lator, operating directly at the heterodyning frequency, which is locked
or otherwise controlled by the output of the stable generator. The noise
spectrum of the crystal oscillator is found to be much smnller than that
of the controlling source.
App VII page 3 -
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release . 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release . 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Not.
All R4.31
All Gin 11111unloos olhormse shown
1.2,11_G/
? L.. L3. L4 . 1... L. , AND L7
3 TURNS NO, 16 FORMVAR CLOSE?WOUND
IN CENTER OF NATIONAL 31150 COIL FORM
RFC 130 TURNS FORMVAR CLOSE?WOUND
ON I/4" CERAMIC FORM.
T ? HALLOORSON 0 6001 2 WATT DRIVER
TRANSFORMER-8-12K a PLATE TO PP GRIDS
?IX jI_AC
Li. L.. L3. 1.4, L.. L.. AND L7
5 TURNS NO. 16 FORMVAR CLOSE?WOUND
IN CENTER OF NATIONAL XR50 COIL FORM
RFC AND T SAME AS 411.11 MC
Capacitors in microfarads
L-0.3 Henry (Burnell 5-7475)
Cu Adjusted for mon voltage across Co
with output shorted. C2 adjusted for min. voltage
across Cu with output open
?
R
IN
1
100K
ILvlc0.1
.,._see note__ _,.
?
0.1
L
47K
OUT
i
Capacitors in microfarnds
L-0.3Henry (Burnell S-7475)
0.1 capacitors adjusted torescoance
at ini.o cps wit h.0047capacitors disconnected
Capacitors in microfarads
L-3 Henry (Burnell S-6762)
Tuning capacitors adjusted to
s000 cps with .05capacItor shorted.
%
,AEMIECORDING:RnEi E
tommeo worsT tor
B -BOTTOM VIEW OF IF AND DETECTOR SECTION LEFT,
AND RF SECTION, RIGHT, DUST COVERS REMOVED
PHOTOGRAPH OF RECORDING RECEIVER
FIGURE VII-6
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
?
APPENDIX VILE
THREE KILOWATT VHF TRANSMITTER FOR
RADIO PROPAGATION STUDIES
W. B. Harding and. D. C. Whittaker
1. INTRODUCTION
The 3 KW transmitter is designed to operate at a fixed frequency in
the range 30 to 75 megacycles, and is capable of long period, unattended
operation. Compactness compatible with ease of servicing was an impor-
tant design objective, since many of the transmitters are to be installed
in small field laboratory-buildings or trailers. The single phase power
supply design was intended to provide for operation at remote sites where
three-phase power is often not available.
One unit has been in substantially continuous operation on 49.80 Nic/s
at Cedar Rapids, Iowa, since March 1956. In this service it has operated
at a reduced power output of about 2.2 kilowatts as a means of avoiding
service calls. Apart from thrice weekly service calls to log the readings
of its meters, the transmitter has required replacement of components or
tubes on the average less than once every two months.
Additional units have been built in connection with other projects
for 30, 36, 4o, 50, 54 and 74 Mc/s operation. Application to other fre-
quencies within or beyond this range involves the relatively simple matter
of designing appropriate inductors for the tuned circuits.
2. GENERAL DESCRIPTION
. A complete setup is shown in Figure VIII-1. A, block diagram of the
system is. shown in Figure The system includes the following
separately identifiable units. The 3 KW Power Amplifier unit includes
plate, screen and grid bias power supplies, heater current supply, certain
protective circuits, and a VSWR indicator in addition to the amplifier it-
self. The Driver is a two-stage linear amplifier powered. by the Driver
Power Supply. The Control Unit Contains an automatic Reset Circuit, high
VSWR Protector, Off Air Alarm, and forward and back power recording cir-
cuits; The reset circuit attempts to return the transmitter to operation
if a transient overload occurs. The VSWR Protector interrupts translis-
sions during periods when the voltage standing wave ratio on the antenna
transmission line exceeds a preset limit. The Off Air Alarm is triggered
when the transmitted power falls below a:preset level. The Recording ;Cir-
cuit facilitates graphic recording of the output voltages from the forward
and reverse power diodes in the directional coupler. The Break Keyer
interrupts the transmission for, an adjustable period of about two minutes
each time a one-second pulse is received from a programmer clock. The
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Identification Keyer provides for the transmission of the station call
letters each time the Break Keyer produces a two-minute interruption.
The Multiplier-Exciter accepts the approximately 1 Mc/s signal from the
Near One Megacycle Standard Oscillator and multiplies its frequency to
the desired transmission frequency; this unit amplifies the signal for
adequate feed to the Driver. (See Appendix IX for a description of the
Standard Oscillator and Multiplier-Exciter.)
The various units are described in the following sections. All units
except the 3 Kw power amplifier mount in one standard 19 inch wide by 73
inch high relay rack. The 3 Kw power amplifier unit is housed in a special
rack 71 inches high, 29 inches wide and 19.5 .inches deep, and weighs 1500
pounds complete. Power required is 60 cycles, 230 volts A-C in a three-
wire balanced single phase 50 ampere entrance. The total power consump-
tion is 5.91(w.
3. POWER AMPLIFIER
The power amplifier is a completely self-contained unit with bias,
screen and high voltage supplies. The output circuit contains a direc-
tional coupler and meter to read forward power (power fed to the trans-
mission line from the transmitter), back power (power reflected to the
transmitter from. the transmission line due to mismatch), Voltage Standing
Wave Ratio or "SWR". Output connectors permit recording of both forward
and back power voltages generated by the directional coupler. The unit
contains all of the conventional protective circuits, such as: time de-
lay for application of high voltage, overload protection, low bias voltage
or "under-bias" protection, and chassis interlocks. The amplifier is
capable of 3.2 kilowatts output into a 50 ohm load, althouel an output of
2.0 kilowatts is recommended where reliability is of utmost importance.
Conventional circuit design is used throughout the Power Amplifier
and its associated circuits, as shown in the complete schematic wiring
diagram, Figure VIII-3. The amplifier proper uses two 4-1000A tetrodes
in a,grounded cathode, push-pull configuration. The amplifier chassis
together with the front upper cabinet panel on which the operating meters
are located is binged and swings forward and outward through 90 for ser-
vicing. When in the operating position, the upper part of this unit meets
internal shield panelling of the main cabinet _so that the 4-1000A tubes
and the entire RF part of the plate circuit ai-e completely shielded except
for one smn11 opening. This opening is the entry for the adjustable pick-
up loop which couples the plate tank power into the directional coupler
and hence the antenna. Figure is a rear view showing component
arrangement. Figure VIII-5 illustrates the plate tank section as seen
from above for a 50 Mc/s unit. Figure VIII-6 shows the bottom view of
the 4-1000A sockets with associated grid drive circuits for the 50 Mc/s
transmitter, and the fuse and main circuit breaker panel as well as parts
of the bottom of the 4-1000A chassis, seen from the rear of the cabinet
with the amplifier section swung forward into the servicing position. The
-Interlock and Protective circuits are described in connection with the
Control Unit since the two operate together.
- App VIII page 2 -
4. DRIVER AND DRIVER POWER SUPPLY
The Driver is a two-stage linear amplifier. The first stage is a
2E26 operated class A. A, potentiometer, labelled "Output Level", varies
the screen voltage and thereby permits control of the stage gain. The
grid and plate circuits are tuned and link coupled. The second stage
design depends on the frequency of operation, one model applying to fre-
quencies below about 60 Mc/s, another having been designed for 74 Mc/s.
A complete schematic diagram of the Driver, including both forms of the
second stage, is presented in Figure VIII-7; the power supply is shown
in Figure VIII-8.
The second stage, for frequencies below 60 Me/s, is a pair of 6146's
in push-pull operated class A32. (For frequencies above 60 Mc/s, the
second stage is a single 4X150A operated class A32.) Link coupling, is
employed and the second stage is neutralized. All elements of the tubes,
with the exception of the second stage plate current, are metered by means
of a meter switch and a 0-50 microampere meter. Second stage plate cur-
rent is metered with a separate 0-300 milliampere meter. In some models
there is a relay which is used to break the cathode lead of the 2E26 for
keying purposes.
The Driver Power (Figure VIII-8) Supply is of conventional design.
(The 4X150.A model Driver required 800 volts for the plate, in contrast to
the 600 volts required for the 6146's of the lower frequency model, and
therefore employs a capacitor input filter. The 4X150A. model Power Sup-
ply also furnishes 110 volt A-C to operate the blower in the 4X150A air
system.)
The lower frequency model furnishes approximately 60 watts of RF
power into a 50 ohm load. The higher frequency model furnishes approx-
imately 80 watts of drive into a 50 ohm load. The higher driving power
is required for the 4-1000A's as their maximum operating frequency is
approached. Any source of RF energy at the operating frequency, which
supplies about 2 volts rms to a 50 ohm load., is sufficient to drive the
Driver. The multipliers and exciters are discussed in Appendix IX.
5. CONTROL UNIT AND PROTECTIVE CIRCUITS
The Control Unit, used in combination with the Interlock and Pro-
tective Circuits of the Power Amplifier cabinet, is described in Figures
VIII-9 to VIII-11. The functional block diagram of the complete system
(Figure VIII-2) shows the role of the control circuits in relation to
the rest of the transmitter.
The Overload circuit interrupts the A-C power supplied to the high
voltage supplies when the direct current flowing in the high voltage
plate supply exceeds a predetermined amount -- usually about 1.2 amperes.
However, it offers no protection against a short circuit in either of the
main filter capacitors. When the Overload circuit is tripped, the Over-
load Reset circuit automatically commences operation, attempting four
App VIII page 3 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
times within two minutes after the overload-to re-close the main A-C
contactor. This automatic reset will in general correct a transmis-
sion failure caused by a temporary overload, such as night accompany
lightning striking a nearby object. If the overload condition per-
sists during the reset cycle, it becomes necessary for the transmitter
operator to manuarly reset the system.
The SWR monitor turns the transmitted carrier off when reflected
power from the antenna indicates an excessive standing wave ratio (SWR)
on the transmission line. This is intended to protect the transmitter
and also to avoid transmissions during periods when the accuracy of
measurement of radiated power is under some doubt. High SWR may be the
result of weather conditions, ice, snow, rain, or it maybe the result
of a failure in some component of the antenna system. A portion of the
reflected RF power is rectified by a crystal diode, called the "back
diode", that is located in the directional antenna coupling unit. The
voltage thus produced will be zero under conditions of perfect match and
will increase with the degree of mismatch present. At an adjustable
preset voltage, representing a standing wave ratio of about 2:1, a relay
disables the transmitter and simultaneously starts a 4-RPE motor and cam.
After a 15-minute interval, the relay is de-energized, and, depending
upon whether conditions have changed, the transmitter will either return
to operation or will again be disabled for another 15-minute cycle.
The Off-Air Alarm is triggered when the forward power fed to the
antenna transmission line falls below a preset level, usually 75 per-
cent of the desired level. The output of the "forward power" diode in
the directional coupling unit is a voltage representing the output power
level of the transmitter. When this voltage drops to the predetermined
level, due to a malfunction of some component of the transmission sys-
tem, an alarm sounds and the carrier indicator lamp on the control unit
goes out.
In the.Recording.Circuit the output voltages from the "forward"
and "back" power diodes of the Micromatch coupler are amplified and
fed during alternate periods of 7i minutes length to a single 0-J. ma
Eaterline Angus graphic recorder. Adjustable voltages from the Off-
Air Alarm and SWR Protection amplifiers are used for this purpose.
Switching circuits permit calibration of the recording in terns of the
readings of the directional coupler indicator in the Power Amplifier
rack. The recordings are used as a means of standardizing the strength
of signals received from the transmitter to one arbitrary level of trans-
mitted power. Atypical record is shown in Figure VIII-12.
6. BREAK KEYER AND IDENTIFICATION XElER
?
Two keyers are used in the system, the Break Keyer (Figure VIII-
13) to make the two-minute noise break on the hour and half hour, and
the Identification Keyes (Figure VIII-14) to send the call sign at the
beginning of the break. The two-minute half hourly break is required
- App VIII page 4 -
.18
OP
in the receiving operations this transmitter is meant to satisfy. A
programming clock controls both keyers by a pulse sent every half hour.
This pulse operates a time constant circuit which opens the keying line
for two minutes and starts the identification keyer which repeats the
call letters twice in a period of twelve seconds.
- App VIII page 5 -
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
PHOTOGRAPH OF COMPLETE SYSTEM IN A TYPICAL INSTALLATION
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
OFF-AIR
ALARM
115 VAC
INC
OUT
I WIRE CABLE
DRIVER
iimool
INC
RG5 el/U
???
CONTROL UNIT
115 VAC
RGS BM BACK
1
ANC
FiG5 OM FORWARD
0-
0
0
41
01
BREAK KEYER
II5 VAC
WIRE CABLE
IDENTIFICATION KEYER
ROD
INC
OUT
MULTIPLIER
115 VAC
REG.
II
lID VAC
ANC
0-43
OUT
FREQUENCY STANDARD
ONO
IN
05 VAC
DRIVER POWER SUPPLY
g.r,22,2
o ? I + +
1
11111
II
NS SAC
LINE VOLTAGE REGULATOR
0-?
0-?
t?J
41
ea
5 WIRE CABLE
LC
0
OUT
KEY TO OCTAL SOCKET
I- YELLOW -
115 VAC FROM CONTACT
2-BLACK-
115 VAC FROM CONTACT
3-WHITE-
INTERLOCK
Al-BLUE-
INTERLOCK
3-GREEN -
5 -NOT USED
1-RED-
OVERLOAD RESET
0- BROWN-
OVERLOAD RESET
OCTAL SOCKET
EA
RECORDER
I MA FS
2 WIRE CABLE
0
tI5 VAC
to Go,.
15 AMP
PAOGRAM CLOCK
BLOCK DIAGRAM OF COMPLETE TRANSMITTER SYSTEM
FIGURE VIII-2
III
230VAC
3 WIRE
I# BO
50 AMP
?
?
?
?
?
?
?
e
?cE
C'
EEO
0115 1011
1101101 ISIS
0
CASA 100
0
171101*1 A011
311110A MI
0
010013
655555
0
15110110 A011
015013
41000
111101111 Loll
OISA 101
010015
0101 Ill
nrinn *00
00111
7 7
ea
F. F..
=22
a2
ggfE:32:7:0
E-E;g:.T7?T
'ERE
Ngegt
!it=
e.149-14
sag
ti)
sin NS
IA!
S.
;.1
?t"
?Q4
t-fg
Z
"1111
2t.
? IP411
C
(INCLUDING POWER SUPPLIES,
Ui
cc
Ui
a.
Ui
Ui
?.J
a.
2
Cc '-',-,0 ??)"- Imo, ;_,. (..-.4--1-i ;"-- r: , -...--,-.. , rr, .11' -', ? c r.,-, 1 arr-c -
t.? Fil"L'i ,' 011-,IT.Ill lion in fizt,i,-?12111,f r ?104 111, vr?-,-,? ir.v? ',..r.nr.-1.
te,nna ^_olltz,:r.I.-_-7.1c.r, ., ....a-. -..'r...::-It.- . _r-.=-2---.7e7-.-
an,,, 1 a.z, ,,,, t i i_ci-.. .,,,, e.viikly rid CI.1,12.-.......11,.._
, -..nrgte-,s v?Inc-,..:-,, _.
nee wn MI( n t Ff..4ti_l,i p_
t, n#I;-_i-_1i- ..r.t.".-r.?.o(:?f;ir,.rr:n-..1t -:e,-a:-_:-:-,j,-r,)...,?.:-_- ---I !:-,i?, ,i,_;i1:-, L-i.1 .4--1.-L-..,-, 1"*a.-,i,.7,,,- 7.-. c-1...1..7-,,,,,i-"-.7.'.""?,1i i"2,?-" .5,7.- )...
1.-.7 v-.--%-tc_i 1_11/2 --5-.2._?..-In-_,7 ..3.!.:ZinT...":1- 7.1%. ?.,.Z.;e..;
than in- t lc,- (1;rittrn, it
t ?,-ai,.: ,::41 1.3--t..
r ..,.a....-: 7.'.,::.... d . ,. ? _ . ? i -- t ' .. . a. _ ::-. .".. tar
sptieric. _in kips)! n r - bei,e;.h i :_of f_,..45;..,-..kilt.,-irir ?
4.)1-9?14f.h.tri il
usE u
..,,.,.._-:. UTIATV-&-t'A-141'ir.,-..,i=?1,7i.:1,-.?:',.1-ii---.LL.2:4:.''.-a.,-,
rjcerr:igligg..1.21iiff:fi.41.:1:1:1)(iifel..-LIF.1%--
low
),I41 121P-WA:1-2 f.(1:D.,;:_.(?21.14.ti:t?Lt.) .tfjl.g1"11Plis-2..iF.5147-A.r:_-?-
'' 1 -e"Wria -Ifiddli t,,Piftr-La
If elk/9k WeWligi8V7).' g:).-u_Ogi 11 e_1;11
:--uc_c.,_.essful,pric2t,_121mii&
'
pinseit4sover ( rscia4es 55_12i.itisr.c..1,4,_1,11:_a_, ti.J.,_;:,..2,7,,,nymu...._
- j , ?r . I , j1 - ( '''? - ?,111- , vi
3: ... Dos, over 1:.... . nce- ..-- ,,,..--. -..-;-_--- , ,- .r? -,.-;
tan er, 41S r'S.,CCIp..p811.07.) MC.4?y.ii. -iir,irl,e.r..!,1.....,-1.;,:',..L.i- i,-A.1..a4-c. Let_rite_
4,. taneog. r-,..
,-....' enc e nutset
p retiuire- conskp.stuOlfti.poJ:or .ruiizu .__
'.., ence pu.ses, require co.:___ . -
it
- 2'.'..N'Iccad4 L. Raw-,w?P(1...P, i'L;-.t.,.:.-A..-w4:-.,_.'_?..S0437., ,._
.y, rad, a iii7nI.--a tirad try ikririe'n des-L.31-M_ ft, tehcrii,,.-.4.,..11-'_.."47).r0C.?::::::
q.',f7r1;r:c; .?"5 i riltfccpr4gr-g2-5...%g41Zt f),-1,,..,i,-,_- - - - -
' ." 9
i'cp.2. -TI fralifeyi li
lni.tiaff-I-A3t-lialtifi:-Ifit'''f,' ` '- !-:- I
"17,1?-
'_crmr.rae.
4ite.r.c? ;4-Z>.
.1,-TXWtr.
? I.,- 4:31TMerarara"..=
?
f-t".4141 ? ? tudb - ettr_
r
arr.
?
--Tl.ikG .1 04,1-- .C.AaLkstt
77:417?1_`: ?=efs
1Q-AITC?1.41.:,"
-
1-FEFF-FZE,PME :134t
')12`.E3'?". "
tr?
" -J9v.itierrt f qi57,14tat.r.
_
' t-: l'iu#11:citto.:.11.VJA
)13 4-46citt:
n iCit,t4tr.
I tx&n:tv..wdAtzlile.t'
.7.,1
zfroCi4)
,43
;;-.F..41U,i:.-c.:-1.-P.511/1:' of -cilla 1 1111,-',:q.s.i
_ _
V. r x.4114151.4)
TIT-17;4.4
1
' h taEk_e_"iE.-5?9:t4.T141..bsiff-4,v..tk-e"c?L1:11.1111)10;
I r;ij,.j) atldaI 0-._.1}1
_Flrai 1tUtll.p" .
)(:.9i.s-4-..c;*.1-00..6.?_8.s- 11.;Lwltkjii. ;so
th
Pe EISIS4141p14Yetal4VI Cjurrjr$-_,t -?ssag ens.1;Al
-14
-
5,_.:41-1,.ti-id.14nc_.-Airil"zi.:
- _L ?
`1).Tibt!P -:9.413a.Vicf4Eql'crr
.dav-TIre*we?1.1erOxLtecorc.le.ci Au id, riec9rgetie,
.0U ID
r..
009) RP
KIN
((G; OC):, )
I II 5unlO OG ) 11)1
0
0 0PNi vsNA?@ rt5g
0
- 0 051
(1=40 009 ) qv
11 'IDA I b3
I I
1 1 1 1 1 1 1 11
("Ago us ) G9
t.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
PROCEEDINGS OF THE IRE
October
1955 Bailey, Bateman, and Kirby: Radio Transmission at VHF by Scattering in the Lower Ionosphere
($A040 cas)q? wisokiaba
VP 51 X
II I I
(su)I u os) qp
88 8 2
/-.41-1;1111401111_Thilil
-
lip 1; 1111101i
111:
II
I '
d
-
je 1;,Lw;
ih
?-frrri:?
10. .1 ? .2/1
I 1
' 4L4--T'n
I ';? lit: c
ii;111 tilt IP
E 1111! I
hi
11; (17.)
ilIIi1111
jij
swqo 009) qp
("WO 009) qp
2 8 ' 8
I I
(1.43 009)qp guemApM3
; ; g
I I I I I
3 3 Z 2
1111111
(suAto os)qp
("40 OG)11P
50-
40-
30
20
I0-
CEDAR RAPIDS - STER UNG Ing
I243km 49.80Mas
TOTAL NUMBER Of HOURS OBSERVED 82
-
- --,..._
.--
-
02 04 06 08 10 12 14 16 18 20 22
75? W TIME
1251
FARGO -CHURCHILL 70714. ItUM9ER Or NOuRS OBSERVED 219
1326km
49.70Mc/s
?
7-1--)?,--- --F-
02 04 06 08 10 I2 14 16 18 20 22
90 W TIME
1952
ANCHORAGE'- BARROW 707AL NUMBER OF HOURS OBSERVED 82
I156km 48 87Mc/s
-

?
08
02 04 06 08 10 12 14 16 18 20 22
150. W TIME
DCC 1952- rm
TOTAL
NUMBER OF HOURS OBSERVED.
277
GOOSE BAY - SONDRE STROMFJORD _
1608km 48.02 hicis
?
?1?.-. ,?-- ---
06 08 10 12 14
606 W TIME
CEDAR RAPIDS - STERUNG
I243km 4980Mas? i.
Ir
1.
PERIOD
OF OBSERVATION
r-
?
1 rn-1-11
=/=. r
ITh
_ .1-
,- )
JFMAMJJASONOJFMAMJ JASONDJFMAMJJASONDJFMAMJJASOND
1951
-, 1952
1953
1954 .
1
FARGO -CHURCHIL
I326km 49.70Mas
.
1
14
41
'
?
?
'
M.-.
?7
-r
-um -n
,
JFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASOND.
195 1952 1953
1954 ,
'
ANCHORAGE-BARROW
- I156km
48.87Mds
_
--
-ri-
--F11-,4--n-
1
JFMAMJJASONDJFMAMJJASONDJFMAMJJASONOJFMAMJJASOND
)-- 195 1952 1953
1954
1.-
?
114
)., GOOSE BAY - SONDRE STROMFJORD
1608km 48.02Mc/s
-
-
-
,
Fig. 25?Diurnal variation of occurrence of sporadic-E
propagation for yearly periods indicated.
JFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJEMAMJ.JASOND
1951 1952 1953 1954
a mples of.routine recordings illustrating such signals
are in Figs. 23, 24 at left. Special records of these
events have been kept in order to determine their diur-
nal and seasonal characteristics. Examples are consid-
ered in this analysis if, at some time during a period of at
least twelve minutes during which the normal signal is
masked, the signal levels exceed one millivolt for a refer-
ence impedance of 600 ohms. A criterion of this sort is
necessary in order to eliminate the occasional long-
duration meteor bursts. Whenever possible, interpola-
tion through the sporadic-E event is performed in con-
nection with the data analysis of the normal signals.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Fig. 26?Seasonal variation of occurrence of sporadic-E
propagation.
The sporadic-E observations are summarized in Figs.
25 and 26 for the paths operating at about 50 Mc/s. In
these summaries are included the observations at 48.02
Mc/s for the experimental communication path from
Goose Bay, Labrador, to Sondre Stromfjord, Green-
land. For the temperate-zone path from Cedar Rapids
to Sterling most of the sporadic-E propagation occurs
during the summer season. It may occur at any time
of day, but is more likely to occur during the day?
light hours. A secondary but much weaker maximum
occurs during the winter months, which may have
somewhat different diurnal characteristics. Sporadic-E
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
PROCEEDINGS OF THE. IRE
October
1955 Bailey, Bateman, and Kirby: Radio Transmission at VHF by Scattering in the Lower Ionosphere
,ropagation is not observed at this frequency over this
path during the equinox months. These findings agree
!with other mid-latitude observations by other methods
'it _lower frequencies. The characteristics observed on
the arctic paths are totally different. Most of the spo-
:-adic-E propagation takes place between about 18 hours
,ocal time and about 02 hours. On the Goose Bay to
13ondre Stromfjord path the diurnal maximum occurs
;thout two hours earlier than on the other arctic paths.
,Sporadic-E propagation is almost completely absent
;hiring the late night and the daytime. This diurnal
? tharacteristic is observed in all months. There seems
:o be more sporadic-E propagation during the summer
?
, season, although the summer of 1953 had less sporadic-
E propagation than the summer before. Nevertheless
t cannot be said that these data show any certain solar
,:ycle influence. The occurrence rates for a particular
3ath were very sensitive to frequency, as would have
'peen expected. At 107.8 Mc/s during the summer of
, :1952 only two brief occurrences were observed. They
:oincided with longer events at 49.8 Mc/s. At 27.775
,Mc/s more events and a generally longer duration for
specific events were observed. A similar result was found
n Alaska at 24.325 Mc/s.
A connection between the aurora and sporadic-E
,
propagation has long been suggested. At vertical inci-
lence there appears to be a one-to-one correspondence
3etween the zenith or near-zenith appearance of well-
...lefined active auroral forms and 117 observations of
!choes from about 100 to 120 kilometers. On the other
, sand, nearly all attempts to associate the strong vhf
signals of the type under discussion with the appearance
, 3f aurora in the illuminated common volume in the E
region have proved unsatisfactory. Observations from
the terminals of the Anchorage to Barrow path and for a
briefer period from Fairbanks, where the path midpoint
I in the E Kegion is more easily examined; failed to estab-
,
I !ish any significant, connection. In the case of the Fargo
r.o Churchill path the results are somewhat less negative
, ?
lhough the connection sought is not very definite. Fig.
? '27 (facing page) shows an example where correlation be-
'ween Es-influenced signal recordings and the presence
?;)f active auroral forms in the direction of the path mid-
,point is fairly direct. In this figure the auroral forms
were actually photographed from the receiving hut and
?some of the poles of the receiving rhombic antenna can
seen. Fading twilight masks the observations at first.
; The discussion given below in connection with the
sputters phenomenon should be borne in mind in any
thempt to understand the relationship between Es-
', 3ropagation and auroras.
! Sudden Ionospheric Disturbances (SID's)
Signal enhancements at 49.8 Mc/s coinciding with
3ID's producing sky-wave fadeouts at hf have been ob-
served from the beginning of the program and have al-
1 ready been reported. During 1952 when the 107.8
Mc/s signals from Cedar Rapids were also available
at Sterling only a few occurred. They sufficed however
to demonstrate that the enhancements were of the
same amplitude at the two frequencies and that they
occurred simultaneously. Best example obtained is
in Fig. 28 (page 1206). Quite recently a moderately
strong SID has occurred when signals were being re-
corded simultaneously at 27.775 and 49.8 Mc/s. In this
case a good enhancement was observed at 49.8 Mc/s but
the 27.775 Mc/s signal showed at first a trace of en-
hancement which was followed by a sharp decrease a
few minutes after the start of the SID. This is inter-
preted as indicating that absorption associated with the
sharp increase in ionization in the D region was more
than enough to offset the enhancement once the event
had become well developed. It further indicates that
most of the scattering takes place in or above the ab-
sorbing layer during these events. At minimum the
27.775 Mc/s signal was still many decibels above the
background noise level. It is a matter of some interest
that the maximum absorption at 27.775 Mc/s occurred
several minutes ahead of the maximum enhancement at
49.8 Mc/s. The details of this event are in Fig. 29 on
page 1206.
In examining the routine records some care is neces-
sary in identifying SID's. From time to time enhance-
ments or signal variations of very similar appearance are
observed which are not associated with unusual absorp-
tion. It is also observed that the weaker SID's produce
no identifiable enhancement. The simultaneous record-
ing of some distant hf station usually provides an excel-
lent guide. Also whenever the received noise is predomi-
nantly cosmic noise and a transmission break occurs
during the event, a significant decrease in the cosmic
noise level is observed unless a simultaneous burst of
solar noise contaminates the observation. The SID be-
havior is believed to be directly related to an increase
in fN, the plasma frequency corresponding to the mean
electron density, N, in the scattering volume.
Polar Blackouts
The term polar blackout has been given to events oc-
curring in regions of high auroral activity which are
characterized by periods of very intense nondeviative
ionospheric absorption lasting from somewhat less than
an hour to six or more hours. These events, unlike SID's
which are usually less than an hour in duration and can
only occur during daylight, are observed at all hours
though they are much less frequent in the evening pe-
riod from about 18 hours local time to midnight. As
with SID's they produce complete fadeouts at hf, but
unlike SID's which occur simultaneously all over the
sunlit hemisphere, polar blackouts are fairly local. By
local it is meant that they occur over an area a few hun-
dred miles wide, outside of which conditions may be
fairly normal. They are thought to be associated in
some complex way with the corpuscular radiation from
27
t' , i? 24
MAW-- NONE
OBSERVED NOISE LEVEL
AT 48.87 Mc/s
..? -
TRICE
0.12 I 0.02
27 DECEMBER
TUT POSE MAXIMA
/PRECEDING DAT
FOLLOWING D,Y
00 02 04 06 08 ,16 12 14 16 18 20 22
ALASKAN (I50?W) STANDARD TIME
Fig. 42-Precipitation noise at Barrow chiring a severe blizzard.
If the antennas and receiving equipment are operated!:
correctly and are properly designed, receiver noise is
never a limiting factor at 50 Mc/s, though it influences,',!
significantly at times, the observations of background:
cosmic noise at 107.8 Mc/s.
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
PROCEEDINGS OF THE IRE
PART II. THE ROLE OF THE ANTENNAS
INTRODUCTION
11 iv In the preceding account it has been repeatedly em-
LtI phasized that the behavior of the observations and the
interpretations of this behavior cannot be divorced
' from consideration of the antennas actually used. In
the sections that follow the part played by the antennas
in the propagation studies will be examined more closely
and from several points of view.
;1
it;
1
Fig. 43?Geometry for Case I scattering.
GEOMETRIC CONSIDERATIONS
iSome generalization of the concept of the geometric
factors, q, defined in a particular sense by (3) and (4),
is indicated. Consider Fig. 43 where P represents the
position of an elementary macroscopic volume of the
scattering medium which is presumed to be distributed
i in a shell concentric with the earth, having a thickness
.b, small compared with its height Jr above earth. Let
M= midpoint of the path in the scattering medium,
P =a point in the scattering shell displaced from
AI in the plane containing the transmitter T,
receiver R, and center of the earth 0,
n = the angular position of P with respect to M
measured at the center of the earth,
aR = the angular elevation of P as seen from T and
R respectively,
/T, IR = the ray lengths to P from T and R respec-
tively,
ET; ER = the angles between /T and IR respectively, and
the tangent plane to the scattering shell at P,
= the scattering angle,
2d= surface distance from T to R,
October
a= radius of the earth, and
h= height of the thin scattering stratum.
The following relations may be used to determine the
a's, /'s, c's, and 7 as a function of the displacement of P
from M as measured by n according to the conventions
of Fig. 43:
tan aT ?
tan aR ?
cos Ed ? 77)
a
a
a + h
a
a + h
sin CI + 77
a
1 (a + h) sin (--d ? 77)
a
7, ?
cos ar
1R ?
cos Er =
COS ER
and finally
Or
(a + h) sin CI + 77\
a )
a
cos an
a + h
a
cos c2n,
a + h
= ar + CYR + 2?,
a
(8)
(9)
1955 Bailey, Bateman, and Kirby: Radio Transmission at VHF by Scattering in the Lower Ionosphere 1215 ;
and
2a2 sin2 ? ?
a
cos 7 ?
12
swinging transmission arrangements, which are chosen
to give further insight into the interpretation of the
(17) observations made with less than ideal arrangements.
(a + h) cos ? cos r - a
a
sin a = (18)
1
(a -I- h) sin r
sin fl = (19)
1 cos a
sin o =
(a + h) ? a cos ? cos
a
1
(20)
In Case II if horizontal polarization is used at T, the
downcoming signal at R from P will contain both ver-
tically and horizontally polarized components when
O. The power received at R from P with a horizontal
(10) antenna is measured by sin4 x, and with a vertical an-
tenna by sin2 x cos2 x.
(14)
(15)
If horizontal polarization is used the angle x previously
discussed is 90?. In vertical polarization x =90? ?7.
Displacements of P from M of the type illustrated by
Fig. 43 will be referred to as Case I.
Now consider Case II, as represented by Fig. 44 in
which P is displaced from M in a direction forming a
right angle at the midpoint with the displacements of
Case I. That is to say, P is displaced from M in such a
way that ar = aR= a, = /R =/, and c7,=?R=e. As shown
in Fig. 44 let r represent the angular position of P with
respect to M measured at the center of the earth. It is
necessary for Case II to define an azimuth i9 as the di-
rection of P as seen from T or R with respect to the
great-circle path connecting T and R. The following
relations may now be used to calculate the principal
geometric quantities, as a function of the displacement
of P from M as measured by r:
12 = (a + 102+ a2 2a(a + h) cos cos r, (16),
a
3
or
Fig. 44?Geometry for Case II scattering.
To obtain x the following relationship is used,
sin ? cos ? sin
2a a a
COS X = ?
I V? cos2? cos2
a
COS
a 2d
x = ?I sin ?a sin )3.
Fig. 45?Geometry for first beam swinging model?
"omni to beam."
FIRST BEAM SWINGING MODEL?"OMNI TO BEAM"
As a first trial consider Fig. 45, which represents Case
I geometry. The transmitter at T is connected to an
omnidirectional antenna having an aperture A0 and
the receiver R is connected to a highly directional an-
tenna having a single lobe of solid angle M2 which is in-
versely proportional to its aperture AD. The receiving
antenna is so directive that its beam Mt is formed in-
dependently of the ground no matter how it is steered
as long as the lower edge of the beam is above the
horizon. The directivity M2 is such that throughout the
intercepted elementary macroscopic volume of the
scattering medium, the values of the /'s and 7 may be
regarded as constant to a high order of approximation.
The position of the intercepted scattering volume may
thus be represented by the point P as shown in Fig. 45.
The received power is now expressed as a function of
the position of P in the plane represented by Fig. 43,
that is to say in a Case I manner. In accordance with
the definition of the scatter cross section, the following
relation can be written
P, b/n2Al2 sin2 x AD
Pr cc.
1T2 sin oR
(a)
(sin --
)"
2
(b) (c) (d)
(23)
where
(a) is proportional to the incident power density,
(b) is the macroscopic element of volume illuminated
by T seen by the receiving antenna,
(c) is proportional to the scatter function,
(21)
(d) is the solid angle intercepted at R, ?
but AD cc 1/6,St, so that the geometric aspects of the
transmission may be expressed by:
(22)
Pr
?
Pt )4
17,2 sin ER sin ?
.
2
b sin' x
(24)
With the relations above it is possible to compute the
geometric quantities required for certain idealized beam and a geometric factor, q, generalized in this manner
_
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
E 1216
r A may be defined as
f
tl
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
PROCEEDINGS OF THE IRE
1
9 - (25)
(
? 2 sin ER Sin ?
7 )"
1 7
2
cIf the beam if is swung according to Case
I !analogous generalized geometric factor is:
?
111.1
It I
e
43
+2
1
Psin e (sin I)"
2
MIDPOINT
DIRECTION
OMNI TO BEAM
BEAM TO BEAM
3 4 5
aR DEGREES
October
n=4, 6, and 8. Fig. 46 shows q as a function of aR. In
Fig. 47 q is plotted as a function of 13. For higher values
of it the q curves fall more steeply from their maximum
values, and the position of the maximum for Case I
beam swinging is displaced in the direction of the path
II, the midpoint. In Fig. 47 the polarization correction sin4 x
corresponding to horizontal antenna elements at both
terminals has been included.
(26) The symmetrical properties of the q for Case I above
may be easily established by interchanging T and R
and by swinging the beam to a point displaced by an
equal value of n in the direction of R.
OMM TO BEAM
AND
BEAM TOBEAA1_
Fig. 46?Generalized q factor for Case I geometry for Cedar
Rapids to Sterling distance.
'Li -2
5
A -4-
o
Iaj
-6-
4
?1
(/)
-8-
I)
-ID-
I04
6
$, DEGREES
a
I 0
Fig. 47?Generalized q factor for Case II geometry for "omni to
beam" model for Cedar Rapids to Sterling distance.
; Figs. 46 and 47 illustrate the approximate behavior of
(25) and (26) respectively in the vicinity of the path
;midpoint for the Sterling-Cedar Rapids path for
Fig. 48?Geometry for second beam swinging model?abeam
to beam."
SECOND BEAM SWINGING MODEL?"BEAM TO BEAM"
For this trial consider Fig. 48 which represents Case
I geometry. Both T and R are connected with identical
highly directional antennas of the kind used for re-
ceiving in the previous example. The two beams are
directed toward P. The received power is now expressed
as a function of the position of P as it varies in the plane
represented by Fig. 43. As before the following relation
may be written for P closer to T:
Pe b/T2A2
Pr 0C
12,2M2 sin Er
(a) (b)
sin2 x
AD
(',sin n 1122
2
(c) (d)
(27)
where (a), (b), (c), and (d) have the same meanings as
before. Thus:
Pr
Pt
ADb sin2 x
/n2 sin Er (sin-1)n
2
If P is closer to R, the result corresponding to (28) is:
(28)
Pr
Pi
Apb sin2 x
ir2 sin ER ( n
sin
2
In both cases the denominator contains the larger land
larger e. If // and ei are written for the larger values of 1
and e according to the position of P for Case I geometry,
the generalized geometric factor q may be defined as:
(29)
1
q
I/2 sin et (sin ?7)n
2
(30)
1955 Bailey, Bateman, and Kirby: Radio Transmissio
If the beams are now swung in azimuth corresponding
to Case II geometry, the intersection of one beam with
the scattering shell no longer entirely contains the other
beam and a further factor is required in the generalized
q. Since it is always assumed that the thickness of the
scattering shell is small compared with its height, the
scattering volume can again be written as the area of the
thin shell illuminated multiplied by its thickness. Thus
the volume illuminated by the transmitting beam is
bl'Aft
V= ? ? (31)
sin e
and if Vc is defined as that part of the illuminated vol-
ume within the receiving beam, then the further factor
in the generalized q for Case II geometry is V,/ V, and
j2 sin E
:)n
2
(32)
The factor V,/ V can be approximated as follows. De-
fine 2r as the exterior angle between the two beam di-
rections projected on a tangent plane at P. T may be
obtained from the following relation:
sin ?
a
cos r
1/ 1 - cos2 ? cos2
a
Or
(33)
sing d
cos sin ? ? (34)
sinl a
Then as long as 2rID, which includes the region of
principal interest:
? 1
sin 2r
JT
?ir sin e
?
(35)
it at VHF by Scattering in the Lower Ionosphere 1217
THE BEAMWIDTH OF THE SCATTERING MECHANISM
In general there will be an angular width in azimuth
ct. and an angular width in elevation %A which together '
define the range of directions from which the power'
received mainly arrives. With the aid of the foregoing '
beam swinging models it is possible to examine what
may be termed the natural beamwidth of the scattering ?
medium. The natural horizontal beamwidth associated '
with the scatter mechanism is called cfi. and is defined
by the angular distance between the points three dec-
ibels below the maximum q's for Case II geometry. ?
Table III shows values of (1). obtained from Figs. 47 ?
and 49 for typical values of it. It will be noted that (fo.
A less convenient approximation for IT,/ V which is us-
able for all possible values of r is
V, 2 cot r\ tan T
arc tan (- - arc tan (7??)] . (36)
V 7r sin e sin e
The two arc tangents represent 2nd and 1st quadrant
angles respectively. This expression is exact for beams
of circular cross section in the limit as 6,52-)0. Figs.
P 46 and 47 illustrate the behavior of (30) and (32) re-
spectively for it = 4, 6, and 8 in the vicinity of the path
midpoint for the Cedar Rapids to Sterling path. Fig.
49 includes the polarization correction sin' x corre-
sponding to horizontal antenna elements at both
terminals. It will be seen that the maximum transmis-
sion takes place when the two beams are directed at the
path midpoint. The two beams are then matched ex-
actly in the scattering medium. As before if a greater
value of n is selected the curves fall away more steeply.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
-10
14 68 10
p,DEGREES
? I
Fig. 49?Generalized q factor for Case II geometry for "beam
to beam" model for Cedar Rapids to Sterling distance.
does not vary greatly with extreme variations in the
beamwidth of the transmitting antenna. The values in
Table III are representative for all practical path
lengths. At the greater distances beyond 1,800 kilo-
meters, the q5.'5 are about one degree narrower, and for .4
the shortest distances of practical interest, about 1,000
kilometers, the Om's are about one degree wider.
TABLE III
Om, DEGREES, CEDAR RAPIDS TO STERLING PATH
Di
Omni to beam
Beam to beam
4
6
8
11.3
9.4
8.0
8 . 6
7.5
6 . 8
In the case of the vertical plane, as represented by
Case I geometry, what may be termed the natural fi
beamwidth pr, is defined as the angular distance be-1
tween_points three decibels below the maxima shown in k
Fig. 46. It is .to be noted that the transmitting antenna
plays a significant part in determining the angular g
elevation at the receiver of the upper and lower limits
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
111 1218
r of ifr?,. The case of the omnidirectional antenna to beam
antenna represents an extreme condition exceeding any
11 found in practice. The beam to beam case comes closer
- ? to practice. Table IV shows values of 44,, obtained from
' Fig. 46 for typical values of n.
PROCEEDINGS OF THE IRE
Zit
October
by Fig. 51 approximately by the sum of the dips of
the radio horizons at the two terminals. As a practical
matter, ck will not be limited by common volume
geometry effects even for very long paths as can be seen
TABLE IV
5
DEGREES, CEDAR RAPIDS TO STERLING PATE
y 4
ii
Omni to beam
Beam to beam
,e
E 3
4
5.2
4.2
6
5.2
3.9
8
5.0
3 . 7
.4 2
These results are discussed below in connection with
problems associated with practical antennas.
EFFECTIVE VOLUME FOR SCATTERING
In practical situations, the transmitting antenna will
have considerable directivity. If the transmitting an-
tenna beamwidths, ch. and IP., in azimuth and elevation
4 respectively, are substantially smaller than ck. and V.,??
. (A and 4, will be largely determined by cb. and 1/./.. For
the longer paths, the common volume geometry shown
1 in Fig. 50 will be the limiting factor in determining ;fr.
Ib) PLAN
Fig. 50?Common volume geometry.
In this case 4, is approximately determined by the inter-
section of two planes tangent to the earth at the path
terminals with the scattering stratum in the vertical
plane passing through the terminals and the path mid-
point. IP thus limited is called %frc, and its value is given
as a function of path length and scattering height in
Fig. 51. For very long paths for which elevated antenna
sites must be used 4'c is greater than the values shown
?.
? ? ?
-4--)- )????
I I
? ? 1 I ;
IONOSPHERIC HEIGHT km
1
_?
I
1
90
i
-,76--:-
1
1
I I
1
i
i
' I
! I
I 1
17- I I
I I
I I
I 1
1500
1600
aoo
lO MOO
2000
2100
SURFACE PATH LENGTH. KILOMETERS
2200
2300 2400
Fig. 51?Limiting effect of common volume geometry on maximum
vertical angle of arrival or departure for Case I geometry for
antennas at zero height, including effects of representative mid-
latitude refraction.
from Fig. 52 and consideration of the influence of path
length on IA,, as discussed earlier. The cone widths 43 and
IP as influenced by Om and cf),,, and IP., IP., and ific, re-
spectively, determine in a rough way what may be
termed an effective volume in the height range over
which scattering takes place.
100
90
80
.c
70
60
1
I
1 1
i
!
, '
I
I 1
::, Da Is d
1 , 1
tang.nry al eartn's
I II
I I
I ts="4.241,11,1,11.1.14
turf K.
' 1
r----
t- tt?
1 .
i
? '
I I
'
?
I
?"- ?
I.I
. i 1
n?
1
'
t ?
1
1
I i
1
10 0 10 20
30 40 50 60
AZIMUTH, DEGREES
Fig. 52?Limiting effect of common volume geometry on maximum
azimuthal angle of arrival or departure for Case II geometry,
including effects of representative mid-latitude refraction.
It is useful to consider the influence of the effective
volume in the design of antennas for use in transmis-
sion by ionospheric scattering. When the transmitting
and receiving beamwidths both vertically and hori-
zontally are substantially larger than, and include, the
cone of -angles associated with the effective volume, the
received power will increase as the directivity of the
antennas is increased in approximate proportion to the
product of the plane-wave gains of the antennas. Under
these circumstances increased transmitting directivity
produces more intense illumination of the effective
volume and, in reception, power from the effective
1955 Bailey, Bateman, and Kirby: Radio Transmission at VHF by Scattering in the Lower Ionosphere 1219 '1
volume is received with a larger aperture resulting in a
further increase of received power. If, however, the
antenna beamwidths are smaller than the cone of
angles determined by chm, and IP. or Is% whichever is
smaller, the received power will increase more slowly
with increasing antenna directivities. For this condi-
tion, the power received from the central portion of the
effective volume, resulting from the increased intensity
of illumination in the maximum of the transmitting
antenna beam, and the more effective scattering in the
vicinity of the path midpoint, more than offsets the loss
of received power resulting from the reduction of the
effective volume imposed by the smaller beamwidths.
When the antenna beamwidths are approximately equal
to Om, and IP. or IP, whichever is smaller, the effective
volume will just fill the beams and the antenna beam-
widths may be said to match the effective volume. The
plane-wave gains of transmitting and receiving antennas
should in these circumstances be largely realized.
REALIZED GAINS OF THE ANTENNAS
From early in the experimental program it has been
apparent that antenna characteristics have an impor-
tant influence on the received signal levels and on the
diurnal variations of these levels. Fig. 53 shows the
computed directivity of the transmitting and receiving
Ty-z-r7-rrY-T tiTTIT rtrrit 111
35 [
1-
30F
(HORIZONTALLY POLARIZED COMPONENT)
(0
25
cr
ID 1.?
ILI 1?
o
1
-20
E
It 15-- Coat opts opts are numb ere
in decibels relative
- to maximum field
10
5
,
?
Iii
-25 -20 -15 -10 -5 0 5 10 15 20 25
AZIMUTH, DEGREES
Fig. 53?Computed directivity of rhombic antennas used for the
Cedar Rapids to Sterling path.
rhombics used for the routine signal intensity recording
program on the Cedar Rapids?Sterling path. When an-
tennas of lower directivity are used at Sterling for re-
ceiving it is found that the average signal intensities
available from these antennas are generally lower than
those available from the rhombic antenna but that the
ratios of the powers received on the rhombic antenna to
those received on the less directive antennas vary char-
acteristically with local time. These effects are illus-
trated in Fig. 54 which displays the received intensities
at 49.8 Mc/s on the rhombic antenna and on a five-
element horizontal Yagi antenna. The free-space, half-
power beamwidths of the Yagi antenna were approxi-
mately 65 degrees in the plane of the elements and 52
degrees in the plane normal to the elements. The plane-
wave gains of the rhombic and the Yagi antennas in the
beam maxima are approximately 18 db and 9 db re-
spectively, relative to a horizontal half-wave dipole at
the same height above ground. Comparisons of received
28
024
LU
20
IC
216
400
02 09 06 08 10 12 14 16 18 20 22 00
751. W TIME
Fig. 54?Median signal intensities observed at 49.8 Mc/s at Sterling
with rhombic and Yagi receiving antennas for a 19-day period in
December, 1951. Rhombic transmitting antenna used at Cedar
Rapids.
signal intensities using rhombic transmitting and re-
ceiving antennas and five-element Yagi transmitting
and receiving antennas were made on the Anchorage to
Barrow path at 48.87 Mc/s in March, 1953. Results of
experiment are in Fig. 55. Fig. 56, page 1220, shows
18r
16t
2
14L
1:110MBC ANTENNAS
BEAMWIDTH ID 6?
o
CO 02 04 06 08 10 12 14 16 18 20 22 00
150 W TIME
Fig. 55?Median signal intensities observed during March, 1953
at 48.87 Mc/s at Barrow with rhombic to rhombic and Yagi to
Yagi antenna arrangements.
distributions of realized gain for rhombic-to-rhombic
antenna systems used in the routine recording pro-
?
gram as compared with dipole to dipole transmission.
The realized gain of an antenna is defined here as 1
the ratio of the signal intensity observed -when the
antenna is used to that observed when a reference or
comparison antenna is used during the same period of
time. In stating results of observations of realized gain
it is necessary to specify the antenna arrangements used
at both terminals. Measurements of realized gain were
made on all of the 50 Mc/s test paths by means of alter-
nate half-hourly transmissions on a rhombic antenna
and on a reference dipole, with reception performed
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
1220
;
1.! simultaneously with a rhombic antenna and a dipole.
Table V summarizes the results for the Fargo to
I I
0 I Churchill path for a three-day period. It is evident that
ki A
TABLE V
REALIZED GAIN, DECIBELS
PROCEEDINGS OF THE IRE
IPer Cent of
Comparisons
in Which
Gain Equals
I or Exceeds
Value Shown
Transmitting Rhombic
Antenna
Receiving Rhombic
Antenna
Measured
with
Receiving
Rhombic
Measured
with
Receiving
Dipole
Measured
While
Transmitting
on Rhombic
Measured
While
Transmitting
on Dipole
5
10
SO
90
95
19
18
14
10
9
14
13
9
4
2
19
18
15
11
10
17
15
10
3
1
the realized gain of a directive antenna is a function
of the directivity of the antenna used at the opposite
1:1 terminal and is influenced strongly by diurnal variations
in the propagation mechanism or mechanisms. It is con-
venient to discuss this behavior by considering the case
i of a receiving antenna. The controlling factor in deter-
; mining the realized gain of a receiving antenna is the
angular size of the cone of angles from which power
I mainly arrives at the receiving location in relation to
- , the angular dimensions of the receiving beam. The ob-
served greater realized gain of the rhombic receiving
; antenna shown in Fig. 54 during the daylight hours sug-
"ri gests that most of the received power is arriving from a
cone of angles determined largely by the directivity of
:1 the transmitting antenna or by a highly directive scat-
tering mechanism. On the other hand, the behavior at
other times when the realized gain is low suggests that
I; a larger cone of angles exists from which power is ef-
fectively receivable. Table V further illustrates these
effects.
4
3
3
2
2
12
8
4
*---1
--1-14-1-1-1--1
I
I FARGO -CHURCHILL
1 1 _ 1 I I
1
I
r T
CEDAR RAPIDS - STERUNG -
I I I I
ANCHORAGE - BARROW
I
414
r
I
4,
1
I
...
90 95 98 99
PERCENT CFnuc GAIN EQUALS OR EXCEEDS ORDINATE
Fig. 56?Cumulative distributions of realized gain at about 50 Mc/s
of rhombic to rhombic antenna arrangement over dipole to dipole
arrangement, observed for 3-day period for each path.
Factors which can give rise to an increased angular
size of the cone during these periods are decreases in the
angle dependence of the scattering mechanism, large
scale irregularities in the efficiency of scattering over
the extent of the common volume, and contributions to
. .
- -
October
from meteor ionization which lies outside the effective
volume as determined by Ons, 4?8, tp., 0., and IP,. The
first of these factors, in view of the frequency depend-
ence measurements, and the calculated values of 0.
and IPm appears to be incapable of producing the ob-
served wide variations in realized gain. Little detailed
information is available concerning the nature and be-
havior of the large scale irregularities in scattering
efficiency throughout the common volume, but there is
hardly any doubt of their existence. Influence of meteoric
ionization is probably most important single factor in
causing observed variations in realized gain.
INFLUENCE OF METEORIC IONIZATION
Much of the preceding discussion has been based on
the assumption that a scattering mechanism is effective
in propagating the received signals. To the extent that
the received signals are appreciably contaminated by
the presence of components resulting from meteoric
ionization, the principal influence on the behavior of
highly directive antennas will be a reduction in their
realized gains as determined by the larger cone of angles
associated with the meteoric components. If, however,
the contributions from meteoric ionization are domi-
nant, greater signal intensities could be received if differ-
ent antennas were used.11.12
In an earlier paper the' possibility was suggested that
the diurnal variation of observed signal intensity could
be interpreted as the resultant of a solar ultra-violet
influence having a maximum at midday, and a meteroic
influence having a maximum at about 06 hours and a
minimum at about 18 hours local time. To illustrate
this suggestion the hourly median signal intensities ob-
served at 49.8 Mc/s at Sterling were shown for April,
1951. Observations of the rate of occurrence of Doppler
components differing from the carrier frequency by at
least 200 cycles per second have been made at Sterling
at 49.8 Mc/s over a long period. The resulting average
rates for the month of April, 1955 are shown in Fig. 57
14
12
3 10
2
8
36
4
2
00 02 04 06 08 10 12 14 16 18
LOCAL TIME
Fig. 57?Diurnal variation of the average rate of occurrence of
meteor Doppler components observed at Sterling at 49.8 Mc/s
during April, 1955.
for Doppler components having intensities exceeding
0.25 microvolt, open circuit voltage at 50 ohms. The
maximum and minimum rates are seen to occur one to
two hours later than 06 hours and 18 hours local time
at the path midpoint. The secondary maximum at
20 22 OD
4
1955 Bailey, Bateman, and Kirby: Radio Transmission at VHF by Scattering in tne Lower lonospherc
The principal results of an experiment designed to
establish more clearly the relationship between con-
tributions resulting from a scattering mechanism and
those resulting from meteoric components are now de-
scribed. In this experiment, 49.8 Mc/s transmissions
beamed towards Sterling, were simultaneously received
at Sterling and at a site selected as being nearly opti-
mum for reception of the meteoric components. This
:EDAR
RAPIDS
621 51,n MIDPOINT
621 5km STERLING
6' 40' 0 1230 2
13. 2I5 ST476
CARYSBROOK
Fig. 58?Geometry for Carysbrook experiment.
144 3km
site was near Carysbrook, Va., at a distance of 144.3
kilometers from Sterling in a southerly direction as
shown in Fig. 58. At Sterling, the rhombic antenna
normally used in the recording program was employed.
An identical rhombic antenna was erected at Carys-
brook, and directed toward the midpoint of the Cedar
Rapids to Sterling path. Fig. 59 displays the hourly
median signal intensities received simultaneously at
Sterling and at Carysbrook during the course of the
experiment. The integrated intensities were observed
to be higher at Carysbrook between 23 hours and 04
hours local time at the path midpoint. During these
hours the results would thus appear to be consistent
with the theory that the meteor ionization trails con-
tributing most effectively to the integrated signal in-
tensity occur on one side of the great circle path be-
tween the transmitter and the receiver." On the other
hand, the existence of greater signal intensities on the
Cedar Rapids to Sterling path most of the time pro-
vides further support of the view that the observed
signals are the result of more than one propagation
mechanism." For the conditions under which the ex-
periment was performed, it is concluded that systems
designed to work through the universally observed
periods of low signal intensity occurring at about 20
hours local time should employ antennas directed along
the great circle bearing between the transmitter and
receiver. It should be pointed out that the rhombic an-
tennas used on the Cedar Rapids to Sterling path since
the beginning of the experimental program were de-
signed for an assumed midpoint height of about 105
n It is of interest to calculate the expected intensity of the
Carysbrook signals relative to the simultaneously observed signal
intensities at Sterling on the assumption that propagation is pre-
dominantly by a scattering mechanism, as might be expected during
the afternoon. Previously given geometric considerations are em-
ployed and for angle dependence it is assumed that n =6. The loss
resulting from the larger scattering angle at Carysbrook is found to
be about 4 decibels. In addition a further loss of 1.i to 2 decibels, af-
fecting the Carysbrook observations, is associated with Vc/ V as
given by (36). Finally there is a loss of about decibel associated
with the use of horizontal polarization, because x=90? at Carys-
brook. A net loss at Carysbrook relative to Sterling of about 6
decibels is therefore predicted for propagation by scattering un-
contaminated by meteoric or other components. This result is seen
to be in fair a eement with Fig. 59, during the midafternoon period.
kilometers for the scattering region. In the light of the
knowledge acquired later concerning the heights at
which scattering occurs, it would have been more ap-
propriate to direct the beam maxima toward a midpoint ,
height of about 85 kilometers. Consequently, it is
thought that the experimental results throughout the
observing program at Sterling including the results pre-
sented in Fig. 59 are biased to some extent in favor of
the meteoric components.
Further experimentation of the kind above is desir-
able in order to obtain information on the relative con-
tributions of the scattered components and the meteoric
components at other times of year and for other path
orientations. Some specific suggestions for related ex-
perimentation were made earlier in connection with
polarization dependence.
17 JUNE - 14 JULY 1955
1215
10
STERLING
CARTS8ROOK
5 TRANSMISSIONS FROM CEDAR RAPIDS
- DIRECTED TOWARD STERUNG
0
OD 02 04 06 08 10 12 14 16 18 20 22 00
LOCAL TIME AT PATH MIDPOINT
Fig. 59?Diurnal variation of hourly median signal intensities ob-
served simultaneously at Sterling and at Carysbrook, at 49.8 I
Mc/s.
DIVERSITY CONSIDERATIONS
In an earlier section the measurements of the correla-
tion between the envelopes of signals received on spaced
Yagi antennas were described. Fig. 20 illustrates the
observed behavior for spacings along and transverse to
the path. Envelope correlation coefficients of 0.5 were
observed for spacings of about 3.5 wavelengths trans-
verse to the path and about 40 wavelengths along the -
path. A practical implication of these results is that
smaller antenna spacings will suffice for effective di- Ll
versity action if the diversity antennas are disposed in :1
a line normal to the path rather than along the path.
Gordon" has studied the effects of the tropospheric
scattering mechanism in relation to the correlation ex-
pected between spaced antennas for diversity action.
Using somewhat similar, but not identical arguments
diversity spacings for transmission by ionospheric 11
scattering can be estimated. While the precise calcula-
tions would be involved, important features can be seen 1,
from some very simplified considerations. As such con-
sider that the antenna spacing required to produce a ?
low envelope correlation is that necessary to give a dif-
ference of one wavelength in the paths of the signals .1
from opposite boundaries of the effective scattering
" W. E. Gordon, "Radio scattering in the troposphere," PRoc.
IRE, vol. 43, pp. 23-28; January, 1955.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
1222
PROCEEDINGS OF THE IRE
volume. Fig. 60 illustrates the usual diversity arrange-
ment in which diversity action is obtained by spacing
? the receiving antennas transverse to the great circle
path at a constant height. The angle 4) in Fig. 60 is the
width of the cone of angles over which most of the power
? is received. The diversity distance Sg is defined to be
the distance through which a receiving antenna must be
' moved to produce a change in the difference between
the ray path lengths l and /2 of one wavelength. The
ray lengths, /1 and 41, and /2 and 12', are much greater
? than SI so that they may be considered to be parallel.
Fig. 60?Geometry for transverse spaced-antenna diversity.
From simple geometrical considerations,
. ct.
ii'? /1 = Se sin
2
; (1)
121 12 =
2
and the diversity distance is:
X
2 sin ?
2
(37)
(38)
(39)
The smaller of the cone widths, 0. and 0. is used with
(39) for estimating appropriate spacing of the antennas
for diversity action.
The diversity distance in a vertical direction at the
-I receiver is, on the basis of similar considerations,"
?
X
St, ? (40)
2 sin ?
2
iThe smallest of the cone angles, IP., and 1pc, is used
for IP in computing an appropriate diversity distance S.
The longitudinal diversity distance Si along the path
rand parallel to the tangent plane to the earth under the
Ireceiver is influenced by both cone angles 0 and t(i. It
i is convenient to state the results separately. For 0,
and for IP,
34 In these equations certain cosines have been set equal to unity
because of the smallness of the angles.
X
Si(0) = (41)
2 sin2
4
Sg(1,&) =
2 sin ? sin 0
2
October
(42)
where 0 is the angle between the tangent plane at the
receiver and the center of the cone width IP. When the
combined effects of 0 and IP are taken into account, the
appropriate diversity spacings will be intermediate
between the values determined separately. This matter
has not been further explored since even the smaller
of the diversity distances obtained from the above ex-
pressions will, in practice, be many times the diversity
distance Sg required for spacings transverse to the path.
For this reason longitudinal diversity is not convenient
in practice.
As will be discussed in Part III, a useful siting pro-
cedure is to align the transmitting and receiving beams
on the ionospheric midpoint height by putting the
antennas at a height z above a smooth reflecting surface,
as determined by,
X
5= g
4 sin a
(43)
where a is the angle of elevation of the path midpoint
for the design height of scattering. For the shorter paths
in which the antennas are sited in accordance with (43),
a will be roughly the same size as IP. With this in mind,
and comparing (43) with (40) it will be seen that for
small a and tp, St, will be roughly twice as great as 2z,
the distance between the antenna and its image. Ac-
cordingly, the correlation between the antenna and its
image will be rather poor, and coherent gain will be
realized only for the contributions arriving from the
central portion of the effective volume. The net increase
in signal power resulting from the presence of the image
receiving antenna when both transmitting and receiving
antennas are thus sited is expected to be about three
decibels or slightly greater. An alternative explanation
is that the- receiving antenna will be receiving power
from two poorly correlated, sources of approximately
equal intensity, the effective volume and its image.
The transverse diversity distance Sg for the Cedar
Rapids to Sterling path is 9.6X on the assumption that
=0.= 6 degrees. Experimentally observed values of
envelope correlation as a function of antenna spacing
were given in Fig. 20; extrapolation indicates that the
correlation is very small for antennas spaced at the
computed distance of 9.6X.
Height-gain observations made early in the 49.8
Mc/s recording program at Sterling showed that the
signal intensities as received on a Yagi antenna were
not very sensitive to height over a range of heights
from about 35 to about 80 feet indicating poor correla-
tion between the Yagi antenna and its image in the
ground. The diversity distance S? for the Cedar Rapids
to Sterling path is 11X as computed for IP =ik?, = 5.2
degrees. In the distance-dependence tests, height-gain
1
1955 Bailey, Bateman, and Kirby: Radto Transmission at VHF by Scattering in the Lower Ionosphere
comparisons at Homestead given in Table II for an-
tennas at heights of 40 and 100 feet showed that the
daytime signal intensities observed at a height of 100
feet were greater than those at 40 feet by 8i decibels.
For the path length 2,088 kilometers between Cedar
Rapids and Homestead, tfrc will be the limiting angle in
determining cone width By using values for Vic from
Fig. Si, for an 80 kilometer height, and making ap-
proximate allowances for the effects of the heights of
the transmitting and receiving antennas, a cone width
iPc of about 1.1 degrees is determined for this path, cor-
responding to a diversity distance St, of about 1,030
feet. The spacing of 200 feet between the antenna and
its image for the upper antenna height is substantially
less than the computed diversity distance for this path.
At night, however, the observed height gain between
the 40 and 100-foot heights was 6i db indicating a
somewhat larger cone width tpc. This behavior might be
anticipated on the basis of the higher observed heights
during the night hours and the contributions from
meteoric ionization which occur at the greater heights.
Similar observations made at shorter distances during
the Florida distance dependence tests showed lower
height gains than those observed at Homestead, indi-
cating greater effective values of IP,. The observed be-
havior in the experiments discussed above is thus seen
to be in good qualitative agreement with the behavior
predicted from diversity considerations. ?
THE SIGNIFICANCE OF THE FREQUENCY AND
ANGLE-DEPENDENCE RESULTS
In the light of the preceding discussion on the role of
the antennas it is of interest to consider the extent to
which the frequency and angle dependence measure-
ments may be regarded as providing a basis for judging
models of the scattering medium having frequency and
angle dependence of the kinds discussed earlier. From
the theory of Villars and Weisskopf, or from more ele-
mentary physical considerations, it is expected that
the angle-dependence as expressed by an exponent of
1/sin 7/2 will vary with frequency and with conditions
in the medium. The medium will scatter with a sharper
polar diagram as the frequency is raised confining the
downcoming waves at the receiving point to a smaller
effective cone. Correspondingly the cone angle at the
receiver will decrease with an increase in the effective
frequency exponent at a fixed frequency. Thus, meas-
urements of frequency dependence using scaled anten-
nas will not accurately represent the frequency de-
pendence characteristics of the scattering mechanism
unless the -antenna beamwidths are small relative to
the beamwidths associated with the scattering mecha-
nism. The beamwidths of the antennas used for the
measurements were, in fact, comparable to the natural
beamwidths of the scattering mechanism shown for the
second beam swinging model and it would be expected
that variations in the characteristics of the mechanism
with frequency or with time at a given frequency would
1223
influence the measured values of n by no more than
plus or minus 0.5. A more serious source of difficulty in
interpreting the observations results from the existence
of meteoric components having varying intensities
relative to the scattered signal. In this connection, the
relative behavior of the signal intensities at Sterling
and at Carysbrook, suggests that for several hours dur-
ing the afternoon period, the contributions from me-
teoric ionization are small in comparison with the
scattered components. The frequency dependence ob-
servations made during this portion of the day are
therefore thought to provide a fairly reliable estimate
of the frequency dependence of the scattering mecha-
nism. Correspondingly, the observations made during
the early morning hours particularly during the winter
months can probably be used to provide a reasonably
good estimate of an effective frequency dependence of
the integrated signal intensities produced by reflections
from meteoric ionization for the experimental arrange-
ments. Considering the various uncertainties including
experimental errors associated with the observations
and the interpretation of the observations it is thought
that the effective exponent n representing frequency
dependence of the scattering mechanism is being meas-
ured to within about twenty per cent when selected data
are used.
The interpretation of the angle-dependence observa-
tions is similarly beset with difficulties. First, a rela-
tively small systematic error is introduced by the finite
beamwidths of the antennas since the scattering angle
increases on either side of the midpoint in the scatter-
ing region. Thus the effective would be expected to be ,
somewhat greater than the midpoint value. Secondly,
because of experimental difficulties associated with the .
measurements of pulse signal intensities over the shorter
paths used for this experiment, it was not possible to
make measurements except during periods of very high!
signal intensity and consequently only a limited quan-
tity of data was obtained. On the other hand, the data
were taken during periods considered to be relatively
free from contamination by reflections from meteoric
ionization. On the basis of the preceding discussion it is
thought that the exponent n associated with the angle!
dependence of the scattering mechanism at 50 Mc/s was I
not greater than nine nor less than six, for the particu-
lar periods during which the measurements were made.
PART III. DESIGN CONSIDERATIONS FOR
COMMUNICATION APPLICATIONS
INTRODUCTION
Sufficient experience has been acquired with the prop-
agation mechanism under study to permit an evaluation
of some of the communications possibilities. Expe-
rience in the Arctic and middle latitudes has provided
evidence of the utility of the propagation mechanism
for communication purposes. With systems gains com-
parable to those used on the test paths, or somewhati
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
1224 'PROCEEDINGS OF THE IRE
/ greater, and for somewhat lower frequencies, say from
i 30 to 40 Mc/s, multichannel radio-teletype 1 opera-
1 1 1 i tion can be realized with a very high degree of reliability.
/ Under these circumstances, radiotelephone communica-
tions and transmission of facsimile are also feasible.
i In the Arctic the known characteristics with respect to
.1 magnetic disturbances are of particular value.
, I
i The most useful range of path lengths is from about
1,1 1,000 to 2,000 kilometers. Recent experience with the
t
test path between Newfoundland and the Azores has
indicated that, subject to the availability of high sites,
11, overlooking the sea for example, path lengths as great
as 2,300 kilometers may be used successfully. As dis-
cussed earlier in this paper, received signal intensity de-
creases rapidly with decreasing distance below 1,000
' kilometers and for this reason path lengths shorter than
about 1,000 kilometers are to be avoided.
!i
USEFUL PATH LENGTHS
USEFUL RANGE OF FREQUENCIES
, The range of frequencies of greatest usefulness is
from 25 to 60 Mc/s. Frequencies lower than 25 Mc/s,
wherever used, will be subject to and will cause inter-
! s
ference to a much greater degree as a result of normal
? ; and sporadic ionospheric propagation. The interference
I will be particularly serious in years of high solar activ-
ity, and will be influenced by diurnal, seasonal and
?. geographical factors. Back scatter will be experienced,
; particularly in middle and low latitudes and in years of
1 high solar activity.
, In general, the lower frequencies in the range are
less susceptible in high latitudes to the various diffi-
culties described above. The lower frequencies in the
range of 25 to 60 Mc/s have other inherent natural
i
i advantages. For fixed transmitter power and antenna
11; directivities, use of the lower frequencies will result
1: in greater signal-to-noise ratios, less interference from
; meteoric effects, and slower fading. Effects of SID's
1 and polar blackouts, although more pronounced on the
I lower frequencies, will not affect the reliability of com-
munications circuits operating in the lower portion of
i this range.
It should not be concluded that there is little use for
. ?
frequencies as high as 50 or 60 .Mc/s. In certain seasons,
1 these higher frequencies may be of considerable value.
If potential interference from regular-layer propagation
and back scatter is to be practically eliminated, the
higher frequencies must be used for applications re-
quiring extreme reliability, even at the expense of
higher transmitter power.
BANDWIDTHS AND CHANNEL SEPARATIONS
tolIt seems unlikely for some time to come that sys-
tems will require bands greater than about 50 kc/s.
; In many instances substantially smaller bands will
suffice. Because of problems associated with slopes of
October
the pass bands in receiving equipment, which may arise
during occurrence of very strong signals, it is probably
advisable to separate channels about 2 to 3 times the
width of the nominally occupied bands. It is also nec-
essary to allow for the effects of sporadic-E propaga-
tion, transient meteoric enhancements, and regular
ionospheric propagation.
TRANSMITTERS AND RECEIVERS
In order to take full advantage of the propagational
reliability of the mechanism it is important that trans-
mitters of sufficiently high power be used and that the
equipment employed at the path terminals be designed
to provide a very high degree of reliability.
The transmitters and receivers should be capable of
operating in the frequency range 25 to 60 Mc/s. For
most applications there is no need to incorporate pro-
visions for continuous"front-panel" frequency chang-
ing as large changes in frequency will almost always re-
quire careful tuning of new antenna arrangements and
need only be provided for within the equipment itself.
For some applications, involving low information rates,
such as single-channel radio-teletype transmission,
transmitter powers of the order of five kilowatts are
likely to be found adequate at frequencies of the order of
35 Mc/s, particularly in the Arctic.
In the frequency range 25 to 60 Mc/s external noise
?will be the limiting factor in communications provided
that the receivers have sufficiently low noise figures. A
receiver noise figure of two can easily be realized and is
adequate.
MULTIPATH LIMITATIONS
Various types of multipath propagation have been
observed in the experimental program. The most com-
monly observed multipath effect is produced by re-
flections from meteoric ionization in the ionospheric
volume common to the transmitter and the receiver.
Values of maximum .multipath delays and correspond-
ing azimuths have been computed from the common
volume geometry illustrated in Fig. 50 for the longest
and shortest geometrically possible ray paths and are
plotted in Figs. 61 and 62. Multipath delays as great as
the values derived from Fig. 61 for the shorter paths
will be extremely rare since the probability of occur-
rence of meteor ionization becomes vanishingly small at
the points where the tangent planes intersect at me-
teoric heights. Furthermore, the intensity and duration
of received signals will be very low for reflections within
the common volume but near the intersections of the
tangent planes as a consequence of increased inverse
distance attenuation, the decreased obliquity,24 and the
additional directivity discrimination at low angles of
elevation resulting from interference between the di-
rect and ground-reflected waves. Observations of such
multipath delays were made using pulsed transmission
on the 811 kilometer experimental path between Ster-
ling, Va. and Bluffton, S. C. The results summarized
s,
?
4
1955 Bailey, Batenzan, and Kirby: Radio Transmission at VHF by Scattering in the Lower Ionosphere
MAXIMUM METEORIC MULTIPATH DELAY. MILLISECONDS
6
5
4
3
2
nm ? i00 km
No ? 85km
REFRACTION INCLUDED
o
600 800 1000 1200 MOO 1600 1800 2000 2200 2400
SURFACE PATH LENGTH KILOMETERS
Fig. 61?Maximum meteoric multipath delays for antennas
at zero height.
in Fig 63, were obtained from range-time oscillograms
of the type shown in Fig. 15.
A second type of multipath effect, associated with
sputter, is illustrated in Fig. 35. This type of multipath,
characterized by rapid fading is of practical importance
as the associated delays may be as much as several
milliseconds. It occurs most frequently for paths cross-
ing or near the zone of maximum auroral occurrence,
where it may be expected to cause difficulty at times
with antennas having typical directivity characteristics.
These effects are illustrated by Figs. 35 and 36. Fig.
50 illustrates the common volume geometry associated
with this type of multipath as well.
Other possible sources of multipath include off-path
reflections from banks of sporadic-E ionization and
from ground-scattered energy propagated by F2 or
region reflection reflection for several different types of multipath
geometry. The delays associated with multipath arising
from back scatter may be as great as 50 milliseconds or
more and are potentially capable of seriously limiting
the transmission speed. The harmful effects of back
scatter may be reduced or eliminated by using antennas
designed for suppression of the back lobes.
For transmission by ionospheric scattering, multi-
path from strong offpath reflections from meteoric
ionization is, during its existence, a factor reducing the
available radio path bandwidth. The antennas which
have been used in the experimental program do not
have ;sufficiently low radiation in the minor lobes, as
compared with the main beam, to provide adequate
suppression of the off-path meteoric components. Con-
sidering the preceding discussion in conjunction with
the experimental observations, it seems reasonable to
expect that multipath components having delays ex-
ceeding half a millisecond will occasionally occur with
sufficient intensity to introduce transmission errors on
the shorter paths. However, the available radio path
bandwidth, as estimated from the reciprocal of the de-
lay, for these shorter paths is greater than 2 kc/s for a
large percentage of the time. For the intermediate and
longer path lengths the available bandwidths will be
1225
70-
GO-
0
30-
20-
600 860 1000 1200 i 400 1600 1800 2000 2200. 2400 1
2
1
Fig. 62?Azimuth of meteor giving maximum multipath delay
for antennas at zero height.
SURFACE PATH LENGTH, KILOMETERS
180
160
C, 140
w 120
co
KO
80
u.
6 6?
40
2
to) Observing
Range
0.25 To 5 Maliseconds
22 JANUARY 1953
0745-0845
0845-0945 75? V/ TIME
0945-1045
?
Ib)
Observing
005
2744NUARY
0845-0900
0945-1000
Range
To 06Milliseconds
1953 -
75'W TimE
_
_
0.0 05 lb IS 2.0 2.5 00 01 02
DELAY, MILLISECONDS
Fig. 63?Distribution of observed meteoric multipath delays obi
served at Bluffton at 49.7 Mc/s for a transmission path of 811 knl
from Sterling.
somewhat greater but still largely determined by thi.j
minor lobes of the antennas and the reflecting proper4
ties of the ground. 1
With antennas of improved directivity, multipatli
4
1
associated with meteoric ionization will no longer be ;?
,
problem and the radio path bandwidth will be limitet-
by the antenna beamwidths or the natural beamwidtl,
associated with the scattering process. While the sub i
ject of wide-band transmission at vhf by ionospherii
scattering has not been extensively analyzed, it is o
interest to estimate the inherent bandwidth limitatio)
imposed by the scattering mechanism. Values of .
generalized q computed for the Cedar Rapids to Ster
ling path are presented in Figs. 46 and 47. Multipatl
delays haye been derived from the ray paths associate(
with the ionospheric midpoint and positions in th
common volume corresponding to the half power value
of q, assuming n =6 and the use of an omnidirectione
transmitting antenna. With these idealizations, th
maximum delay is 21 microseconds and is associate
with IP.. The available path bandwidth for the assume
conditions is 48 kc/s. For practical situations, the effect
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
,1.226
PROCEEDINGS OF THE IRE
of ground reflection will usually be such that the delays
associated with 4, will be greater than those associated
with IP. Assuming sharply beamed antennas 4,,Th is
limiting and the delay is 9 microseconds. This corre-
sponds to an available path bandwidth of 110 kc/s. If
the transmitting and receiving antenna beamwidths
are smaller than bandwidths of the scattering mecha-
nism, still greater bandwidths are in principle available.
ANTENNAS
With the results of the experimental program (Part
I) and the discussion of the role of the antennas (Part
III), together with the multipath considerations given
above, it is possible to establish the characteristics of
antenna systems for use in communication applications.
For purposes of discussion, Table VI indicates certain
desirable characteristics without regard to whether
antennas having such characteristics are at present
practicable.
TABLE VI
DESIRABLE ANTENNA CHARACTERISTICS
Horizontal beamwidth
8 degrees or less
Vertical beamwidth in-
cluding ground effects
6 degrees or less
Maximum radiation in
any minor lobe
At least 40 decibels below radiation in
the maximum of the main beam
Radiation efficiency; i.e.,
ratio of total power radi-
ated to antenna input
power
90 per cent or greater
Horizontal orientation of
beam .
Normally on great-circle bearing be-
tween transmitter and receiver
Vertical orientation of
beam
85 kilometers above midpoint of great-
circle path
Provision for reducing di-
rectivity or for varying
direction of the beam
For use during periods when scattering
is not homogeneous and during periods
when scattering from meteoric compo-
nents will provide higher signal-to-noise
ratios
Bandwidth over which
characteristics are to be
maintained
200 kc/s
Although most of the above set of characteristics are
self-explanatory some additional comments are indi-
cated. When a low noise-figure receiver is employed,
somewhat lower radiation efficiencies for receiving
antennas can be tolerated since the available noise
power from galactic sources is much higher than that
,resulting from internal noise generated in the receiver.
For the longer paths, increased vertical directivity is
necessary in order to utilize effectively the reduced
common scattering volumes. From the common volume
point of view, horizontal beami.vidths smaller than 8
degrees will not be required for any except the very
longest paths. The suggested techniques for varying
directivity or for orienting the main beam in directions
:other than the great Circle-bearing have not been tested
October
under operational conditions and a determination of
the practical effectiveness of this scheme will require
further study and trial. It remains to be seen to what
extent antennas having characteristics approaching
those envi?aged above will find their way into practice.
The use of spaced-antenna receiving diversity is defi-
nitely desirable for good communications. From the
spaced-antenna observations of envelope correlation
between spaced antennas, and from the diversity-dis-
tance considerations discussed in Part II, it is con-
cluded that for effective diversity action the component
of spacing transverse to the path should not be less
than about 10 wavelengths.
The effects of the scattering mechanism with regard
to the polarization of the transmitted waves have been
considered earlier and it was shown that the scattering
losses are somewhat less for horizontal than for vertical
polarization. Horizontal polarization is generally to be
preferred for additional reasons associated with the
reflection characteristics of the ground."
PRACTICAL ANTENNA SITING
A corollary to the observed failure to realize, for a
large fraction of the time, antenna gains comparable
to the plane-wave gains, is that the additional gain re-
sulting from ground reflection will not be fully realized,
even though ideal sites are employed at the transmitter
and receiver. The reason for this was discussed in Part
II in connection with vertical diversity spacing. Never-
theless there are demonstrable advantages in siting the
antennas with respect to a ground surface so that the
plane-wave ground-reflection lobe patterns are well
formed. In fact, if either antenna is poorly sited, the
beams may only partially intersect in the height region
where scattering occurs.
For the idealized beam-swinging models discussed
in Part II two limiting cases were considered with re-
spect to the transmitting antenna patterns. They were
an omnidirectional antenna and an extremely directive
antenna. The generalized q-curves representing Case I
geometry, in which the effective scattering volume is
displaced toward one terminal, are shown in Fig. 46 as a
function of ?R. In practice the directivity of the trans-
mitting antenna is intermediate between the two ex-
tremes. It is therefore anticipated that slightly greater
signal intensities will result if the transmitting and re-
ceiving beams are directed toward a point in the scatter-
ing stratum slightly displaced in the manner above pro-
vided appropriately different directivities are employed
at the terminals. While unsymmetrical operation of this
kind, employing negligibly increased scattering angles,
would represent no disadvantage, it is nevertheless
recommended in practice that antenna beams be de-
signed to have their principal lobes directed toward the
path midpoint in the ionosphere. The recommended
ionospheric height for antenna design and corresponding
site selection is 85 kilometers.
1
1955 Bailey, Bateman, and Kirby: Radio Transmission at VHF by Scattering in the Lower Ionosphere
Some exceptions to this design recommendation can
be made with advantage for paths longer than about
2,200 kilometers. For such extreme path lengths some
increase in received signal intensity is likely to be
realized with antennas having practical vertical free-
space directivities if use is made of heights greater than
required to direct the first ground-reflection lobe at the
path midpoint in the ionosphere. As the antenna heights
increase, tfr, increases with the increasing depression of
the radio horizons. As a consequence the total power
radiated into the common volume is increased. Thus, for
very long paths greater heights should be used wherever
practicable and especially at one terminal if height
limitations exist at the other terminal. The design
heights indicated for lobe alignment at the path mid-
point in the ionosphere should be regarded for paths of
extreme length as minimum rather than optimum.
The curves of Fig. 11 show the vertical angle of ar-
rival or departure for various ionospheric midpoint
heights as a function of the surface distance between the
transmitter and the receiver. They have been found
useful in connection with antenna design and siting
problems. The values given by these curves have been
calculated for elevations near sea level for the assumed
condition of no lower atmosphere and for representative
radio-wave refraction in mid-latitudes. Some remarks
concerning the general applicability of the curves are
indicated. First, the refraction corrections decrease with
elevation of the terminal above sea level. Secondly, in
warm, humid regions, such as many tropical locations,
the increase in angle of arrival over that for the case of
no lower atmosphere may be about 11 times the values
shown. Thirdly, in polar regions, particularly under
winter conditions when the absolute humidity is very
low, the increase in angle over that for the case of no
lower atmosphere may be only about two-thirds of the
values of increase shown. Lastly, the corrections for
angles below about one or two degrees may, as a practi-
cal matter, be extremely variable, particularly if condi-
tions conducive to superrefraction occur.
In siting an antenna intended to function effectively
for small angles of arrival and departure, it is insufficient
merely to provide a suitable site on which to perform the
construction. Nor is it sufficient simply to have an un-
obstructed horizon in the desired azimuth for the de-
sired angle of departure or arrival. When the angle of
departure or arrival is small, the ground for a consider-
able distance in front of the antenna plays a critical role
in formation of the lobe pattern. The problem of ground-
reflection lobes and general requirements for a smooth
first Fresnel zone has been given considerable study,
particularly with respect to the ground radar siting
problem. For the purposes of the following discussion it
is assumed that the lowest lobe will be effectively formed
when the terrain in front of the antenna is flat and
smooth over an area no smaller than the first Fresnel
zone. Horizontal polarization only is considered so that
1227
lobe formation is not complicated by large variations in
the ground-reflection coefficient over the range of angles
of interest. Mathematically, the antenna is assumed to
be at a point and the ground-reflected wave is assumed
to have the same amplitude as the incident wave but to
undergo a 180-degree phase change at reflection.
The first Fresnel zone is defined, with the aid of
Fig. 64, as the area of ground, assumed smooth and
plane, in front of the point antenna A from within which
ROCRLACE RAY
4, 4.1
d" ---------
Fig. 64?Geometry of first Fresnel zone.
all secondary wavelets, according to Huygens' principle,
contributing to a plane-wave front advancing in the di-
rection of the positive x-axis with an inclination upward
of a, differ in phase by 180 degrees or less from a refer-
ence ray, obeying strict geometric optics. The origin of
the coordinates, 0, is situated at the ground-reflection
point for the reference ray.
For alignment of the maximum of the first ground-re-
flection lobe at an angle of elevation a, it is necessary for
the point antenna A to be at a height, z, given previously
by (43), as follows:
X
z
4 sin
where X is the wavelength. The distance from the an-
tenna base at B to the geometric ground-reflection point
-
is given by:
d= (44)
tan
Now let P(x, y) be a general point obeying the condi-
tions defining the boundary of the first Fresnel zone, so
that
AP ? (AO ? OR) = ?x --(45)
2
where R is the point along the reference ray at which the
phase comparison is made with the wavelet originating
at P.
By using (43) to eliminate X in (45) and expressing AP,
AO, and OR as functions of h, a, and the coordinates x
and y, the locus of P is found, after some simplification,
to be:
4.4.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
1111228
2z\2
(X --tan a)
852
sin2 a
13
[7his is seen to be the standard Cartesian form for an
As a practical matter it is of interest to know the dis-
itance from the antenna base at B to the near edge of the
first zone dN, to the far edge dp, and the maximum width
or the minor axis, of the zone. These distances, in-
creased suitably for the finite dimensions of a practical
'Antenna, determine the extent and location of the
:ground in front of the antenna which must be flat and
'ifree of obstructions for effective lobe formation. From
}: (46) when y=0,
PROCEEDINGS OF THE IRE
y2
8 2 1. Z
160
(46) 140
iso that,
?
2z(1 ?
x ?
tan a cos a/
dbr ?
tan a cos a)
22-0\
tan a ( cos
3 ?
and the maximum width of the zone is:
2
120
100
BO
60
40
20
October
1111
11111?1111?111111111M1111?111111161111
1?511111111111INIMIIIIIIMIE11111MIIIINI
1011111111EMMIN111111111111.1111111MIN
1111111111111?111111?11111111111111111111111111111?11111
NIE011111111?11111111?11111111111111?111111111111
N111111111111M111111111111111111111111111111
MIIIIM111111111111111111111111111111111111111111111111111
1111111111111111111111111111111111111111111111111111111
N?NW= cn 1111111?1111?11?1111111111
11111111kr=13?M11111?11111111111111?11111111111
111111MANNIM111111111111111M11111111111
MINEM?11?111111111111111111111111101111111?111
? 5.m.'s SNISIMINIIIMIIIM0111111111111
111?11111?11111SOMMILSIMMIIIINIMMI
1111111111?11111?111111MFSTINSIMIEM
111111111111111MMIIIMil?1111?1111?11111
11111?=11111111111?11111111111111111?111111111111
5 6
a DEGREES
10
(47) Fig. 65?Antenna heights for aligning first ground-reflection lobe
maximum at indicated elevation angle, for horizontal polarization
and a plane earth.
(48)
(49)
w = 4-V2z. (50)
The effect of the curvature of the earth on the first
'Fresnel zone will be appreciable only if the angle of ar-
rival or departure, a, is very small. The qualitative re-
sult of curvature is to cause the elliptical area to -be re-
duced in size and to be altered into an egg shape elon-
gated with the broad end near the antenna. The quan-
rdty dN is almost unaltered, whereas d is somewhat re-
uced, and dp is considerably reduced. The maximum
!width will be slightly reduced and will occur somewhat
inearer the antenna. An additional effect of some practi-
cal importance is the reduction in the height, z, of the
antenna required for the first lobe maximum to be
formed at a specified angle a. For the plane-earth case
this height was given by (43).
Curves of z and d have been derived, and are plotted
in Figs. 65, 66, and 67 as a function of a for plane earth
and for the curved earth, for three fixed values of X cor-
responding to frequencies of 30, 40, and 50 Mc/s. Some
compensation for the effects of tropospheric refraction
has been introduced for curved-earth case. Fig. 68 (next
page) gives values of di? as a function of a for the three
sample frequencies for a plane earth according to (49).
When a is greater than two or three degrees the values
for a curved earth will be smaller by a negligible amount.
41
I
II
I
I
50mcfs
4011cis
1
Ill
1
1?1
000
900
BOO
Sk
?
PLANE EARTH
CURVED EARTH
? ? ?
1--
II
Ii
i
700
600
1
tv30Must
..\
11
/
ill
1 I
\
I''
400
300
,
A\
\
.
\ _.
\___.
ii
\
\\
200
i00
iiiiii:.
Alb.
.."44111.N.
1%k
__..\
-?
00
as
a DEGREES
10
Fig. 66?Antenna heights for aligning first ground-reflection lobe
maximum at indicated low elevation angle for horizontal polariza-
tion, for plane and curved earth.
When a is less than two or three degrees the curves
shown give values of di1 substantially greater than the
actual values for a curved earth, and provided they can
be satisfied in practice no difficulties can be expected to
result from their use.
It is of practical importance to know what sort of de-
partures from the ideal first Fresnel zone may be
tolerated. For a well-developed first ground-reflection
lobe the entire first zone at least, should be flat, and the
horizon from every part of it should subtend less than
4
,41
1955 Bailey, Batentan, and Kirby: Radio Transmission at VHF by Scattering in the Lower Ionosphere 1229
100
90
80
70
cn
4;1 60
4
3
1
III
I
---=---
II
11
PLANE EARTH---
\
1121
CURVED EARTH
11
=I
11
1
It
1
I
I
I
\
1 11
I
309ths
1---*-- -
1
- -
? ?
.
I- -
-
- -
11
\A
?
-VI
_
\\
\
). .
--. - 1 -1--\1\--
\
--1
..
____
-09
00
09
a DEGREES
LO
Fig. 67?Distance to ground-reflection point from antenna having
first ground-reflection lobe maximum at indicated low elevation
angle for horizontal polarization, for plane and curved path.
10
1
9 ,
8.
30Mcts
0
50 NU,
:11k
1111011111111111wm?
11111
MU=
40 Mc/s
2
3
4
6 7 8 9 10
a DEGREES
Fig. 68?Distance to far edge of first Fresnel zone from antenna
having first ground-reflection lobe maximum at indicated eleva-
tion angle for horizontal polarization and plane earth.
the angle a. Norton and Omberg," applying Rayleigh's
criterion, state that irregularities in the terrain in the
:5 K. A. Norton and A. C. Omberg, "The maximum range of a
radar set," PROC. IRE, vol. 35, pp. 4-24; January, 1947.
first zone should have a departure from ideal smoothness
of not more than one-fourth of the antenna height. As a
practical matter, water surfaces make excellent first
Fresnel zones, particularly for small values of a.
The ray treatment leading to the results presented
above for the curved earth should not be relied upon for
values of a less than about 0 degree at which the hori-
zon, for the heights involved, is actually below minus
half a degree. For one thing the divergence factor begins
to reduce the effectiveness of the ground-reflection in the
formation of the lower lobes, though this may be offset
by other considerations, as discussed for example by
Burrows and Attwood."
MODULATION TECHNIQUES
In the development of systems utilizing vhf iono-
spheric propagation it is important that consideration
be given to the special behavior characteristics of the
received signals in order to realize a system capable of
minimizing or eliminating the undesirable transmission
effects produced by re4ctions from meteoric ionization
And the fading and multipath characteristics of the
nornially received signals. Considerations of path
geometry and meteor velocities indicate that the differ-
ence between the transmitted frequency and the Dop-
pler frequencies should not be greater than about 6 kc/s
for the Cedar Rapids to Sterling path at 49.8 Mc/s. This
difference in frequency is, for a particular meteoric
event, directly proportional to the transmitted fre-
quency.
ESTIMATED SYSTEM PERFORMANCE
A considerable body of signal intensity and noise in-
tensity data has been acquired in the experimental pro-
gram. It is possible, using these data, to estimate the
expected reliability of a system employing a typical
type of service such as single-channel radioteletype
transmission. System reliability estimates are made for
several assumed types of service.
Montgomery37?38 has studied the behavior of several
types of modulation for narrow-band transmission of
binary-coded messages in the presence of fluctuation
noise and has considered the effects of diversity action in
the reception of narrow-band frequency-shift transmis-
sion. His results will be used for estimating ratios of
average signal power to rms noise power required to
establish a specified system performance for radiotele-
55 C. R. Burrows and S. S. Attwood (Eds.), "Radio Wave Propa-
gation," Academic Press, Inc., New York, N.Y., pp. 80-81,119-120;
1949.
37 G. F. Montgomery, "A comparison of amplitude and angle
modulation for narrow-band communication of binary-coded mes-
sages in fluctuation noise," PROC. IRE, vol. 42, pp. 447-454; Febru-
ary, 1954.
33 G. F. Montgomery, "Message error in diversity frequency-
shift reception," PROC. IRE, vol. 42, pp. 1184-1187; July, 1954. '
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
1230
PROCEEDINGS OF THE IRE
type transmission. The required ratios of average signal
power to rms noise power given by Montgomery for
specified error probabilities are based on the use of ideal
" receivers. These ratios will be increased by three decibels
j as an allowance for the nonideal receivers used in
practice. For the examples illustrating radioteletype
operation it will be assumed that a binary error prob-
ability of 2X10, corresponding approximately to a
teletype character error probability of 10-3 for syn-
chronous radio-teletype transmission, will provide a
satisfactory service.
Fig. 69 presents distributions of the values of the
ratios of hourly median signal intensity to rms noise
t} power in a 2 kc/s band observed on three experimental
60
40
3
0 1
WIZ 20
I
1
10
1
I REFERENCE POWER IsPUT TO TRANSMiTTER FINAL t.10PuriER a0KW;
, ESTIMATED ANTENNA POWER 30.1W
1 1 1 i I
FARGO -CHURCHILL
ANCHORAGE- BARROW
I I 1 III II
CEDAR RAPIDS - STERLING -
1

1
i I . 11--- 1 I
i , i
.1 1 4+
1
1
05 I 2 5 10 20 30 40 50 60 70 80 90 95 98 99995999999 9999
PERCENT OF TIME t EQUALS OR EXCEEDS ORDINATE
Fig. 69?Cumulative distributions of the ratios of hourly median
signal intensity to rms background galactic noise observed at
about 50 Mc/s in a 2 kc/s noise band for the year from October,
1951 through September, 1952.
paths over a 12-month period. Fig. 19 may be used for
estimating systems performance at other frequencies.
The assumption will be made that the distributions of
the ratios of hourly median signal intensities to noise
intensities in a 2 kc/s noise band correspond to the
actual distributions of the ratios of average signal power
to rms noise power in a 2 kc/s band. This assumption is
not expected materially to affect the results, as measure-
ments have indicated small differences between the
recorded signal intensities and the rms signal power.
For the first example, an estimate is made of the ex-
pected reliability for synchronous single-channel fre-
quency-shift radioteletype operation over the Cedar
Rapids to Sterling path at 49.8 Mc/s using dual spaced-
antenna diversity and suitable dual-filter receivers ar-
ranged for elimination of meteor Doppler errors. In this
example, the bandwidth of each of the filters is assumed
to be 100 cycles per second.
Required ratio of average signal power to rms +34 db
noise power '(Fig. 6 of ref. 37)
Allowance for use of nonideal receivers
Allowance for diversity gain (Fig. 2 of ref. 38)
Allowance for bandwidth ratio (2 kc/s to 0.1 kc/s)
Required ratio of average signal power to noise
power in a 2 kc/s band
Expected propagational reliability (Fig. 69)
+ 3 db
?14 db
?13 db
+10 db
99.5 per cent.
October
For the second example, an estimate is made for
synchronous four-channel, time-division, radioteletype
operation with dual, spaced-antenna diversity at a fre-
quency of 49.7 Mc/s, using the Fargo-Churchill path.
Each of the dual filters in the receivers is assumed to
have a bandwidth of 400 cycles. Following the pro-
cedure used in the preceding example it is found that the
required ratio of average signal power to rms noise
power in a 2 kc/s band is 16 decibels. From Fig. 9 a.
value of 99.6 per cent is obtained for the estimated re-
liability.
Finally, an estimate is made for the expected reli-
ability using narrow-band frequency modulation radio-
telephony on the Fargo to Churchill path at 49.7 Mc/s.
Dual, spaced-antenna diversity reception is assumed.
It is also a value of 14 decibels for the ratio of hourly
median signal intensities to rms noise power in a 2 kc/s
band is required for satisfactory radiotelephone service.
Reference to Fig. 69 gives a value of 99.9 per cent for
the estimated propagational reliability.
Some additional comments are in order with respect
to the above derived estimates of propagational reliabil-
ity. As discussed previously, the controlling noise is
usually of galactic origin. There are, however, periods
when other kinds of noise, such as atmospherics from
local thunderstorms or from thunderstorms located
within the beam of the receiving antenna and within
optical range, precipitation noise, and man-made noise,
have to be reckoned with. No allowances have been
made for the effects of noise sources other than galactic
and the above estimates are probably optimistic by
several tenths of a per cent. Outages resulting from the
effects of sputter on paths crossing or near the zone of
maximum auroral occurrence are estimated to reduce
the reliability by no more than half a per cent. Further-
more other sources of outage, such as equipment fail-
ures, power failures, and shutdowns for maintenance
are likely to be greater than those resulting from in-
sufficient signal-to-noise ratios. Assuming that outages
caused by these and related factors do not exceed three
per cent, a systems reliability of about 96 per cent or
greater should be realized for the types of service con-
sidered in the above examples.
ACKNOWLEDGMENT
It is desired to acknowledge the valuable contribu-
tions made by our colleagues at the National Bureau of
Standards, and in particular the parts played by V. C.
Pineo in the experimental program, and K. W. Sullivan
in the observational program. The following organiza-
tions have made important contributions to this work.
The Massachusetts Institute of Technology, the Rand
Corporation, the Collins Radio Company, Engineering
Experiment Station of North Dakota Agricultural Col-
lege, E. C. Page Consulting Radio Engineers, and the
Department of Defense.
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
U. S. DEPARTMENT OF COMMERCE
Sinclair Weeks, Secretary
NATIONAL BUREAU OF STANDARDS
A. V. Astin, Director
THE NATIONAL BUREAU OF STANDARDS
The scope of the scientific program of the National Bureau of Standards at laboratory
centers in Washington, D. C., and Boulder, Colorado, is given in the following outline:
Washington, D.C.
Electricity and Electronics. Resistance and Reactance. Electron Devices. Electrical Instruments.
Magnetic Measurements. Dielectrics. Engineering Electronics. Electronic Instrumentation.
Electrochemistry.
Optics and Metrology. Photometry and Colorimetry. Optical Instruments. Photographic
Technology. Length. Engineering Metrology.
Heat. Temperature Physics. Thermodynamics. Cryogenic Physics. Rheology. Engine Fuels.
Free Radicals.
Atomic and Radiation Physics. Spectroscopy. Radiometry. Mass Spectrometry. Solid State
Physics. Electron Physics. Atomic Physics. Neutron Physics. Nuclear Physics. Radioactivity.
X-rays. Betatron. Nucleonic Instrumentation. Radiological Equipment. AEC Radiation In-
struments.
Chemistry. Organic Coatings. Surface Chemistry. Organic Chemistry. Analytical Chemistry.
Inorganic Chemistry. Electrodeposition. Molecular Structure and Properties. Physical Chem-
istry. Thermochemistry. Spectrochemistry. Pure Substances.
Mechanics. Sound. Mechanical Instruments. Fluid Mechanics. Engineering Mechanics.
Mass and Scale. Capacity, Density, and Fluid Meters. Combustion Controls.
Organic and Fibrous Materials. Rubber. Textiles. Paper. Leather. Testing and Specifica-
tions. Polymer Structure. Plastics. Dental Research.
Metallurgy. Thermal Metallurgy. Chemical Metallurgy. Mechanical Metallurgy. Corrosion.
Metal Physics.
Mineral Products. Engineering Ceramics. Glass. Refractories. Enameled Metals. Concreting
Materials. Constitution and Microstructure.
Building Technology. Structural Engineering. Fire Protection. Air Conditioning, Heating, and
Refrigeration. Floor, Roof, and Wall Coverings. Codes and Safety Standards. Heat Transfer.
Applied Mathematics. Numerical Analysis. Computation. Statistical Engineering. Mathemat-
ical Physics.
Data Processing Systems. SEAC Engineering Group. Components and Techniques. Digital
Circuitry. Digital Systems. Analogue Systems. Application Engineering.
? Office of Basic Instrumentation ? Office of Weights and Measures
Boulder, Colorado
BOULDER LABORATORIES
F. W. Brown, Director
Cryogenic Engineering. Cryogenic Equipment. Cryogenic Processes. Properties of Materials.
Gas Liquefaction.
Radio Propagation Physics. Upper Atmosphere Research. Ionosphere Research. Regular
Propagation Services. Sun-Earth Relationships. VHF Research. Ionospheric Communications
Systems.
Radio Propagation Engineering. Data Reduction Instrumentation. Modulation Systems. Navi-
gation Systems. , Radio Noise. Tropospheric Measurements. Tropospheric Analysis. Radio
Systems Application Engineering.
Radio Standards. High Frequency Electrical Standards. Radio Broadcast Service. High Fre-
quency Impedance Standards. Electronic Calibration Center. Microwave Physics. Microwave
Circuit Standards.
1.6C00,11125-0L
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4
Declassified in Part - Sanitized Copy Approved for Release @ 50-Yr 2014/06/09 : CIA-RDP81-01043R003000180001-4
Department of Commerce
National Bureau of Standards
Boulder Laboratories Postage and Fees Paid
Boulder, Colorado U. S. Department of Commerce
Official Business
,
.?
Declassified in Part - Sanitized Copy Approved for Release ? 50-Yr 2014/06/09: CIA-RDP81-01043R003000180001-4