MAPPING FROM SATELLITE PHOTOGRAPHY

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP79B01709A000400010001-7
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 25
• Document Creation Date: December 14, 2016
• Document Release Date: March 24, 2003
• Sequence Number: 1
• Publication Date: December 1, 1967
• Content Type: REPORT

Approved For Release 2003/04/23: CIA-RDP79B( 9WA6bM* ( 1 I IMA DoD
MC&GWG Minutes 1-24 (Apr 65-May 67)
MC&GWG Documents 1-8 (Apr 65- Mar 67)
There are 3 MC&G Files in this office-
1965-67 file (this one), 1968 file, & 1969 file.
(To be combined later for archiving. )
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
BEST COPY
Available
THROUGHOUT
FOLDER
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
(Revised for publication, December 1967)
Frederick J. Doyle
Chief Scientist
Autometric Operation
Raytheon Company
AMERICAN SOCIETY OF PHOTOGRAMMETRY
National Science Foundation
Visiting Scientist Program
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
Frederick J. Doyle
Within the memory of many men now alive, maps were made by the
ground surveyor lugging his theodolite, plane table, and alidade across
the plains and up the mountains. In the years following World War II,
the ground surveyor was largely replaced by the cartographic aerial camera
and the photogrammetric plotting instrument. This innovation produced
quantum jumps in production, geometric accuracy, and content of topo-
graphic maps.
Where Do We Stand Today
At the present time fully ninety percent of all new map compilations
are produced photogrammetrically. Yet in spite of more than half a century
of effort, the mapping task is woefully incomplete. Around the world only
about half of the land area is covered by principal arcs of triangulation,
and much less than half by first and lower order triangulation. The status
of compiled maps is indicated in this table.
World Map Coverage
Quality
Small Scale Medium Scale
Large Scale
Remarks
1:600,000 1:75,000
1:75,000
and to
Smaller 1:600,000
and
Larger
Adequate
--- 15%
5% }
Principally
Require Revision
30% 5%
5%
U. S. and Europe
Inadequate
70% 40%
10%
Nonexistent
--- 40%
80%
Adequate
U. S. Map Coverage
100% 96%
64%
Inadequate
--- 4%
4%
Nonexistent
--- ---
32%
Data compiled by Army Map Service, and exclude Antarctica
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B011709A000400010001-7
Of primary concern is the fact that the rate of obsolescense of
existing maps nearly equals the production of new maps, so that with
present techniques the job will never be completed. Furthermore, the map
production cycle is about three years from photography to printing so that
the new map is three years out of date on the day it is published.
Geographers would like to see the million scale map of the
world (IMW) completed. They state that most 1:250,000 maps are deficient
in content. They need large scale 1:25,000 maps of all populated areas.
Geologists, engineers, and other map users need similar scales. However, of
first priority to all is the rapid revision of existing maps. Some sort of
Parkinson's Law operates to make maps most difficult to compile and most
rapidly obsolete in precisely those areas where they are needed most
urgently. Maps of large urban areas should be,recompiled annually. The
current cycle in the United States is five to ten years.
The Gemini Photography
The use of the artificial satellite as a camera carrying vehicle
is expected to provide a jump in mapping capability comparable to that
which the airplane made over the ground surveyor. The Gemini photography
was made with an ordinary hand held Hasselblad camera. On missions 5 and
7 photographs were made of the Cape Kennedy launch area. These two pictures,
never intended for cartographic purposes, were used to revise the plani-
metric detail on the existing Army Map Service 1:250,000 map of the area.
In another application of the Gemini photography, the U. S.
Geological Survey compiled a mosaic of most of Peru, parts of Bolivia
and Chile. Control points were selected from existing 1:1,000,000 maps
and identified on the individual Gemini Frames. These were then rectified
and photograph tilts of as much as 40? were removed. Despite 20,000 feet
of topographic relief a reasonable match was obtained, and the resultant
photomap gives a view of the country never seen before.
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B0-1709A000400010001-7
But these are baby steps. What could be done with a system
actually designed with cartographic objectives in mind?
Map Requirements
Before exploring the potentialities of space cartographic
systems, it would be well to recall the requirements for producing maps.
A topographic map contains three kinds of information. The
first is content, i.e., the details which are represented on the map.
Content is provided by photographic resolution and scale, or more directly
by ground resolution. In this area the exact capability of space photo-
graphy remains to be demonstrated. For a variety of reasons it seems
probable that the ground resolution obtainable with a given lens-film
resolution will be higher from space than a simple geometric extrapolation
from the scale of airplane photography would indicate. A useful criterion
to apply is that the photography can be enlarged until its resolution is
equivalent to between 10 and 20 line pairs per millimeter. This will
present all the information which the human eye can extract without en-
larging the map scale by magnification. Not all map information is
obtainable directly from photography, regardless of its scale or ground
resolution. Data such as political boundaries, place names, and detail
obscured by vegetation must be compiled on the ground or from other sources.
It is estimated that if the suggested resolution criterion is applied,
about 80 percent of the total map information can be extracted from the
photographs.
The second kind of information is the position of the objects
shown on the map. For some applications the relative positions of all
objects will be sufficient, but it is usually necessary and always
desirable to attempt to specify all positions with respect to some well
defined coordinate system, either local or national. Map positions are
indicated by the reference graticule: longitude and latitude, state
coordinates, or military grid lines.
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B01709A000400010001-7
The third kind of information is elevation - generally shown
by contour lines above a reference surface - usually mean sea level.
In the United States, criteria for position and elevation on
maps exist in the National Map Accuracy Standards. Applied to photo-
grammetric mapping,these standards, and the higher resolution criterion
defined above, result in the values given in the following table. A fixed
contour interval does not necessarily go with a given map scale. An
interval fine enough to depict the terrain will be chosen.
MAP ACCURACY REQUIREMENTS
Map
Std. Error
Ground
Contour
Std. Error
Scale
Position
Resolution
Interval
Elevation
1,000,000
300 meters
50 meters
500 meters
150 meters
250,000
75
12.5
100
30
100,000
30
5.5
50
15
50,000
15
2.5
25
8
25,000
7.5
1.3
10
3
The numbers in this table represent the objectives against which
a space cartographic system should be evaluated.
Orbital Constraints on Photographic Coverage
It is immediately clear that if full coverage of the Earth is
required, a near polar orbit is necessary. Of course, if mapping is to
be restricted to specific areas, orbits of lower inclination can be
employed. But for elementary discussion only near polar orbits will be
considered.
An orbit is approximately fixed in inertial space and Earth rotates
beneath it. At practical altitudes the satellite period is approximately
1 1/2 hours and in that time the Earth will rotate some 22 1/2? of longitude,
i.e., about 2500 km. Since no reasonable camera can cover 2500 km on a
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B0-1709A000400010001-7
single photograph it is necessary to arrange the mission such that consecu-
tive days will fill in the gaps. There exist so called "resonant" altitudes
at 4, 145, and 303 nautical miles, at which each day's coverage would
exactly duplicate the preceding day and the gaps would never be filled.
In order to perform the gap filling function efficiently, it is necessary
to make compromises in selecting orbital altitude, eccentricity, and
inclination. Although orbits as low as 80 n.m. can be flown, for several
reasons including spacecraft lifetime, an altitude of about 125 n.m. (232 km)
is desirable.
As illustrated in Figure 1, the other critical parameter is the
width of the ground track covered by the camera. This dimension, divided
into the 2500 km between consecutive orbital passes will determine the
minimum number of days in orbit which would be?required to obtain complete
coverage in the gaps. Quite obviously, if the spacecraft can remain in
orbit for more than this minimum time, it will get more than one look at
each spot. This is clearly desirable in view of the cloud cover which may
be expected.
Resolution and Map Scale
The relationship between camera focal length, orbital altitude,
lens-film resolution, and ground resolution is shown in Figure 2. For
a wide angle cartographic camera, current technology limits average lens-
film resolution to approximately 50 lines per millimeter. Thus, as indicated
by line 1, a standard 6 inch camera, flying at 125 n.m. altitude, with this
resolution, would produce a ground resolution of about 27 meters. Comparing
this number against the resolution requirements stated earlier, it is evident
that such a camera system would provide map content adequate for maps at
about 1:500,000 scale. In order to obtain the 12 to 15 meter resolution
required for maps at 1:250,000, a frame camera of 12 inch focal length,
indicated by line 2, would be required.
To produce adequate resolution for the larger map scales with
wide angle camera systems restricted to 50 lines per millimeter would
Approved For Release 2003/04123 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B01709A000400010001-7
COVERAGE WIDTHv = MINIMUM DAYS IN ORBIT
Figure 1 - Coverage of Satellite Photography
Approved For Release 2003/04/23 : CIA-RDP79B01709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDF2979B01709A000400010991-7
Figure 2 - Relation Between System Parameters
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B0,1709A000400010001-7
require cameras of extremely long focal lengths and unreasonable film
formats. For this reason consideration is given to panoramic cameras
which are capable of producing resolution between 100 and 200 lines per
millimeter. Such cameras, however, have inherently poor geometric fidelity,
and cannot satisfy the requirements for position and elevation accuracy.
Line 3 on the chart shows that a 24 inch panoramic camera at 150 lines per
millimeter could produce about 2 meters ground resolution - adequate for
standard maps at scale 1:50,000, or, by relaxing the resolution criterion
slightly, for maps at scale 1:25,000. Panoramic cameras require sophisticated
and expensive photogrammetric instrumentation not generally available. For
this reason an eventual operational system for producing or revising large
scale maps may well go to longer focal length, narrow angle, frame cameras,
which might attain 100 lines per millimeter and a corresponding ground
resolution of 3.5 meters. The ground width covered by such a camera would
necessarily be small. As a consequence the satellite would require a very
long lifetime in order to be able to photograph any desired area with
vertical pictures, or else the camera would have to take oblique pictures
to the side of the ground track.
Geometric Map Accuracy
At an elementary level, the position and elevation accuracy
obtainable by photogrammetric procedures is:
dP = f dx
dP = ground position accuracy
dH = ground elevation accuracy
H = flight altitude
B = distance between exposures making up a stereo pair
dx = accuracy of image measurement
-8-
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79BO4709A000400010001-7
Stereo base B is obtained by exposing the photographs at time
intervals such that some part of the ground area covered by one photograph
is also covered by a following photograph. With the 6 inch standard
camera, consecutive photographs overlap by 60 percent and B = 0.6 H. In
order to obtain adequate B with a 12 inch camera, a film format of 9 x 14.5
inches is proposed with a 9 inch dimension perpendicular to the flight
direction. Consecutive photographs will overlap by 67 percent, and
a stereo model will be composed of alternate photographs. This arrangement
will provide an effective B = 0.8 H. Because 24 inch cameras have a
narrow field of view, they cannot achieve adequate B by overlapping ver-
tical photographs. Consequently two cameras will be required in a
"twin convergent" configuration with one camera directed forward along
the flight line and the other directed aft. If the angle off the vertical
is 20? for each camera, the effective base in each stereo model will be
B = 0.7 H.
The current level of accuracy in recovering the position of
an image on a single photograph is approximately dx = 0.005 mm. This
is representative of the relative accuracy which can be obtained in a
single stereo model. However, as every photogrammetrist knows, a stereo
model must be scaled, positioned, and levelled before geometric map
data can be extracted from it. Conventionally this is done by reference
to ground control, and errors accumulate alarmingly as one departs from
the control.
In this regard, satellite photography will have an enormous
advantage over aircraft photography. The satellite orbit is mathematically
predictable, and if the time of each camera exposure is recorded precisely,
the position of the camera can be accurately determined. Furthermore, as
shown in Figure 3, a photograph of the star field can be made in synchronism
with each terrain photograph, and measurement of the stellar photograph
will provide the absolute angular orientation of the camera to a few seconds
of arc. These data are equivalent to having ground control in every stereo
-9-
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B01709A000400010001-7
b - TIME BETWEEN
EXPOSURES + 0.5 MS
H - EPHEMERIS
RECONSTRUCTION
w 4 K - ATTITUDE
RECORDING CAMERA
Figure 3 - Exterior Orientation of Satellite Photography
Approved For Release 2003/04/23 : CIA-RDP79B01709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B0'1709A000400010001-7
model. The consequence is that errors will not accumulate to the same
extent when the photographs are triangulated in a strip or block, and a
value of dx = 0.020 mm is expected to be a reasonable estimate of the
absolute accuracy with which image positions can be recovered.
If the appropriate values of H, f, B and dx are applied for the
three camera systems under consideration, the values listed in the following
table are obtained.
GEOMETRIC MAP ACCURACY OBTAINABLE
Altitude H (125 n.m.)
6 inch
camera
232 km.
12 inch
camera
232 km.
24 inch
camera
232 km.
Focal length f
152 mm.
305 mm.
610 mm.
Relative accuracy position dP
7.7 m
3.8 m
1.9 m
elevation dH
12.8 m
4.8 m
2.7 m
Absolute accuracy position dP'
30.7 m
15.2 m
7.6 m
elevation dH
51.3 m
19.0 m
10.9 m
If these geometric numbers and the resolution numbers previously
discussed are compared with the requirements for mapping at different
scales, the capabilities of the three camera systems can be summarized.
CAMERA SYSTEM CAPABILITY
Relative Mapping
6 inch
camera
12 inch
camera
24 inch
camera
Content for map scale
500,000
250,000
25,000 to 50,000
Position accuracy for map scale
25,000
25,000
10,000
Elevation accuracy for contour interval
50 m
15 m
10 m
Absolute Mapping
Content for-map sclae
500,000
250,000
25,000 to 50,000
Position accuracy for map scale
100,000
50,000 -
- - -
Elevation accuracy for contour interval
200 m
50 m
- - -
-11-
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B0t709A000400010001-7
The 6 inch system could satisfy the requirement for world wide
small scale 1:1,000,000 and 1:500,000 mapping. The much more serious
problem of medium scale 1:250,000 mapping could be satisfied by the 12
inch camera system which also has the important capability of providing
adequate geometric control for the preparation of large scale 1:50,000
and 1:25,000 maps. The content for these large scale maps could be
provided by the 24 inch camera systems. Thus an ideal system would be
composed of both the 12 inch and 24 inch cameras. This would largely
satisfy all current requirements for mapping at scales smaller than
those needed for actual engineering construction.
What Are The Prospects
A year ago, the Department of the Interior announced its project
EROS - for Earth Resources Observational Satellite. Although a number of
proposals are under consideration, the most promising seems to be a camera
system designed and built by RCA Astroelectronics Division. The camera is
the ultra sophisticated child of the highly successful camera used in the
now operational TIROS TV weather satellite system. The characteristics of
the TIROS and the new vidicon are as follows:
TIROS
EROS
1/2 inch
2 inch
1/4 inch square
1 inch square
Resolution
400 lines
8000 lines
Resolution elements
160,000
64,000,000
Sensitivity
0.4 ft. candle sec.
0.01 ft. candle sec.
-12-
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B01709A000400010001-7
In order to meet the requirements of a large number of scientists
in the fields of agriculture, forestry, geology, geography, hydrology,
and other natural resource disciplines, it is proposed to use three cameras
to acquire photography in three different spectral bands.
These bands are selected to provide:
(a) The sharpest demarcation between land and water areas,
(b) The maximum' discrimination of vegetation types,
(c) The greatest penetration of water.
Each frame of the proposed pictures will cover an area of 96 x 96 nautical
miles and will provide a ground resolution of 100 to 200 feet from a
circular orbit at 300 nautical miles. The orbit inclination will be 97?
sun synchronous so that the illumination conditions will be identical for
adjacent orbital passes. The satellite will weigh about 850 pounds,
and can be launched by a Thor Delta from the Western Test Range. Solar
cells and batteries will provide power for the cameras and for a 4
megacycle communication bandwidth required to transmit the pictures to
ground stations. A video tape recorder will store the pictures until the
satellite is within range of a ground receiving station. A lifetime of
at least one year is planned so that repeated coverage can be obtained
to determine the time variant characteristics of areas of special interest.
NASA's Lunar Orbiter program has clearly demonstrated the
ability to acquire and transmit extremely high resolution photographs
from space. However, photogrammetrists have learned to be suspicious of
the geometric integrity of transmitted and reconstructed pictures. This
fact and the lack of adequate stereo overlap makes the use of EROS photo-
graphy for geometric mapping marginal.
Approved For Release 20031b4f23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B0,1709A000400010001-7
In the Apollo program, it will be necessary to perform a
number of Earth orbit missions to check out various parts of the system
and procedures. NASA is studying the possibility of using one of these
missions to carry a number of Earth sensing experiments. Among these
would be a 6 inch focal length, 9 x 9 inch format, cartographic camera
with a coupled stellar camera.
A study has been performed by Martin-Marietta Corporation to
define the integration of this experiment with the other sensors and the
spacecraft. They have proposed a new equipment carrier module which would
replace the Lunar Module. It-will consist of a welded aluminum truncated
cone enclosure 84 inches in diameter at the experiment mounting end
and 110 inches long overall. A truss, which will support the cone in
the spacecraft adapter, will also serve to support all experiments not
requiring in-flight access or pressurization. The cone itself will be
pressurized and the camera system will be among the experiments in the
pressurized section. The astronaut will have access to the experiment
section through the air lock for such functions as changing the film
magazines.
In operation the command and service module with the equipment
carrier module will have its longitudinal axis normal to the Earth's
surface and the cartographic camera will look down through the base of
the cone. This configuration will provide the astronauts maximum
terrain visibility through the Command Module windows.
The proposed parameters for the Apollo mission are:
? Cartographic camera
6 inch focal length, 9 x 9 inch format
140 n.m. circular, 500 inclination
(provides complete U. S. coverage)
-14-
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B01709A000400010001-7
14 days
(provides 2 looks at every point)
900 frames each covering 210 x 210 n.m.
(limited by stowage in CM for return to Earth)
Total coverage
13 million square miles
Proposed launch date
Spring 1969
Manned missions are extremely costly to fly, and they are restricted
in the amount of photographic film and other data which can be physically
returned to Earth. For these reasons, NASA is also considering unmanned
photographic missions, and a study has been performed by Lockheed Missiles
and Space Company to define the characteristics of such a system.
NASA envisions a spacecraft, illustrated by Figure 4, carrying
three 6 inch focal length, 9 x 9 inch format, cartographic cameras. The
use of three cameras will provide the multi-spectral capability for re-
source evaluation in addition to cartography.
The exposed film would be returned to Earth in a data recovery
capsule. This part of the system has been developed and proved by General
Electric Re-entry Systems Division for use in several Air Force experimental
programs.
The general procedure is to mount the experiment instrumentation
(cameras in this case) in the spacecraft and to feed the data (exposed film)
to the attached re-entry vehicle. When the data acquisition mission is
completed, the recovery vehicle is separated from the spacecraft and re-
enters the atmosphere. A parachute is deployed and the data package is
snatched by aircraft. The spacecraft itself is then deboosted and splashes
into the ocean. The recovery technology is clearly available. It remains
ILLEGIB
-15-
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
REENTRY VEHICLE
Figure 4 - Schematic for Film Recovery Satellite
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B0'1709A000400010001-7
to adapt the re-entry vehicle to the handling of photographic film. It
is estimated that between 100 and 200 lbs. of exposed film could be
returned from a single mission. With the vehicle in a sun synchronous
polar orbit, with a lifetime of 3 weeks, a single camera could photograph
nearly 30 million square miles. If the film load were divided among three
cameras, the system could cover the entire United States in several spectral
bands with a high probability of getting successful coverage.
What Are the Chances of Success
When a spacecraft is in orbit, its lifetime is necessarily limited.
Since all costs have been accrued when the lifetime is terminated, the
success depends entirely upon whether the mission objective has been accomplished
during the lifetime. For a photographic mission, this is critically depend-
ent upon the weather--or more specifically on the percentage of cloud free
area during the daylight hours. Many studies of world wide cloud distri-
bution have been performed, and the results of a number of them may be
summarized as follows:
(1) With one look, a satellite will probably photograph 50
percent of the desired area. A second look will probably get 50 percent
of the remainder; a third look 50 percent of what is left. This series
would require an infinite number of looks to get 100 percent coverage.
On the other hand, 4 looks would give 94 percent coverage and 5 would
give 97 percent.
(2) To acquire photography at least 84 percent cloud free
over the United States, a satellite launched in September would require
2 looks for the total southwest and a major portion of the midwestern
and eastern sections; 3 looks would get most of the northwest but would
still lose a section through Texas, Missouri, and the Dakotas.
(3) The probability of successfully photographing an area as
a function of its percentage of possible sunshine is:
-17-
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B01709A000400010001-7
Percent of
Probability of success
sunshine
2 looks
4 looks
Over 90
0.99
0.99
80
- 90
0.95
0.99
70
- 80
0.90
0.99
60
- 70
0.82
0.96
50
- 60
0.75
0.93
(4)
The percentage of coverage with an 0.9 probability of 1 or
more cloud free passes is:
2 looks
4 looks
U. S. - summer
84%
98.5%
U. S. - winter
20
77.5
World - all year
17
65
As a generalized conclusion, these studies seem to converge
on the fact that a system providing 4 looks at the areas of interest is
approaching the point of diminishing returns. With a 4 week lifetime
for the spacecraft, the 12 inch camera would get 2 looks, and the 24
inch camera 1 to 4 looks depending upon the configuration selected. One
or two satellites would probably achieve adequate coverage of all areas
which are not perenially cloud covered. To hope to photograph such
areas from a satellite is probably not realistic.
Is It Economically Feasible
Presume that an unmanned satellite is launched carrying both
the 12 inch and 24 inch camera systems, and that a 200 lb. film load is
distributed so that both cameras would be able to photograph the same
total area. Using thin base black and white film and the orbital altitude
of 125 n.m., each of'the camera systems would be able to photograph 9 x 106
square miles. If the total mission cost is $15 x 106, the photography
costs $1.67 per square mile for double coverage. Even if the photography
is only 50 percent useful, the cost is $3.34 per square mile.
-18-
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B01709A000400010001-7
Compared to these costs, conventional aircraft photography in
the U. S. costs the U. S. Geological Survey between $2.50 and $4.00 per
square mile on contract basis. Foreign photography, based on 650,000
square miles in South America, costs the U. S. Air Force about $12 per
square mile for single coverage. Thus purely on the basis of cost per
square mile, space photography, particularly of remote areas, is clearly
more economical.
The problem with these figures is that 1000 square miles from
a satellite would cost the same $15 x 106 as the 9 x 106 square miles.
Looking at the problem in this way, and using $4 per square mile as the
cost of airplane photography, the breakeven point would occur at 3.75 x 106
square miles. That is, if more than 3.75 x 106 square miles of photography
are required the satellite is the economic way to get it.
The fact of the matter is, however, that the total map producing
capability of the United States could not turn out 3.75 x 106 square miles
of conventional mapping in a year. However, the basic reason for this is
found in the number of photographs involved.
Figure 5 shows the coverage produced by conventional aircraft,
photography compared to that which would be given by the 12 inch frame and
24 inch panoramic cameras. Also shown is the area of a standard 1:250,000
scale map sheet.
To photograph the 3 x 106 square miles in the United States,
a standard 6 inch mapping camera flown at 30,000 feet would require a
minimum of 100,000 stereo pairs. The proposed 12 inch camera system
flown in a satellite would cover the same area in about 500 stereo models.
To process 100,000 stereo models is unreasonable, whereas 500 is clearly
within the capability of most agencies.
In addition to the simple processing of 100,000 stereo models,
mapping by conventional photogrammetric procedures would require several
hundred thousand ground control points at about $300 per point, whereas
the 500 stereo models of space photography could be compiled with a few
thousand control points. Of even greater importance is the fact that
-19-
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79B0t709A000400010001-7
DIRECT
ION
24" PANORAMIC 740 SWEEP
24" PANORAMIC 440 SWEEP
1:250,000
MAP QUADRANGLE
f = 12 INCH
9 x 14.5 INCH FORMAT
FLIGHT
f = 6 INCH
9 x 9 INCH FORMAL
H = 30,000 FEET
f = 6 INCH
9 x 9 INCH FORMAT
H = 12,000 FEET
Figure 5 Comparison of Coverage Obtained
Approved For F 1@TAAi200j3M4 :bI 14L5Pq9BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
the control points for the space photography could be established by
triangulation of the photographs themselves. This is not possible with
the conventional photography because the errors in the triangulation
accumulate with the square of the number of photographs involved. This
is basically what makes it possible to predict that space photography
will be able to do the job at all.
The remaining question is whether the 500 stereo pairs will do
the same mapping job as the 100,000. This is the great imponderable,
because there is no experience in mapping from space photography. The
figures indicate that the proposed systems will probably do the job, if
other parts of the mapping system are given the same attention as the
spacecraft and its cameras.
The final fact is that with space photography, useful products
can be made which are totally impossible from conventional aerial photo-
graphy. These include:
a) A synoptic mosaic of continental areas at scale of 1:1,000,000
or 1:500,000 which is obtainable from the 6 inch photography.
b) Photogrammetric control for maps at scale 1:24,000 anywhere
in the world. This is obtainable from the proposed 12 inch
photography.
c) Compiled maps at scale 1:250,000 anywhere in the world. This
is also obtainable from the proposed 12 inch photography.
d) Large scale, rapid response, mosaics and revised maps for
any selected area in the world. These are obtainable from the
proposed 24 inch photographs.
No economic analysis of space cartography would be complete
without consideration of the data processing part of the map production
routine. Only about one third of the current processing involves the
photographs. A cartographic satellite does not improve the remaining
two thirds. But it drastically alters the inputs, both in type and in
quantity. The formats and focal lengths of space photography may be,
to a large extent, incompatible with the current data reduction instru-
mentation. Clearly, if satellite photography is to be useful on a pro-
Approved For Release 20031(4/23 : CIA-RDP79BO1709A000400010001-7
 Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7
duction basis, detailed consideration and planning is required throughout
the whole course of the map making cycle. This extends as far as a re-
education of map users who may find it necessary to revise their notions
of what is an acceptable map.
Conclusion
We see before us both an opportunity and a challenge. We have
the prospect of obtaining a knowledge of the Earth's surface and its
resources in detail which we could not have imagined ten years ago.
Our generation may be in the position to complete the world mapping
task which was started 5000 years ago when the Babylonians first put
stylus to clay to guide a caravan across the desert.
-22-
Approved For Release 2003/04/23 : CIA-RDP79BO1709A000400010001-7