A CONTRIBUTION TOWARDS SOLVING THE QUESTION OF THE OCCURRENCE OF CHROMOSPHERIC FLARES BEFORE GEOMAGNETIC STORMS

• Document Type: FOIA
• Collection: FOIA Collection
• Document Number (FOIA) /ESDN (CREST): 06131786
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 29
• Document Creation Date: December 28, 2022
• Document Release Date: September 25, 2017
• Case Number: F-2015-01649

'.. UNCLASSIFIED PC4 OgWITkragseaM9niug4TiiENTIAL
SECRET
CENTRAL INTELLIGENCE AGENCY
ROUTING AND CONTROL RECORD
DATE
26 October 1964
TO:
/15
OSIi
(b
ATTN: I
(b
BUILDING
ROOM NO.
TITLE
- Scientific Paper "A Con- (b
tribution Towards Solving the Question of the
Occurrence of Chromospheric Flares Before Geo-
magnetic Storms" (CZECHOSLOVAKIA)
REMARKS
RETATN
yam RETAIN
ni ON LOAN
DOCUMENT(S) MUST BE RETURNED TO
DOCUMENT(S)
CONTACT DIVISION
(b
FOR RETENTION
BY ADDRESSEE
BY (Deadline) I
b
(
FROM:
BRANCH
__
I
-- CONTACT DIVISION
(b
BUILDING-......
1717 H StAllillimigim,._
ROOM NO.
511
EXTENSION
2491 1
)(3)
)(3)
)(3)
roved for Release
2017/09/11
C06131786
Approved for Release: 2017/09/11 C06131786
� -
fAL
6? 4
q
A CONTraLiCTI L,OLVING ZUE22jON OF THE OCCURRENCE OF.C100-
MOSi;HERIU FLAHES BE:FORE GEOMAGNETIC STORMS.
by Bohumila Bedndi'.ova�Novakova
Geophysical Institute,
Czecosl. Acad, Sci., Prague
ABST?.!;.CT
The incorectness of the hypothesis according to which flares
should be the source of corpuscular solar radiation, which is
responsible for the origin of geoLlagnetic storms, is proved on the
basis of determining the number of geomagnetic storms not pre�
ceded by any flares. Using the is of guaranteed large flares
from the IGY of C.S. Warwick 71.4 and making an analys of the
solar situations occuring simultaneously with flares but on the
CM � in the centre of the solar disc, it was shown why some fla-
res were followed by a geomagnetic storm while after others, on the
contrary, a period of absolute geomagnetic qtiet set in.
Since it was shown that geoincic storms are preeeded by changee
in the nagnetic fieltoin the solar chromosphere and because flares
also occur durir-, such changes, the question arises, what role
do flares play here? Are they the cause of such changes or are
they their p-....oduct ? It is deduced, and observations confirm this,
hat olthough flares are produced as a sort of consee_nc_ of
ng-es in the photospheric fie1:3, they can simultaneousl,; be
.2,srded to a certain de:re as the cause of substantial changes
in the chro:,esphere which. lead to annulling of the field at
least in th;:.t component 'nich is relJresented by chromospheric
structure and filaments. Their relationship to geophysical effects
of a solar corpuscular origin is,.holwver, quite intarect and
very limited.
The fact that after the first historically recorded observation of
a flare on September 1, 1859, by Carrinton and Hodson and also after
the two consecutive flares on August 3 and 51 1872, which were observed
by Young 1 a strong geowsznetic storm always followed., led to the hy�
wil.moloNy
Approved for Release: 2017/09/11 C06131786�
�
�Approved for Release: 2017/09/11 C06131786
2
pothesis of a connection between flares and geomagnetic storms. This
'hypothesis was also supported by the classical observations of G.E.
Hale, carried out from 1892, using the new instruments developed by
him : a spectroheliograph and later by means of a spectrohelioscope.
After the discovery of the rad-iotelescope and after it was found by
means of it that in the period when a flare occurs on the Sunlradiation
is sometimes magnified on some e-lengths (much longer than the opti-
cal), it was hoped that this meant an indicator of an effusion of cor-
puscular radiation had been found, i.e. such radiation which is respon-
sible for the origin of geomagnetic disturbances. It was only later
that it was discovered that not even this fact can be used for exam-
ple for prognoses etc.
From the hypothesis on the connection between flares and geomagnetic
storms the opinion gradually civeloped that flares were the direct sourCe
of geoactive corpuscular radiation. However, flares in themselves are not
7 .
high formations land than th:::; are observed on the edge of the solar
e
disc, they do not exhibit any iiirked rise of matter to any great height E2.
They are often accompanied by :c.ising prominences which, as will be seen
later, may be conceived as the consequence of changes in the local mag-
netic field which occur often vAhout flares and sotimes during a flare.
It is this random simultaneity of the two effects which is the cause
of frequent mistakes so that tare are some authors who speak of fla-
res but obviously have in mind rising prominences. That the are two
quite different effects, whib probably have no direct physical
inter-relationship, is borne o..1-t by the fact that disappearing filaments-
rising prominences occur ri-,2e--z4ci:7---zu:-.:-:.,%:-xzadxcdoce-J.1,1krz.x-ijczazxx-igatanate
very often in absence of flares. It seems that the two phenomena are
likely the consequences of one and the same cause. Sometimes they
occur simultaneously and other times only one of them appears.
Approved for Release: 2017/09/11 C06131786�
-
- Approved for Release: 2017/09/11 C06131786
Flares occur elnost universally in active centres and are thus on-
ly one part of a complex of events which occur shortly after one ano-
ther or at absolutely the same time. From the consequences occurring after
either a shorter or a 1Sker interval it is impossible to deduce a conne-
ction with the different effects separately. This is even more difficult
with geomagnetic storms which eibit a longer time lag after Sudden chan-
ges occurring on the Sun. If, he?..ver, the mechanism of the effusion of
corpuscular radiation and its p:-cpagation towards the Earth is to be found,
it must first be determined wheter such a phenomenon can be found among
the known expressions of -solar c,etivity which, in each period and during the.
whole cycle of solar activity, wbuld be reliable indicator of geomagnetic .s
storms. It is also necessary tfJ (c;ive physical reasons for each connection
found. One must also take into consideration an explanation of the fact
that sometimes no connection was; found. As is clear from the many papers
already published, it was found jn the Geophysical Institute of the Cze-
choslovak Academy of Sciences tht of all the known expressions of solar
activity studied. by optical meds the best indicators of the origin of
geomagnetic storms are filamenta and the fine chromospheric structure under
certain circumstances. As will seen below, in the period of greater
solar activity, i.e. in the pericd of flare occurence, these are erup-
tive and disappearing filaments, denoted in the Geophysical Institute by the
cane term - unstable zl filaments - which, only if they are in a suitable posi
tion, safely indicate when a geoaagnetic storm with sac will occur
This is because such phenomena, as will be seen later, provide direct
information on changes in the mEp:netic field of the Sun. And it is then
only these changes which are actually the direct indicators of changes in the
corona and thus also of the origin of geomagnetic storms.
Doubts as to flares being the source of corpuscular radiation were
raised by D.Van Sabben , R.P.'datson 5 land O.M. Barsukov6j. The sta-
',. J
Approved for Release: 2017/09/11 C06131786.
Approved for Release: 2017/09/11 C06131786i
A
tistical paper by C.S. Warwick and R.T. Hansen,7 :shows the not very highs*
xima of the Ap�indices after flares in the period of greater solar activity
for flares nearer to the CM but does not give positive results for the
period before the minimum. It ctrl easily be supposed, and later this will.
be quite 'clearly proved, that the circumstance can regarded as merely a
statistical phenomen,
the more elements contained
two of them are found which
in a certain time interval.
also .lear
by t. two
are her
The t:tfect
the ,,able given in
series compared, the more frequently
simultaneous or follow one another
is the same although one series is
much more nuLlerous thin the oth,.1 as is the case when the nmnber of
f13re5 grectly e.xcecds, particully in periods of greater activity, the
number of geomanetic storms. It is seen from Tab. I. and even better
from ig.1 that a decrease in ti:e number of flares after the maximum
(1959-60) is followed not by deci.c.ase but by an increase in the number of
geomagnetic storms whether flare: of greater importance or whether aZcsia
all, including those with the smLa1est importance, are used for the comm
parison. No connection can be obEe:-ved even if one takes the number of sac
or the greatest storms which , according to accepted world opinion, Should
be connected with flares. but which, as will be seen later, indicate no pro-
nouced connection.
Table I C=narison between nur::tcr of flares and number of geomagnetic
storms.
.11.17ibei
.L L
L:k
O.geU 44�.
Imp.
3,3+
2
1
-L-
total
all
SSC
Yetr
2 I�
5+
1957,2ndha1f
25
205
228 5189
27
17
1958
24
327
351 9668
42
24
1959
30
259
2208
91C
269 3587
50
22
1960
19
118
1490
541
137 2168
72
32
1961
10
38
787
294.
48 1129
35
17
The
as to the
J.
-:\
.'
results in the paper oyylaienka 9 , 10 also give rise to doubts
direct connection connection between chromospheric flares and geomagnetic storms,
4
..-Aporoved for Release: 2017/09/11 C06131786i.,,
,
�
Approved for Release: 2017/09/11 C06131786i
Flares Flares
400- 10000
200
0
Storms
/
50]
25J
1958
5000
1961
ssc
r/7/4
1061
20
10
re)o
1958
1961
1961
�
Fig. 1. Comparison of number of flares with number of geomagnetic storms
in 1958-61 .; top left: flare 2, top right: flare = 1-, bot-
tom left: geomagnetic storm 1Cpraax. 5+ , dashed line - large
�
geomagnetic storm Kpmax. = 7t, Kpmin. = 5-, bottom right: sac.
Approved for Release: 2017/09/11 C06131786.
Approved for Release: 2017/09/11 C06131786
0
Ls given in .-2he eeeher ef e e :esceee e.eper sh:reed 12 that when
studying the flare activity of different active centres during the whole
period of passage over the solar disc a geomagnetic storm occurred only
after the 02 of the active centre in which flares occurred throughout the
period. This means that flares outside the CY, fell flat. This is in comple-
te agreement with the reeu7te of
BerouRov 6 ��
In 1961 a paper was published by B. Bell 13. from which it is
clear that the connection between flares and geomagnetic storms is doubt-
ful if it is guaranteed , and then only in the relatively long interval of
three days, only to 50%. Not even a distribution according to the magnetic
types of spots, in the neighbourhood of which the flares occured, give'S,
4.----
satisfactorT results. Despite these facts flares are continually given in d
relation with geomagnetic storms. But the question of theoretical opinions,
in the sane way so the question of prognoses, necessitates a detailed in-
vestigation. Fr this reason the author decided to supplement the Work of
B.Bell and to find, on the contrary, how- great a percentage of geomagnetic
storms is preceded by flares.
of the heliogeophysicel group
physical Institute which Eli:LS
This kind of research is part of the work
of the geomagnetic department of the Geo-
at leaving no geomagnetic storm unexplained.
For this purpose the material from 50.71. 1957 to the end of 1962
was investigated. Only material feore these years,thanLs to the IGY and IGC,
can be regae ed as prectically complete as regards flares. The work was divi-
ded into several stages.
1) The percentage of geomagnetic, storms for Kpmax 5+ was found which were
not preceded by any flares 2 in intervals of 10-48 hours, 10-60 hours,
10-72 hours. These are relatively long intervals. In en earlier paper th4
present author found shorter intervals, valid for disappearing filaments,
i.e. 28-58 hourai Since , howevee on account of the small number of obser-
vations from which these values ..iere derived they must be regarded as provi-
sional. Since an interval of thre days was used in the work of B.Bell,
such longer time intervals were left in the present paper as well. Indeed
Approved for Release: 2017/09/11 C06131786
"...r�140, 1ft
Approved for Release: 2017/09/11 C06131786
�-�
60-
100
80
60
40
20
0
100
o' 0
60
40
20
0
195?
1950 1951 1962
-4
i
1 1
-
:I
I
I
!
.
1111
1
, ,
-1
-
1
Fig.Z. Percentage of geomagnetic storms Kpmax. from 1957 (2nd half),
1958, 1959, 1960, 1961 and 1962. Vertical. abscissae denote, from left
right, magnitude of percentage for flares from spaces (with respect to
o _ 300 7 00 45 0 900*
CM) : 00 - 100 ; 0� - 200 ; o 0 0 60� ; 00
From top to bottom with respect to commencement of geomagnetic storm
:for intervals of 10�A-Prioved for Release: 2017/09/11 C06131786hr,,,...
�
�
Approved for Release: 2017/09/11 C06131786
if it is ,proved for such inte s, that there 16 no connection between
flares and geomagnetic storms, this is even more likely to be for shorter
intervals.
1) Altogether 248 geomagnetic storms, obtained from the graphs of Kp-
,
indices from Qdttingen, were investigated in the manner indicated. As
regards flares, American catalogues 1 1 15 , 15 , 17 were used. Only
for 1962 was the Quarterly Bulletin from Zurich 18 used . The results are
plotted in Fig.1 , where the magnitude of the number of percentages is
denoted by the vertical abscissa successively from left to right for the
individual years and for flares considered in the areas defined by the
distances from the =: 0�-10� 00-200, 0�-30�, 0�-450, 0�-60� and
aid from the top downwards for the different time intervals, beginning
with the shortest and ending with the longest, as given above.
It is seen that the height of the percentages of geomagnetic storms
not preceded by a flare is considerable in all cases. Of the large number
of flares e.g. in the period of sunspot maximum, there are less geomagnetic
storms without flares than in other years when there were less flares (see
Tab.I and 2ig.1 ). The percentages of geomagnetic storms continually decrease
the
with a gradual increase in flares as the distance from C'A: grows. They
also decrease gradually and disrsportionately as the length of the time
interval increased which is again connected with a rise in the number of
flares, the more likely one is to occur in the critical interval before a
geomagnetic orm. However, such an occurrence need not indicate an inter-
dependence. It is merely a statistical matter. If this fact is taken into.
consideration, it must be deduced from the graph in Fig.2 that the conne-
ction between flares and geomagnetic storms can be only random. Later this
this conception will be made somewhat more exact and supplemented.
2) Since it might be objected that it is actually only large geomagne-
tic storms which are connected with flares, it Was necessary to investigate
the percentage of the number of geomagnetic storms with Kpmax-------= 74- and
Kpmin= 5-.In this part of work the Quarterly Bulletins LIS were used in
APoroved for Release: 2017/09/11 C06131786
Approved for Release: 2017/09111 C06131786
,0
100
Pt.
1
"
195? 1958 195.9 1960 1961
Fig. 3. Pcrcentage of larE,e geomagnetic storms ( Kpmax. = 7+,
Kpmin:t5-) not preceded by flare at distances from 0� to
10� from CM. White strip : importance 7I;: 3 ; lightly dashed
strip : = 2 ; dashed strip : = 1-. For 1959-61 white str
coincides with lightly dashed strip.
Annrnved for Release: 2017/09/11 C06131786
Approved for Release: 2017/09/11 C06131786
0/0
100
80
60
40
PO
.(
1.957 1958 1.959 1960 1961 1962
Fig.4. Percc;ntege of geomagnetic storms with sudden commencement not
pl'eceded by flare ( imp. -2 2) at distances from 00
to 10
from CM.
Approved for Release: 2017/09/11 C06131786
Approved for Release: 2017/09/11 C06131786
7
the hope the; nu.;lber of fl..22e �;:s:,IS be cs ccm-�aete as possible and tn...%
�
portances = 3 , =2 and = 1- occurring between 0o to 100 from the CM in
intervals of -3 days to -1 day with respect to the commencement of the
geomagnetic storm ( first part of Tab. II ) were considered separately. The
results are plotted in Fig.2. Since there were no large geomagnetic storms
in 1962, only material up to the end of 1961 was treated. It is seen that
even for large geomagnetic storms the percentages of cases without preceding
flares were considerable in 1958-61 not only for the largest flares but
also when using flares -1.2 , when they were above 50%. Although the per-
centages of large storms without flares 1- are zero for 1957 and 1959,
in 1961 they already reach 6C. This circumstance can easily be explained
on the conception of a large number of flares in the years around the ma-
ximum (see-Tab. I ).
3) For the investigation of geomagnetic storms in relation to flares
to be comple-e, one must also conaider ssc which are likely to be connec-
ted with sudden solar pheno::ena on �:le CY., such as flares otc. We investi-
gated 128 35C chosen from the ;-:.-a)hs of prof. J. Bartels. In this part
of the work
of importance =2 which occurred in the interval from
,
�) to 100 from the 0i were used. The hei,-;hts of the percentages of geo-
magnetic s4.;.)-rms with sac from 7957-62 not preceded by flares are plotted
in .17,.;g. 4. Ecgnnng w'th 1958 those are values above 53 %. The maximum
is reached in 1961 (9). It is seen that not even with geomagnetic storms
with a sudden comm2nceillent is any connection found with flares in the
neighbourhood of the CY..
Table II. Percenta7e of number o
Kpmin. = 5-), not prec
XaT..lbor of flares befo:
f 1sr1e reomagnetic storms (Xpmax = 7+,
by any flare from OC -100 from CM.
flares from -3 to
Number -1'tw �
Number of flares
Year of --5 -=2 -==1- �ro -to+Jdaylfx:om0tot2dayUDefore+agter_
storm number 5L,number Z:number 3 2 ("5:1-
j
;
2 1-
)57
klf
.9f F�2 , 9
7
78
5
55
2
559 8
4
50
4
� 50
0
7
7
3
�
,
9
27
i 91
1
1
2
5
30
�
�
1+7 +22
+61
2
2
14
LZQ
0
4
39
!+2 +10 ,
+20
C
11
<
0
4
17
+7
+47
25
5
7
56
2
49
-3 !
7
e1/4
0
16
;
0
1
17
/ 0:-i
1
� Approved for Release: 2017/09/11 C06131786__________
Approved for Release: 2017/09/11 C06131786
O+1 2,-
-3 ,-2 ,-1 0 ,+1 t+2
i
�����������
!30 /:.-1957
- � �
womelmonan�
el��������
0101.10�11�111��
16.V11.
g. vii.
22.V11
2? VII.
6'. VII.
29. VIII
25.t'.
_
-J-'-1 01+2
��������=,
1.a/IIMMENN�
0.
.1����������
30. IX.1958
22. X.
24.X.
2? X.
4. XI/.
13.X/I.
15. XII.
1? XII.
9.11959
25.1.
11.1/.
11.//.
22.11.
24./I.
9./V.
23.1V
4.V
?V.
97.v.
24.V.
11.1/1.
29.V1.
11.V//.
`h7VIT
avid.
20.1X.
29.X.
1.2?:
5.n.
-3-2-1 1 ,r-0-7r1+2 XI/
$�����������������
Fig.5a. t2iie ist::,i1),..:tion of occ:vrence of larger chromos,pheric flares
(denotE:d horizontal abscissa on correspondin,-,: c:ay) befor (from
-3 to -1 and after ssc (from 0 to +2 day), t distances from
10� from C - in IGY and IGC - periods f greater solar
activity. Data of stron; storms are denoted by larger letters.
Approved for Release: 2017/09/11 C06131786_
Approved for Release: 2017/09/11 C06131786
,012
- -
I �
10.I.1960
i
1
1
131.
--).--. --i
2./V
5./V
6./VT.
2/1/. .
!
10.1V
I
271V
30.1V
� el 1/
F
c). ;)�
1
11.7
i
16. V
28.V.
3.71.
4
4
_
27/.
29. V/.
30.V/.
,
0,
14.1/11.
g.VII.
.16.
,
,
4
V/il.
19. Vii'!.
1
29.0,1.
)
2.1X.

/
--)., � i/ : �
.',1%,e
,
C
4. I)C. .
4'..
____1
24.;.
113.
2
Xf�
2
laZt
!
1
21. XI
.
, -'A J, A 01
..�1-1
+ +
� �
+1 +
30.
XII.
25. XII.
3.11.1961
4.1/. ,
411.
16.11.
g 1/1.
Pk
lig
a /V
12.V
2.V.
a Vg
giC
O. ix.
0./X.
a X
?8.X.
Q /. /962
g/.
4A
2A.
3./1
(2/V
'5./V.
4. V
VI/
Fig. 5b. Time distribution of occurrence of larger chromospheric flares
= 2 ( denoted by horizontal abscissa on corresponding day), be-
fore ( from -3 to-1 day) and after sac ( from 0 to +2 day), at
distances from 0� to 100 from CM - periods of smaller solar
activity. Data of strong storms are denoted by larger letters.
�
__Approved for Release: 2017/09/11 C06131786
� �r,'"`r
Approved for Release: 2017/09/11 C06131,786
The second part of Tab. II cres the number of flares occurring
in the interval of one to thf.ee days before the geomagnetic storm �
(large storms and flares 7E-5, 2 and --;];: 1-) with the number of
flares not occurring until after the commencement of the storm up to the
+2 day. The restlt is very instructive. The number of flares preceding.
is somewhat higher so that Lie difference in:the two numbers gives
positive numbers except for 1961. For medium and large flares it is.�
zero in 1960. It is seen again, primarily for medium and large flares,
that the value of the difference decreases with ,a decrease in the to-
tal number of flares. However, despite everything, there exists on
the average a sort of predominance of the number of flares occurring.tari
before compared with the number of flares occurring after the commence-
ment of a geomagnetic storm. In order to determine this more exactly for
a larger number of cases, a.similar investigation was made for a set of
all sac. The results are plotted in Fig. 5 a and 5 b ( all storms with,.
ssc, apport from one large storel on 30.111. 1960, which had gradual
commencemcnt),It is seen that here, too, flares occur both before
and after the commencement of a storm.
T
however, add up
,
all the flares we obtain z1 higher number before storms (73)
than after storms (21) .
�
in addition, we consider that
this graph gives a high percentage of geomagnetic storms notirpreceded by
.any flares (68%) and simultaneously also a certain percentage when
flares occurred only after the commencement of the storm(7%) then it
be deduced that if there eists a connection betweengeomagnetic storms
with sec and flares it must be oite limited and indirect.
from
4) What\this interdependence may have can be determined as
follows. It was shown earlier that a geomagnetic storm occurred after
a flare if an unstable filament or active centre with unstable filaments,
surges etc., were simultaneously on the C or even better in the centre
of the solar disc 1-201 12-27...Since the opinion is still held that large
);
flares are after all connected with geomagnetic storms, an investigation
of the solar situations during the occurrence of large flares was made
Approved for Release: 2017/09/11 C06131786,...iy,i,,;�4�_,,,