5
SOME CHARACTERISTICS OF THE DISPERSION OF WHISTLERS
Atsushi
KIMPARA
Abstract.
The author studied some characteristics of dispersion (D) of whistlers for 4 years
since the beginning of IGY, and found that in diurnal variation D decreases in general
from sunset to sunrise and increases thereafter. This tendency is observed very clearly
in winter but not in summer, and it becomes ambiguous in accordance with the decrease
of solar activity. At Toyokawa D is minimum in winter and maximum in summer, while
at Wakkanai it is minimum in winter and summer and maximum in spring and autumn,
contradictory to Helliwell's report, suggesting a geomagnetic latitude dependence.
As to
solar activity, D correlates with the s unspot number with time lag of 1 to 2 months.
Investigating D individually, the author also found that on storm days the diurnal variation
of D is exaggerated.
1.
Introduction
As ts generally known, the dispersion of whistlers is given by
D
1 J
fp
=--zc
path, fH-ds
where
fp = plasma frequency = 9 v Nkc
fH = electronic gyro·frequency = 2800B kc
B = flux density of earth magnetic field in gauss
Therefore, if the path of propagation and gyrofrequency variation are known and a form
for the electron distribution is assumed, the electron density can be determined. It should
be noted that the effect of a given electron density on the time delay becomes greater
as the gyrofrequency is reduced. Thus the dispersion of a whistler is most sensitive to
regions near the top of the path. <9>
The investigation of characteristics of the dispersion (D) at calm and disturbed periods
w ill be very useful to the study of geophysical phenomena such as geomagnetic storm,
aurora, cosmic rays, etc. as well as to the applications to radio communications in the
exosphere.
At XIIIth Genera l Assembly of URSI at London (1960) (2) the resolution was adopted,
for effective latitude coverage, t o observe dispersion at every 5° or 10° in geomagnetic
latitude all over the world, and our Toyokawa and Wakkanai observatories were recom·
mended as responsible stations at 25° and 35° .
6
2.
Diurnal variation of disersion (D)
Investigating "dispersion., observed at Toyokawa and at Wakkanai for 4 years since
the beginning of IGY in July 1957, we found that the monthly average of its diurnal
variation can be classified in 3 types, i.e. the summer type, the spring-autumn type and
the winter type, and its characteristics vary slowly in accordance with the solar activity
from 1957 to 1961 as shown in Figs. 1-4 (Toyokawa) and Figs. 5-15 (Wakkanai) .
The dispersion (D) in general decreases slowly from sunset to sunrise and increases
again thereafter. We can find this tendency very clearly in winter, fa irly well in spring
and autumn and rather poorly in winter, and in accordance with the decay of solar activity
this tendency becomes ambiguous. In average, D at 18 -24 in JST (9-15 UT) is larger
than that at 0-6 in JST (15-21 UT), and that at 6- 12 in JST (21-3 UT) is much larger.
70
u 60
( 2 l ..
QJ
"'c
~
50
"'....
40
QJ
0.
"'
Q
"
"'~
c
0
30
121
( I)
20
0
2
4
6
8
10
Time in
12
lA
16
18
20
22
u'f
Diu1nal variation of D at Toyokawa in September 1957
(1) a nd April 1958 (2)
Fig. 1.
60
...,,
u
50
QJ
"'c
c
0
.,0."'....
"'
40
30
20
Q
10
_L__j__lj
('l
?
4
n
R
10
12
14
lb
18
20
22
Time in UT.
Fig. 2.
Diurnal variation of D at Toyokawa in November 1957 (1),
December 1957 (2), ] anuary 1958 (3) and February 1958 (4)
7
70
.1\.
..,.
.,J
.e"'
60
c::
40
.,"'
.30
i5
20
.2
k
"'""
\
I
141.
_.'J
50
10
0
2
8
4
\
\
10
12
14
16
18
20
22
Time in UT.
Fig. 3.
Diurnal variation of D at Toyokawa in November 1958 (1) ,
Januar y 1959 (2), February 1959 (3) and March 1959 (4)
70
60
'"""'.,
" so
"'
(41
.e
c::
40
.,"'
30
.2
k
""
i5"'
t.ZJ-·- -- ----
<31
20
8
10
12
14
1t1
L.v
22
Time in UT.
F ig. 4.
Diurnal variation of D at Toyokawa in November 1959 (1),
December 1959 (2), J anuary 1960 (3) a nd Februar y 1960 (4)
00
"'-".
70
.e"'
60
.,"
0= so
.,
·;;;
k
""
i5"'
40
~
0
8
10
12
14
16
18
20
22
Time in UT.
Fig. 5.
Diurnal variation of D at Wakkanai in Jul y 1957 (1), August
1957 (2) , September 1957 (3) and October 1957 (4)
90
"""u'
...
"'d
d
0
.."'...
"'""
00
70
60
...
\
\-·-·
/\
50
I
i5
(ii.
I>)
I))
<31
6
8
10
12
14
16
ill
20
22
Time in UT.
Fig. 6.
Diurnal variation of D at Wakkanai in November 1957 {1),
December 1957 (2) , January 1958 (3) and February 1958 {4)
90
..,....
...
80
u
/1} • • • • ~· . . .•.
"'
.s
70
d
.~
60
....."'
50
i5
40
"'""
I
I
141
·2
12
14
16
18
20
22
Time in UT.
Fig. 7.
Diurnal variation of D at Wakkanai in April 1958 {1) , May
1958 {2), July 1958 {3) and August 1958 (4)
)"J
ijiJ
...J
"'d
d
0
....."'
""
i5"'
7')
rh
·o
.1
,,
t; ()
~u
(j
?.
4
r,
8
10
<2
h
tfl
20
22
Time in UT.
Fig. 8.
Diurnal variation of D at Wakkanai in November 1958 {1) ,
December 1958 {2) , January 1959 {3) and February 1959 (4)
9
r•
J'
"""'
"
"'
J::
'\
8(J
QJ
J::
0
'
fN
"'....
-... __ j
/'.
i
QJ
c. rr.,
"'
0
t.J
,,
~
"
·0
.~
r.
Time in UT.
F ig. 9.
Diurnal variation of D at Wakkanai in March 1959 (1),
April 1959 (2) and May 1959 (3)
')f'
"-""
""'
f{J
..
QJ
c
c
0
·o
A
-
(£.,
i:S"'
~
sc
\
'Zl
'. \
'
I "--~~-
.-
Ul
....
QJ
c.
1'. ; \ "''
< .. -'I
UJ
1,1
\'<>/
ll'l
' I
v
<r
.;
j
·t,
Time in UT.
Fig. 10.
D iurnal variation of D at Wakkanai in June 1959 (1}, July
1959 (2), August 1959 (3) a nd September (4)
80
,, ~ ,
~"'
t.i
QJ
1;.)
70
\
Ul
c
c
f:/)
0
"'.... so
QJ
c.
"'
i:S
40
tJJ
}()
0
Fig . 11.
2
4
"'
8
10
'4
;ij
1-0
//
Time in UT.
Diurnal variation of D a t Wakkanai in November 1959 (1} ,
December 1959 (2}, January 1960 (3) and February 1960 (4}
10
..,,
eo
.,t.i
Cll
.5
c:
.."'c.
.~
Cll
i5"'
4
16
10
6
1&
20
22
Time in UT.
Fig. 12.
...,.
<.>
.,
70
Cll
60
c:
c:
0
50
...
40
·;;;
Cll
Diurnal variation o f D at Wakkanai in March 1960 (1), April
1960 (2) and May 1960 (3)
.
f3~;-~-.:
, .. _ ____ _•• (3 )
"'
,,,
_ .. <1)
C2)
c.
i5"'
30
0
2
4
c
8
10
12
14
IG
18
20
22
Time in UT.
Fig. 13.
Diurnal variation of D at Wakkanai in June 1960 (1), July
1960 (2) and August 1960 (3)
70
...,.
<.>
60
Cll
"'
.5
c:
.."'
-~
Cll
.,c.
a
30
I
20
0
2
4
6
8
oO
12
14
IG
18
20
22
Time in UT.
Fig. 14.
Diurnal variation of D at Wakkanai in October 1960 (1),
November 1960 (2), and December 1960 (3)
11
70.------------------------------------------,
I'!
0
40
12)
~ 30
Ill
.,..."'
i5
fl
10
12
14
Time in UT.
Fig. 15.
3.
Diurnal variation of D at Wakkanai in January 1961 (1), February
1961 (2) and March 1961 (3)
Seasonal variation of dispersion
According to H elliwell, dispersion is mm1mum at the June solstice and maximum at
the December solstice, and shows the same variation in both hemispheres. The expla·
nation of this remarkable circumbstance is not clear, but may be related to the eccentricity
of the earth's orbit about the sun or to the seasonal asymm etry in the relation between
the sun-ear th line and the geomagnetic equator. 0)
In our observation at Toyokawa dispersion is minimum in winter (N ov.-Dec., some·
times from Oct. or to March) and m aximum in summer (June-July, sometimes from May
or to Aug. and Sep.), showing the inverse results wi th those of Helliwell, and at
Wakkanai minimum in winter and summer and maximum in spring and autumn (Fig. 16,
a. b., Fig. 17, a. b. ) . Therefore, it seems that the seasonal variation of dispersion depends
on geomagnetic latitude; the conclusion should, however, be postponed to the future
s tudy.
90~----------------------------------------~
....,,
.,"
.e"'
I'!
.~
so
Q)
<10
..."'
0.
"'
Q
·. ;fi'/
( 21
30
20
7 8
1957
10
12
2
1958
4
B
Month
10
12
1959
F'g. 16. a . Seasonal variation of D at Wakkanai (1), Toyokawa (2) and
sunspot number (3)
12
~----------------------------------~
....
Q)
.0
e
c::
c::
:::1
c::
.~
0
c.
"'....
c.
i5"'
"':::1c::
Q)
(/)
20
A
10
10
12
12
2
4
6
M onth
1959
1960
1961
Seasonal variation of D at Wakk anai (1), Toyokawa (2)
and sunspot number (3)
6
Fig. 16. b.
""".;
Q)
"'c::
c::
0
"'....
Q)
c.
i5"'
<I I
40
707L-8~~,o~~
,2~~2~~4~~6~~
8~-,~o~-,~2~-2~L-4L-~s 20
Fig. 17. a.
1957
1958
Month
1959
Season al variation of D of long whistlers at Wakkanai (1)
and sunspot number (2)
280
200
......
80
.;
Q)
200
"'c::
a
c::
.~
"'....Q)
c.
i5"'
.,....
s
.0
~
IQO
00.
..
------
"'
?.!
100
:;
rn
70
I()
4
10
12
B
I?
2
G
2
8
Month
1959
1960
1961
Fig. 17. b . Seasonal variation of D of long whistlers at Wakkanai (1)
and s unspot number (2)
6
13
4.
Geomagnetic activity and dispersion
OutsuOl investigated statistically the correlation of disp ersion with geomagnetic storms
and found that th e average dispersions on storm days are s maller than those on the days
before and after, both in daytime and at night. We inves tigated in detail the c haracteri s tics of geomagnetic storms individually and compared them with the variation of
dis persion, in every phase of s torms. Although there are many data of s torms and di spersions, rather few data are found to be observed simultaneously.
No. ). 4 September 1957.
According to data at Ka kioka Ge omagnetic Observatory,
the storm began a t 13.0 UT with SSe and the horizontal component (H) increased + 48r,
at 14 .4 UT it entered the main phase and a t 18 .0 UT the las t phase. The dispersion
decreased gradua lly from the beginning to the end. On 7 September, 3 days afterwards,
the diurnal variation of D indica ted a very s mall variation, i .e. th e summer type variation
(Fig. 18) .
No. 2. 21 September 1957.
According to Kakioka , the storm began at 10.05 UT
with SSe* and H increased + 79r, at 10.5 UT it entered its main phase (K index 7) and
at 01.0 UT on 22 September its last phase. D decreased gradually, too, from beginning
to the end. On 26 September, 5 days later, D indicated a very s mall variation , returning
to the ordinary tendency in s ummer (Fig. 19).
No. 3. 14 February 1959.
According to Ka kioka , th e storm began at 11.43 UT
with SSe and H increased + 12r/2m. Though it is not clea r when the main and the last
phases began, the s torm ended at 21.0 UT, and the maximum range was 99r. D decreased
suddenly at the beginning, but afterward s it indicated small variations, some times s howing little increases.
Observation of D on days near the storm a re lacking, but we find
that it indicates about the similar tendency with the monthly mean of F ebruary 1959 and
it is characterized by a sudden decrease at th e beginning (Fig. 20).
No. 4. 5 December 1959.
According to Kakioka th e storm began at 06.59 UT with
SSe and H increased +29r. At 10.5 UT it entered the main phase, at 17.0 UT th e last
70
....':'1
u
..,
"'c
00
~
c
0
..,...
(I)
40
"'
(I)
30
0
20
0
2
4
6
8
10
12
14
16
18
20
T ime in UT
F ig. 18.
Va1iation o f D on storm day a nd on the days befo re and a fter:
D on storm day (1) on 4th September 1957, 6th (2) and 71 h (3)
September 1957.
14
(:.)
.....
.,
,_;
70
"'
·=
c:
:,0
0
so
.,0."'....
"'
~
Q
I _I 1
'Jij
;:
l
e
6
4
_l _l _l _l
10
1;
i8
14
iO
Time in UT.
Fig. 19.
Variation of D on storm day and on the days before a nd after:
D on storm d ay (1) on 21 th September 1957, 25th (2) and 26th
(3) September 1957.
q)
-:N
.,u
80
"'
7f)
·=c:
.~
.,"'....
f:l)
0.
"' so
.
i5
40
0
?
<1
I
I
r.
R
I
I
10
I
I
I
14
12
T im e in UT
Fig. 20.
~-__!- .; _L __
. - ··-·
"'
IR
22
.:0
Variation of D on storm day and o n the days before a nd after:
D on storm day (1) on 14th F ebruary 1959, mean variation o f
D in Februa ry 1959 (2)
80
'"'".,u
"'
, ·- -·v '
70
·=c:
60
.,"'....
0.
50
/·
/'
..
:..··
(''
0
"'
Q
'
14)
0
:;,
4
6
8
10
,;,
Time in UT.
Fig. 21.
I
\j
40
30
-·
':•
•3>
14
'"
'J
~('
;>:;
Variation o f D on storm day and on the days b efore and afte r:
D on storm day (1) on 5 th Decembe r 1959, 12th (2) and 13th
(3) December 1959.
15
phase and at 22.0 UT ended, indicating the maximum range of 156r. D oscillates
remarkably and after the last phase it increases, showing a difference with those on calm
days (Fig. 21) .
No. 5. 15 November 1960.
According to Kakioka, the storm began at 13.03 UT
with SSC and H increased +31r/3m. At 16.8 UT it entered the main phase and at
02.9 UT on 16 November the last phase, ending at 03.0 UT on 18 November. The
maximum range of H is 225r.
D decreases at the main and the last phases respectively
and indicates oscillatory characteristics as a whole. This tendency can also be seen on
14 and 16 November, latter, however, showing no oscillations (Fig. 22).
No. 6. 15 March 1960.
According to Kakioka, the storm began at about 12.00 UT
with Sg and it is not clear when it entered the main and the last phases, but it ended
at 17.0 UT and the maximum range of H is 138r.
D decreased gradually, but it has no
peculiality in comparison with the monthly mean of March 1960 (Fig. 23).
...,.
.,"
.e
70
60
CIJ
, .- .
'
())
c::
-~
CIJ
...,
........"<-'·· · ·12)
""
i5
.-"'
.
.-'\
/"
CIJ
2
4
6
8
10
12
14
I()
10
20
22
Time in UT.
Fig. 22.
...,.
.,ti
Variation of D on storm day and on the days before and after:
D on storm day (1) on 15th November 1960, 14th (2) and 16th
(3) November 1960.
80
,,, - '"""-•3'-.. .. ..... ·:·· -~ -.··'2'
70
CIJ
.e
c::
~0
· · · - -........_
0
.,....
CIJ
""
CIJ
i5
so
·· r ··
.
I
\
.
;
\
....
·,
c•
13)
\1
40
<3>
30
I)
2
4
6
8
Time in UT.
Fig. 23.
Variation of D on storm day and on the days before and after:
D on storm day (1) on 15th March 1960, 14th (2) and 16th
(3) March 1960.
16
These examples show that, when a storm begins with SSC or SSC*, the tendency
of D to decrease from sunset to sunrise is exaggerated during the storm in comparison
with calm days around storm days. It seems that there is some correlation between a
geomagnetic storm and the decrease of D observed at lower geomagnetic latitude, and
accordingly the variation of H in the exosphere, but it is not yet conclusive.
5. Solar activity and dispersion
Allcock, MorganC6l and Kimpara<4l(Sl investigated independently the dispersion in
detail and found that the dispersion correlates with the sunspot number with a time-lag
of one to 2 months. Here I examined again the dispersion for 4 years from the begin·
ning of IGY and found that the correlation of dispersion with the sunspot number indicates also about the same tendency as shown in the following table- the correlation
coefficient is + 63.0% for dispersion one month later.
D
6.
Correlation coefficient of D
with sunspot number
Same month
55.5%
1 month later
63.0%
2 months later
52.0%
Conclusion
Since the beginning of IGY 1957 we have observed whistlers for 4 years and obtained
many data on the dispersion of whistlers, which is considered to be very useful to the
study of geophysics and to the application to communications in the exosphere. We
studied, first of all, the diurnal and the seasonal variations of dispersion and found that
the dispersion decreases, in genera l, gradually from sunset to s unrise and increases again
thereafter. This tendency is very defi nite in winter and not so very in summer, while
in spring and autumn it lies between th e tendensies in summer and winter.
As to the seasonal variation we have found that at Toyokawa the dispersion is minimum in winter and maximum in summer contradictory to Helliwell's report, while at
Wakkanai it is minimum in winter and summer and maximum in spring and autumn.
This tendency suggests us the existence of geomagnetic latitude dependence.
We have long had deep interests in the correlation of geomagnetic activity with
whistlers, especially with dispersion, and have reported on it elsewhere. <4H5J(7)(8) This
time we s tudied whistlers not statistically but individually in detail and found that, when
a s torm begins with SSC or SSC*, the tendency of D to decrease from sunset to sunrise
is exaggerated during the storm when compared with calm days around storm days.
17
The correlation of solar activity with dispersion about which we reportedW<s> recently
was confirmed by the 4 years' data of IGY, i.e. the dispersion correlates with the sunspot
number with a time lag of one to 2 months.
7.
Acknowledgements
The work described above was carried out as part of the Japanese IGY and IGC
programme in the Research Institute of Atmospherics, Nagoya University.
I wish
gratefu lly to acknowledge incessant encouragements and advices by Dr. M. Shibusawa,
ex-President, Dr. S. Matsusaka, President, Dr. Hasegawa and Dr. Hagihara. I should
like also to acknowledge the assistance of my colleagues, Messrs. Iwai and Outsu, in
discussion and observation. Finally I appreciate very much the faithful assistance of
Miss Yabuta in preparation of this paper.
References
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Helliwell, R. A.: J . Res. NBS, 64D , 642 (1960)
Xlllth General Assembly, URSI, Com. IV, Resolution 4 (1960)
Outsu, J. and Iwai, A.: Preprint of Inter. Conf. Cos. Rays Earth Storm, Kyoto (1961)
Kimpara, A.: Rep. Iono. Space Res. Japan, 14, 160 (1960)
Kimpara, A.: Proc. Res. Inst. Atmospherics, Nagoya Univ., 7, 40 (1960)
Allcock, G. Mak. and Morgan, M. G.: J. Geophy. Res., 63 , 573 (1958)
Kimpara, A.: Comptes r endus, 250, 2596 (1960)
Kimpara, A.: Nature, 186, ::30 (1960)