• TABLE OF CONTENTS
HIDE
 Title Page
 Acknowledgement
 Table of Contents
 List of Tables
 List of Illustrations
 History of study of decameter-wavelength...
 A radiation source model
 Spectral studies using multi-channel...
 A swept-frequency decameter-wavelength...
 Spectrum of bursts
 Summary
 Reference
 Appendix
 Biographical sketch
 Copyright














Title: Spectral characteristics of the radio-frequency outbursts of the planet Jupiter.
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00091590/00001
 Material Information
Title: Spectral characteristics of the radio-frequency outbursts of the planet Jupiter.
Series Title: Spectral characteristics of the radio-frequency outbursts of the planet Jupiter.
Physical Description: Book
Creator: Chatterton, Neil Ellis,
 Record Information
Bibliographic ID: UF00091590
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000565579
oclc - 13554738

Downloads

This item has the following downloads:

Binder1 ( PDF )


Table of Contents
    Title Page
        Page i
    Acknowledgement
        Page ii
        Page iii
    Table of Contents
        Page iv
        Page v
    List of Tables
        Page vi
    List of Illustrations
        Page vii
        Page viii
        Page ix
        Page x
    History of study of decameter-wavelength radiation from Jupiter
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
    A radiation source model
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
    Spectral studies using multi-channel data
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
    A swept-frequency decameter-wavelength receiving system
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
    Spectrum of bursts
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
    Summary
        Page 125
        Page 126
        Page 127
    Reference
        Page 128
        Page 129
        Page 130
        Page 131
    Appendix
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
    Biographical sketch
        Page 140
        Page 141
    Copyright
        Copyright
Full Text











SPECTRAL CHARACTERISTICS

OF THE RADIO-FREQUENCY OUTBURSTS

OF THE PLANET JUPITER










By
NEIL ELLIS CHATTERTON


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY









UNIVERSITY OF FLORIDA
June, 1961












ACKNO7LEDGZ"MENTS

The author is most deeply indebted to the Chairman of

his Graduate Committee, Dr. A. G. Smith. He encouraged the

author in every phase of this activity, and without his

advice the task would not have been accomplished. To Dr.

T. D. Carr, also on the committee, belongs much of the credit

for arousing the interest of the author in this project. To

the other three members of the Graduate Committee, Drs, R, W.

Cowan, J. S. Faulkner, and G. C. Omer, the author is indebted

for their interest and encouragement.

The collection of data was a large part of the effort

in completing this project. Dr. Federico Rutllant, director

of the National Astronomical Observatory of the University of

Chile, was instrumental in establishment of the Chile station.

Without the support of H, Bollhagen and members of his staff,

who maintained watch in Chile, and of II. F Six, Jr., John

E*. White, and Richard Flagg, who helped maintain watch in

Florida, the data used here would not have been available.

Ralph B. 'Jarren, Shop Pore;nan, G0 C. Garris, C.

Van Veldhuison, and other members of the shop staff helped

with construction and maintenance of equipment.

The preparation of this manuscript would have been

impossible without assistance from my fellow students,










John E. White and N. F. Six, Jr., and from Hans W. Schrader

and Paul L. Watkins, who reproduced the series of spectral

photographs included here. The perseverance of the typist,

Mrs. V. Carolyn Roberts, who spent long hot hours over the

typewriter, is much appreciated.

This work has been supported by the National Science

Foundation, the U. S. Army Research Office (Durham), and the

Office of Naval Research.


iii












TABLE OF CONTEETS

Page

ACKNOWLEDGEMENTS . . . . . . . ii

LIST OF TABLES . . . . . . vi

LIST OF ILLUSTRATIONS. . . . . vii

Chapter

I, HISTORY OF STUDY OF DECAnETER-'IAVELENGTH
RADIATION FROM JUPITER . . . 1

A. Observational Record .. . . 1
B. Radio Noise Features .. .. . 7
II. A RADIATION SOURCE MODEL. .. . . 15

A, Summary of Theories of Source of
Radiation .. . . . 15
B. Assumed Electron Densities . 18
G. Synchrotron Radiation Model. . . 22
D. Application of the Model . . . 30

III* SPECTRAL STUDIES USING IULTI-CHANNEL DATA 32

A. Equipment Utilized . . . . 32
B. Observing Procedure. . . . 33
C. The Small Frequency-Difference
Records . . . . . 34
D. Observations of Onset and Termination
of Storms . . .. . . 39

IV. A SWEPT-FREQUENCY DECAMETER-.'AVELENGTH
RECEIVING SYSTEM . . . . 50

V. SPECTRUM OF BURSTS. .. . . . 63

A. Description of Noise Pulses, . 63
B. Analysis of Peaking. . . . 67
C, A Working Model. . . . . 73
D. Conjectural Note . .* . . 80








Page
vi. SUWIaR . . . . . . . 125

LIST OF REFERENCES . . 128

APPENDIX A. ANGULAR DISTRIBUTION OF SYNCHROTRON
RADIATION . . . . . . . 132
APPENDIX B. FUNCTIONS ASSOCIATED' WITH SYNCHROTRON
CALCULATIONS . . . . . . . 135

APPENDIX C. PARAfETERS OF THE LO;C-lOISE AMPLIFIER
AND THE FILTER, ....... 137

BIOGRAPHICAL SKETCH. r . . 140












LIST OF TABLES


Table Page
1. COiE OF ESCAPE OF RADIATION. . . 11
2. CORRELATION BET";I7T FREQUrENCY<.SEPARATED PULSES 38

3* DRIPT RATES OF ONSET AND ENDING OF JUPITER
NIOISE STORIS. . . . . . . . 4 3

4, OCCURRENCES IN 1960 AND 1961 IN WHICH ONSET
AND ENDING OF STORM WERE IN OPPOSITE
DIRECTION . . . .* .. . 45

5. PEAK SEPARATION AND DRIFT APRIL 18, 1961 . 68
6. PEAK SEPARATION AND RIFT APRIL 25, 1961 69

7. RADIO NOISE RECEPTION DURING THE PEAK ANALYSIS
P RIODS . . 4 . # . . . 81












LIST OF ILLUSTRATIONS


Figure
1. Probability of Occurrence of Jupiter
Radiation Versus Frequency. .. . *

2. Probability of Occurrence Histogram for
System III Longitude Chile 1960 .
3. Electron Motion Coordinate System Used in
Synchrotron Radiation Derivation.* ..
4. Frequency-Separated Record 0.1 Mc/s.*

5# Example of Increase in Ratio with Time of
Amplitudes of Frequency-Separated Pulses.

6. Example of Decrease in Ratio with Time of
Amplitudes of Frequoncy-Soparated Pulses.

7. Drift Tendency of Onset of Noise Storms,*
8, Drift Tendency of Termination of Noise
Storms* . * * . * *

9. Representation of Wide-Band Receiving
Systems* 9 > * * * * 0 *

10. Spectral Distribution Recording
Configuration . . . . q

1ii Rhombic Antenna Design Parameters. *
12, View of Final Configuration of Rhombic
Antenna Showing One Side and Output
Termination . . .

13. Balanced Input Wide-Band Single Stage
Pream plifier . .* *

14. Block Diagram of Panoramic Roceiver, . .

15. Frequency Drift of Amplitude Peaks of
Noise Pulses, . . #. . .


Page


, 8

S 10


* 24
S 36

S 40

41

* L47

48
1 t.8s


*


vii










Figure

16. Brush Record of Time of Peak Crossing at
18.00 Mc/s. . . . . . . . .


17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.


Series

Series

Series

Series

Series
Series

Series

Series

Series

Series

Series

Series

Series

Series

Series


April 9, 1961. 0435:30 E.S.T.,


1.

1,
2.

3.

3,

4.
5.
6.

6,

6,

7.

8.

9.

10.

11


* 0

1961.

1961.

1961.


9 . . .

0440:20 E.S.T,

0520:50 E.S.T,

0441:40 E.S.T.


page 2 .

April 27,

April 25,

page 2.
April 27,

April 18,

April 27,

page 2. .

page 3 *
April 27,

April 27,

April 27,


. *

1961.

1961.

1961.


April 27, 1961.

April 27, 1961.


. 9 ,

0445:45 E.S.T.

0440:25 E.S.T.

0445:00 E.S.T.

0o44:40 B.S.T

0438?05 E3.S.T,


32. Series 12 and Series 13. April 25, 1961.
0450:50 and 0451:35 E.S.T. Single Frames, .

33. Series 14. April 27, 1961. 0445:15 E.S.T..

34. Series 15. April 25, 1961. 0449:30 E.S.T..
Strong station at 1.1 Mc/s . . .

35. Series 15, page 2. . . . .
3b. Series 16. April 25, 1961, 04100 05 E.S.T.

37. Series 17. April 18, 1961. 0520:55 E.S.T..


99

100

101
102

103

104


viii


Page


. 4 0 *. 4 0 .


S 84

. 85
S 86

* 87
. 88

* 89

90

S 91

S 92

* 93

9 94

S 95
S 96

97

S 98


I
t


1961. 0439:00 E.S.T.

1961, 0439:20 E.S.T.











Figure
38. Series 17, page 2. . . . . . .
39. Series 18. April 18, 1961. 0521:05 E.S.T.
Series represents interval of 18 3/4
seconds .* * . . . . * *
40. Series 18, page 2. . . . . . .

41. Series 18, page 3. . . . . . .
42. Series 18, page 4. .* . . .
43. Series 18, page 5. *. .* *. . . .
44. Series 19. April 25, 1961. 0441:35 ES,T.
This is the first example of spectra at
1 Mo/s sweep width. *. . .. . .


Series 20. April 18, 1961.
Series 20, page 2. *. . .
Series 21. April 18, 1961.
Series 21, page 2. . ..
Series 22. April 25, 1961.
Series 23. April 25, 1961.
Series 23, page 2. .. # .
Series 24. April 25, 1961,
This series represents 22.2
at 1 Mc/s sweep width .
Series 24, page 2.. . .
Series 24, page 3. . .
Series 24, page 4. *


0529to4 E.S.T..


0527:40 E.S.T..


0440:50 E.S.T..
0440:25 E.S.T..


0443:35 E.s.T*
seconds shot
* .4 4



3 * 4
3 4 3 .* 44


Page
10l


106
107
108

109
110


111


4$.
46.
47.
48.
49.
50.
51.

52.


53.
54.
55.


112

113
114
115
116
117
118


119
120
121
122









Figure Page
56. Series 24, page 5 ... 123
57- Series 24p page 6 .. ... * . 124
58. Section of High-Pass Filter Used in Spectral
Ronording System -, 139












CHAPTER I


History of Study of Decameter-V'nvclcnrth Radiation from
Jupiter


A. Observational record

Radiation from Jupiter in the deccaoter-wavolongth

region vas first observed, accidentally, by Burke and
Franklin (1), using an intrerferometer-typerecciving sys-

tem at a frequency of 22.2 Mc/s in 1955.

This discovery prompted Shain (2, 3) in Australia to

revicr records of observations mado at 18.3 Mc/s in 1990-

51. :vents which previously had been believed to be inter-
ference of terrestrial origin were found to posccss the

same properties as theo events described by the original dis-
coverers of the Jovian radiation. That is, from a first

series of observations using a narrow beamwidth antenna, the

location of the noise source was found to be in the same dir-

ection as Jupiter to within fL, the noise won sporadic and
vas of high intensity at its iraximum. Sufficient data were

available from the records of Burke and Franklin and Shain

to notice a significant feature of this radiation. When
the activity wro plotted as a function of the longitude of

the central meridian of Jupiter as seen from the earth,

a striking: correlation betwocn activity and longitude




2


was noted when the proper longitude system was used.
When one views the planet Jupiter, it is possible

to observe markings on the dish. Observations over a

period of time indicate an approximately constant period

of rotation for some of these features, which scoem to be

carried alonG by a rotation of the solid nass of the

planet. lHoovcr, features at various latitudes differ

in t-eir apparent rotational periods. The variation in

rotational period across the face of the planet indicates

that the observed features are atmospheric in nature.

Because of the large differential in rotational period

between the equatorial regions rnd the temperate zone

and polar regions, two systems of longitude have been

set up for the visual features of Jupiter, designated
System I for latitudes within 100 of the equator and

System II for markings beyond these limits. There are
some exceptions for particular uarkings beyond the

System I limits, vhich seem to exhibit a rotational

period closer to Systori I than to System II (4).
Using these two systems of longitude, the radio-

frequency activity described above was plotted by Shain

in both Systess and IT. No correlation was observed

for -.ystem I, but the radiation sooucd to be a strong

function of System II longitude.* %hain noted that the

source of noise activity drifted at a small rate with

respect to the System II coordinates, so he proposed a

new rotational ystc::-. producing a bettor corrolotion

botuoon longitude and radio-frequency activity. The







radio-frequoncy longitude system has subsequently been
dcsignatod "System III",'
The initial discovery of decameter-wavelength radi-

ation from Jupiter provoked a large-scale investiUrtion
by many iworkors. F. G. Smith ($), in Tnglid, attempted
to eoitcnd '0brvations to frequencies of 38 Tc/s and 81.5
Mc/s. Shain had noted that the peadn povror of the radio
noise picked up by him was in excr-os of 5 X 102 watts/

m /cps at 18-3 Mc/s. Smith's apparatus was sensotive
to powers of 10"24 and 3 X 10-26 Vatts/m /cpe, respec-
tively, for the two fr6quonclcb abovo. To Vobs rvntions
of radiation were noted at these frequencies. However,
as will be seen later, typical noise bursts are of the
order of one second in duration, and Fnitb used a six-
second time constant, so that some doubt rcmni:nr as to
whether radiation near this power level may be observed
at those frequencies. Douglas (6) reports that Reber
made observations of Jupiter in 1953 at 30 TO/s and re-
ceived radiation on ten of tro hundred observing nights.
Franklin and Eurkcr: continued their observations at 22.2
I.c/s -ith a polarilmtor. The records indicated that the
radiation was nearly always right-handed elliptically
polarized, and was sometimes nearly circularly polarized
(7). Wells, an nosociaft of Franklin and nurko, observed
lG0 radiation oimultancously at 18 Mc/s and 27 Mc/s and
found that the radiation did not occur at both frequencies
at the same timo (7). Kraus (8) observed Jupiter at a







frequency of 26,6 1Ic/s and reported on the shape of in-

dividual pulses. One type of pulse n-pponrcd to be of

very short time duration, of the order of 10 milliseconds

or loss. The pulses occurred separately and in groups

of two and three# The time interval between pulses in

a group seeded to fall into two categories, one time

interval being near one-quarter of a nacond, while the

second, much shorter, time interval was of the order of

one-fortieth of a second, Gallet (9) observed at fre-
quencies of 10 and 20 Mc/s in 1956 and 1957, and reported

that activity at the higher frequency vwas about 0.6
time that at the lower.

The University of Florida Radio Observatory made

its initial observations of Jupiter in the winter of

1956. Theeo first observations creo made at a single
frequency, 18 0.c/s. From this seasonTs observations,

coupled with the work of others in previous years, a

good estimate as to the rotational period of the noise

sources wn maddo (10). This rotational period was de-

termined to be

TX= 7^ 55m 2.8'
To facilitate analysis of records, System III longitude
r:w tnkon as bcin7 coincident with System II lon-*tudo on

January 1, 19957 Previous workers had attempted to

correlate the radio noise storms with visible features.

of Jupiter such as the Great Rod Spot, the Reese white

spots, and the south tropical disturbance, VWhile some

of these workers had noted apparent correlations,







the analysis at this time Crre no indication of an

association between visible features end radio noiso
longitude. Activities at this Obzorvatory wore con-
tinued in 1957-58a and the observing frequencies were
extended to include 18, 22.2, and 27.6 Mc/s. Pola-
rimoter records were obtained at 22.2 M c/s, and again,
the majority of noise bursts appeared to be very
strongly right-handed elliptically polarized. Again,

moro activity was noted on lower frequencies. The

27.6 Mc/s array, a 7-clement Yagi-type antenna, did
not pick up definite Jupiter radio noise, although
this may have boon due, in part, to local interfer-

ence. H. J. Smith and Douglas at Yale began obser-
vations at 18, 21, 22.2, and 23 lc/a during this period.
Activity measured at the two latter frequencies showed
good time correlation between pulses at this small fre-

quency spacing. This laboratory noted a mnrlkod reduc-
tion in over-all activity between 1957 and 1958, and
the Yale observatory also noted the effect, From those
observations, it seemed that solar activity might have
a negative effect upon the measurable radio-frequency activity
of Jupiter, since the sunspot cycle reached its peak in

1958, vhilo Jupiter activity reached a low ebb in 1958
and 1959.
In 1959, a first attcipt was mado at this labora-
tory to determine the approximate vidth of noise pulses
in frequency. This was done by using two receivers set







to nearly the same frequency in this case 18 Mo/s -

and running a high-speed record of the incoming signalrs,

while the frequency separation of the receiverS was varied.
Results indicated a pulse width of the order of one :.c/s.
This method will be discussed more fully in Chapter III

in connection with the 1960 data.
In the sunrer of 1959, this Observatory established
a field station at the University of Chile. During the

1960 observing season, coincident noasuremonts nore mado
on 18 and 2222 Mo/s at those two stations. The two-re-

ceiver method of pulos nnnlysis was continued at the
Chile observatory, while a swept-frequency system was
employed at, the Florida station. Four additional fre-
quoncios were monitored in Chile, 10, 15, 16.7, rrn 20
Hc/s, in addition to 18 and 22,2 Mc/s. The Florida
station continued to make observations at the three fre-

quencies mentioned above. The results obtained from these
rioasurcments will be outlined later.
As was aexpoctod from the solar activity docrense

since 1959, the activity of Jupitor has increased in

1961. This is in keeping with Jovian activity levels
occurring before sunspot maximums, A 5 Tic/s receiving
systeor has been installed in Chile, ':hilo the Florida
station addod a 15 Mc/s antenna and greatly improved
the sweep-frequency system.







B. Radio noise features

Somr. of the features of the rndio noise have been

listed above* The cum of all those features gives a
ratlhr good description of the decameter-wavelength radi-

ation, with only a few cr.p present, such as the lack of

polarinotcr records at many frenuoncies.

The radiation is sporadic in nature. This includes

not only the occurrence of the individual short pulses,

bub also the occurrence of the total "noise storn",

Thces stores havo been observed to last a total of, typi-

cnlly, one to two hours, although numerous much briefer

events are observed. The total probability of occurrence
falls off at high frequencies. Figure 1 gives a plot of
the total probability of occurrence versus frequency for
the years 1958, 1959, and 1960 (11). The first two years

data vero obtnin; at the Florida station alone, while
the last years data is a composite for Florida and Chile.
The probability of occurrence ooees to be a maximumI in

the vicinity of 18 .Ic/s. The radiation, as mentioned

earlier, is associated with a fixed rotational period of
Jupiter. The radiation also seems to be associated with

particular longitudon in this coordinate system, and

radiation fro: a given active longitude is observed over

nuch loss than 1300 of rotation of the planet. Those

last two statencnts indicate that the radiation is direc-

tional in character, which has been the main impetus
which has riven rise to the various theories as to the


















9 0


o 9
1

S.06- 4 \-



S- A FLORIDI958
_j

oC 0 CHILE, 1960
o .02- FLORIDA, 960
S IIA FLORIDA, 1959l l
A FLORIDA,1958

0
8 10 12 14 16 18 20 22 24 26 28
FREQUENCY IN MC







Fig. 1.--Probability of Occurronco of Jupiter Radiation Versus


Frequency (From Reference II)







source of the disturbance* Figure 2 shows some typical

histograms for probability of occurrence vorsus longitude

for data obtained in Chile for 1960. These show several
features of the radiation which have been noted elsewhere

(1, 11, 16). That is, the cone of escape of radiation
as indicated by the width of the probability peaks in
longitude increases for decreasing frequency, there
appears to be a possibility of three separate sources of
the radiation, and the central longitude of the distur-
bance ooeens to decrease with increasing frequency.
Table I lists the cone of'escapo of radiation noted by

various workers at the frequencies indicated.

In addition to the sporadic character noted above,
the radiation does not always occur simultaneously on
all frequencies. Aside from lack of individual pulse
correlation, the activity may be wholly lacI.ing at one
frequency, while another frequency close by is quite
active. Much of this present work will be devoted to
a discussion of the facts known on this subject and
the conclusions that may be reached on the basis of
these facts.
Shain first noted that receiving systems placed a
distance apart, in his case 25 kilometers, did not re-

produce identical noise pulses vhen set at the same
frequency (32). Douglas reported the results of work
done at Yale using a triangular array of receivers tuned
to 22*2 Mc/,s with a ten mile north-south and thirty mile





I10








1960 CHILF-PROBABILITY HISTOGRAM


22.2 mc./sec.


20.0 mc./sec.










18.0 mc./sec.




16.7 mc/sec.


I 10.0 mc./sec.


100 200
SYSTEM III LONGITJDE


Pig. 2.--Probability of Occurrence Histogram for

System III Longitude Chile 1960 (From Reference 11)










TABLE 1
CONE OF ESCAPE OF RADIATION


Year Observer Frequency (Mc/s) Cone


Shain




Burke &s
Franklin



Carr,
A. G. Smith,
etc.
Carr,
A. G. Smith,
etc.


Carr,
A. G. Smith,
oto.


18.3
19.6
27
18,3
22.2

26.75
18


18
,22*2
27.6
10
16.7
18
20
22.2
27.6


135o
500
300
700
600

750
40


45
330
130
1080
54o

.430
360
270
200


1950.51


1956-57


1959




1960


~I-I~- -~----- -- ------







east-west spacing (6). Amplitude distortion between pulses

received at the different sites was prosont, but time of
occurrence of the pulses at the different stations was
correlated. Douglas states that, iJnoring the anrliti-,1io
diotorition, the time correlation wan calculated to be
0.84 On a control run ':ith no Jupiter noise bcini re-

ceived, puloos occurring during this interval shori:d time
correlation of 0.2. Vith the establishment of the Chile
field station, it bocamo possible for the University of
Florida group to maro correlation studies over long

separations. The distance botwoon obsorvihnC points was

70j40 kilomoteos, and the observing frequency was crystal-
controlled to 18.000 Mc/s. The dotailod results of these

observations are published elsewhere (12); A long-period

scintillation, on the order of 30 to 45 second, was

noted, together with an apparent ahorter-poriod scintil-
lation of about the duration of typical Jupiter noise

pulses. Good correlation between individual pulses was
sometimes noticed, indicating that the basic short-duration
pulses are a result of the activity of the source, rather
than a modification in our ionosphere.
The effect of our ionosphere on the polarization of

incoming radiation was also a question that had not been

explored adequately. To the date of the opening of the
Chile station, polarimeter readings in both hemispheres
had indicated predominantly right-handed polarization,

but simultaneous readings had not been taken. These







readings were carried out in 1960 at the two University

of Florida stations at 22.2 Mc/s', with a close agreement
btotoen polarizations being noted. Apparently', then'~ the
earth's ionosphere does not grossly change the polarization

of the incominG radiation'r
Since the long-period activity of the noise sources
is apparently correlated with the solar sunspot cycle,
it was natural to wonder if a close correlation existed
between day-to-dny radio noise activity and solar
activity. A study of the correlation between solar
corpuscular emission and Jupiter radio noise activity was
made at this laboratory by CarrP A; G Sm-ith and Bollhagen

(13). Near the time of opposition of Jupiter; from April
to August' 1960; observations rvore made at the Chile sta-
tion; A correlation betvoen Gooima-notic A-index and

Jupiter activity was apparent if a time ln. of eight to
nine days was introduced in the t-index: time scale.
This time lag is approximately the same time interval

one nould eopoct soln.r particles to take in travelling
from the earths orbit to Jupiter, assuming the same
velocity of travel as noted from the FOun to the earth;
This latter velocity was calculated from the time lapse
between an observed solar flare and the commencement of
the A-indoe chenge.. :'"arlck (14t) reports; from his
analysis; a time leo of one day. However, 7a.rvrick used
solar rndio-froeroncy activity as n moasuro of solar
corpuscular outputs. The analysis indicates that parti-

cles of velocity 0.1 c would be needed to obtain his







correlation. Further work on this subject will be done

by this laboraory sat the Chile'field station during the

1961 opposition,












CHAPTER II


A Radiation Source Model


A. Surmary of theories of the source of the Jovian rarliation

The sporadic decameter-wavelength radiation emitted

from Jupiter, with its wealth of detail, its period,
polarization, and spectral characteristics, leads to the
question of what can be producing this power. Initial

ideas considered the possibility of a li-hitning-bolt

phenomenon, or intense volcanic activity on the surface,

A lightning-discharge phenomenon monitored at a given fre-

quency would appear as continuous radiation if the process

were as on earth, rather than the sporadic type observed.
About 100 lightning flashes per second occur over the

surface of the earth# and at a distance these would appear

continuous in time* The volcanic theory attempts to ex-

plain the radiation as being caused by shock-cxcited

oscillations in a plasma; hence, an ionosphere surround-
inC Jupiter is called for. The "cone of escape" des-

cribed earlier, at first appraisals seems to fit into this
picture, indicating that a magnetic field is present,
Further credence is attached to this by noting the state

of polarization of the radiation# Rishbeth (15) made

calculations as to how likely a plasma oscillation







mechanism is, based on required ion densities in the pro-

posed ionosphere. Taking an oscillation frequency of

20 Ic/s, he found that an ionosphere with the required

ion density is a possibility, after considering the two

most important mechanisms for ion loss radiative recom-

bination and dissociative recombination. Thin ion

density, much higher than that observed in our own iono-

sphere, is a possibility even though ':he ionizing rndi-

ation from the Sun is much less at the orbit of Jupiter
than at the orbit of earth, because of the composition

of the atmosphere of Jupiter, The proportion of heavier

Gases is low, and Rishbeth points out that the molecular

constituents containing the heavier atoms likely to be

present, ammonia and methane principally, will be dise

socintad starting well below the region of highest

ionization* In the region of highest ionization, the

predominant component, H ions (protons), will be able

to attain high enough density to provide plasma oscil-

lations on the order of 20 Mc/s and higher. The

molecular components still present in the region should

be sufficiently fow to prevent loss of net ionization
by dissociation, Tho fact that the probability histo-

gram peak width decrcnsos with increasing frequency is

a difficulty in this theory, for higher frequencies

should exhibit a Inrger escape angle* One ides put

forward to circumvent this difficulty is to suppose that

the decreased power at higher frequencies makes recep-

tion less likely at the fringos of the ponk. A more







complete description of the ionospheric model is given

by Carr (16). The ionospheric model is not dependent on

the proposed shock-producing mechanism which led into

this discussion, but t is dependent on having some shock-

producing source which has a definite relationship with

the surface of the planet.

If a charged particle enters a magnetic field, it

begins to gyrate in this field. If the charged particle

has a velocity comiponont in a direction parallel to the

mag:noetic fold, the motion will describe a belin. This

motion involves an acceleration of the particle, so that

radiation will take place* For slow-moving particles

this radiation will take place at the classical cyclotron

frequency given below; thus it is called "cyclotron

radiation". For velocities near the velocity of light,

the radiation occurs at the relativistic cyclotron fre-

quency given by Equation 2-11 and at harmonics of this

frequency. The higher the energy of the particle, the

greater is the proportion of the energy that Goes into

the harmonics, and the higher the harmonic at which

peak radiation occurs. This phenomenon is called syn-

chrotron radiation* The radiation by particles in a

magnetic field is discussed below, to determine the

applicability of this type of radiation as a source of

the Jovian radio noise

Docausc the mass of the radiating particle affects

the intensity of radiation so strongly, as seen by Equa-

tion 2-1, electrons must be the source of radiation* The







Sun is a copious source of cherrged particles, Those stream

outward as a plasma, a grouping of positive ions and elec-

trons of neutral charge -nlity, Pertinent information

is given below nbout the observed physical charactoristics

of this stronL at the orbit of the ezrth. In Chaptor V

a brief discussion is given of the possible mechanism of

acceleration of these electrons to higher energies.

The synchrotron or cyclotron nodols simplify the

picture in at least two respects. First of all, the

ma-;notic field presents the localizing influence called

for by the data, and secondly, no outside mechanism need

enter the composite picture of electrons and field to pro-
duce the radiation# As shown in the next section, the

density of electrons required to produce the observed

raclition intensity is quite large at loecr relativistic

energies (v/c<0.95), so that on this benis alone a

cyclotron model mechanism is highly unlikely. The

salient features of synchrotron r-diation are next con-

sidered, so that they may be used to see how voll the ob-

osrved phenomena fit the model.


.- Assuumcd electron donsities

Before proceeding with a description of synchrotron
rndintion theory, it is in order to make a few simple

calculations of the energy nocessary for this process to
be observable. From the simple theory of a charge moving

in a uniform circle, the total intensity radiated is (17)








e' 1e3.2 (2-1)
3 mc7(i-`1)
whoro e is the electronic charge, m is the electronic
mass, Bo is the magnetic field in which the electron
gyrates, C is the velocity of liCht, and s is the ratio
of electron velocity to the velocity of light. If
numerical values are substituted in the above equation,
the result is

I- 1,58\x U, j. (2-2)

At the orbit of the earth, the number of ions pre-
sent in interplanetary space is usually of the order of
100 ions/cm.3 This figure may Jump to as high as 105
ions/cm.3 during periods of high solar activity (18)b
The ion velocities range from a value of about 500 km/sec.
up to 1500 km/sec. Recent observations of X-rays pro-
ducod in our atmosphere as a result of solar flares (19),
indicate that a large number of electrons of high energy
are available at the orbit of the earth. The peak flux
from a strong flare amounted to between 2 and 6 X 107
eloctrons/cmi2/sec., with energies greater than 50 keyv
Measurements of the Van Allen radiation belt show-: that
there are 108 to 1011 electrons having energies greater
than 20 kev* trapped within the belt in a tube which
connects the two aurora zones and has a one square
centimeter base measured at the auroral zone.
A typical nodorately strong Jupiter radio pulse can







have a power output, as seen from the earth, of 101
natts/m.2-cps* If a representative value of $ astro-
nomical units is taken as the distance from Jupiter to
the eorth, and a bandwidth of 5 X 10 cps. is taken as
the band of emitted frequencies, then, assuming isotropic
radiation, one would have a figure of


I 3.5x Io'0 watts (2-3)

for the intensity of a given storm. Since the radiation
is certainly not isotropic, it is felt that this is a
conservative estimate on the high side. Since Equation
2-2 above is the intensity of radiation emitted from a
single electron, the energy emitted from N electrons
with a mean onorgy determined bye, vould require that


S2x o32 3- e lec.frons (2-4)

participate in the emission* From this it is immediately
seen that for magnetic fields of the order of magnitude
needed to explain plasma oscillations on the order of 7
gauss, the number of electrons is lnrco indeed, 'save
for extremely high energies. For an order-of-magnitude
calculation, values ofa = 0o95 and Bn =s gauss are used
in Equation 24- to get

N 2x lo0 electrons (2-5)

Now. suppose that electrons radiating outward from the
Sun at the velocities and numbers as givon above near
the orbit of Jupiter before being accelerated, Using







a particle flux of 103 elcctrons/c-.3at at velocity of 108
cn./sc.,t it is found that F, the flux thrnubh unit area
in unit time is

F= /lo elecfrons/cm.."- 5ec.

A storm will last for perhaps 5000 seconds and since the
particles lose energy all the time in the magnetic field,
it is ronsonable to assume that the electrons arrive in
the field during only a fraction of this time. A spec-
tral feature which might possibly be caused by the energy
loss is discussed in Chapter V, For this calculation,
suppose that the injection time is 500 seconds. Then
the flux per unit area is

F, 5x" o electrons/ct.& (2-6)

Surely not all the electrons are accelerated to relati-
vistic velocities, so assume that this number is reduced
by a factor of 103, indicating that only one electron in
a thousand is accelerated, The area over which electrons
must be trapped into the magnetic field, unin- those as-
surmptions, is

As /-xo I0 cm.z

which is an cren comparable to the disk of Jupiter, On
the bnsio of these calculations, the value of 6 chosen
bhre must be taken as an extreme lower limit.
If the trapping area vworo sufficiently lnrr!e to
collect more oloctrons, the total nuriber present in the







radniation belt would even so avontunlly reach a limiting

density, beyond which they would be dinsipated. The num-

ber of electrons needed for the synchrotron model cal-

culated above is already high, so that the value of/

chosen here could not be made much smaller. For a

radiation.belt model, the particle density at high

energies must be much larger than that measured in the

radiation belt of the earth. The number of electrons

needed increases sharply with decreasing energy.


C. Synchrotron radiation niodel

The value of 3 chosen as a lower limit in the pre-

vious section, indicates that the best approach to the

problem of fitting the synchrotron model to Jupiter

radiation should be from approlinations assumiing high

energies. The theory can not be carried through ex-

actly, so that work done sc far has been in the realm
of near-relativiatic particles nnd ultra-rolafivistic

particles. Since the calculations in the abovo section

point to a lowor limit of onerioes nearer the ultra-

relativistic limit, this approach will be followed here*

The outline of the model givcn here is that developed

by Uestfold (20). The primary purpose of the develop-

ment here is to indicate the general features to be ex-

pected of synchrotron radiation and to determine limits

of applicability of this appronimato theory. The dis-

cussion in Chapter V shows that the sir.mplc picture out-

lined hero does not produce a -ood fit to the observed







data, but various quite probable factors that would modify

this picture are also pointed out*
In the following devolopmont, the symbols used are
as pictured in Figure 3, The symbols are as follows:

-o= constant magnetic fiold
f(o)-direction of original electron motion
= direction of electron motion
C = angle between magnetic field and electron
motion
== direction of observer, taken to be in the
yz plane
X angle between f(o)andT
y= angle between n and V(o)

Vostfold has developed the theory for ultra-relativistic

electrons and the approximations used vill be studied
for a lower limit of/s= 0,95.
From the relativistic equation of motion

= efx o (2-7)


hero the symbols are defined as usual, the solution

,= constant (2-8)


may be obtained, so that


=- WSX/3 (2-9)




48.= r(2-10)

This is the cyclotron frequency for an electron in field

Bo with velocity pc. This annular frequency nay be re-
duced to















COORDINATES OF ELECTRON MOTION


Fig, 3.-Electron Motion Coordinate System Used In Synchrotron

Radiation Derivation.









(2-11)


f =2.8o 8JI-/,s Mc/s


As the electron velocity incrcnaso., the cyclotron fre-
quency becomes more and uore reduced from its classical
value of 2.8 Bo Mc/s. The poor flux must be computed
from the Lionard-Wiechert potentials, which yield a re*
suit for the electric and fmlnetic fields

E e= n(-n-x x '] (2-12)
trR 0- )3



B-= nxE (2-13)

Here only the radiative component of the field is shown.
The primes on the above quantities indicate coordinates
at the retarded timos. /( is the pcr;.ioability of free
space, R is the position vector from the electron at its
retarded tino to the observer. Introduce

(2-14)

and expan in teris of ( to obtain

f/-F--) (2-15

The ratio of the last two terms on the right for /3-095
is 1/ qko and the approximation involved in dropping the
third torn becomes better as s increases. It is now
necessary to inveotigato the size of the anglo between
and n defined by









Si/7= cos e (2-16)
/3
In Appendix Ak the component of this angle between the
plane of gyration and the direction of observation is
shown to be of the order




The development of the equations of polarization and
harmonic power is carried forward using this approximan
tion, Since mort of the radiation is confined to.,a small
angular range, the limits of integration can be carried
from the actual limits to infinite limits when they
involve this angular dependence According to curves
given by Ostor (21), the radiation at the lower har-
monic numbers is spread out over greater angles than

at high harmonic numbers. The approximations used hero
have more effect on the lower-frequency components than

on the higher-frequency components.

Returning to Equation 2-12, the factor in the
numerator may be written







Since the radiation is concentrated near the plane of

gyration


n-\ x0()







that is, the absolute value of the described victor is
small and is a function of the angle 9. At this point,
large angles are effectively eliminated, since, for
instance,

(n-i) ^ 0o(e) r Co(f() ( f (rnajude)


is of third order and may be neglected for small angles.
Terms of this type are dropped. Also,

'WB. x-9 W X = (o-




These factors lead to the result that

n x L[f -/3 xl,3 J] ^ (n- )- C -^

If the electric field vector is expanded in a
Fourier series

(rt)= En2knWsf (2-17)

then




will yield information of the type that is sought. In
retarded time, take the distance to the observer to be
largo, so that


Then with th approximations ready introduced, th field

Then with the approximations adready introduced, the field




28

vector F bocoics

Ee J ec UwO. Sin r a ?'(2p ( 1 + t*
'n Yr (-Z+ ( I'+ sin.C

-t, 5in At ( 'coe-l~ <) / 71'
+r Id+ '-i'in )Z


after transferring to angular coordinates, with no fur-,
thor assumptions than those stated nbove, This form
may be integrated, as stated by Westfold, to yield a
result in terms of the modified Bessel functions.

fn r id& Cn_


rY {ff^ cos-in (2-18)irin

Reference to those modified Bossel functions may be
found in the literature (22).
To the approximation used,

x n = n cos I siA C

The polarization is dotermiined by the ratio of the com-
ponent of the electric vector in the ? direction to the
component perpendicular to both the i direction and the
direction to the observer, in direction -,'x For this
one-electron modol, the polarization is given by



S.Xft^ Kis[^2Y^JI (2-19)
Y~ KCiS;IKy?~






The power radiated in harmonic n as a result of the fields
in the direction- xu and 7 respectively, is

( (2-20)


Z. ^ ^ ) (2-21)
h re 1* n-F)
Where F,, and nc are defined by

) Kx)= [ -(x)j
F M(x)= x [JKS ((2-22)x
C 3S
The total power radiated is the sum of Equations 2-20
and 2-21. It will be shown below that in the case of
decameter Jupiter radiation, the harmonic numbers are
such that the radiation cannot be taken as nearly con-
tinuous in frequency, For an ordor-of-jiagnitude calcu-
lation, the following will be used.

2-lr
S-P -P8.
The power radiated per unit frequency interval will be


< <>jf



j e u f &5in<
Using the usual values for the constants in the mks system,
the power radiated in a given frequency interval per unit
area is
and2.3~0X3 / x /0 r-2rod/ s/ (23.
and c 4 .oi/jL Mc/s (8-23)
1-I T







Values of F = FP' + IF and some of the modified Bossel

functions are given in Appendix B.


D. Application of the model
The development above has been for a single energy

and for a constant magnetic field. In actual practice,
both of these quantities vould be variable. Other vari-

ables affecting the observations include the time be-

tween collision of the electrons ..with other particles,
and the kinetic temperature of any associated ions.
From the number of electrons computed to be neces-

sary, they must be confined in a radiation bolt around

Jupiter and contribute over a period of time to the ob-
servod radiation process. If those electrons approach a

region of increasinG magnetic field, the relation

Si'2^,L canstani


will hold (23). They will thus be turned around at some
"mirror point" and be retained in the belt. Because of
the localized nature of the radiation, it must be assumed
that the observed radiation occurs near this point, such
that the electron radiates a largo quantity of radiation
into a small solid anrlo.
One further point will be brought up at this time.

If the region of radiation is such that the wavelength
of radiation is much longer than the characteristic dis-
tnnce between electrons, the radiation w-ill be partially
coherent (24)* The intensity will be a function of







electron distribution in phase space and of the dimensions

of the bunch. In one of the examples given in Reference

21, a Gaussian velocity distribution of electrons produces

an intensity'factor




where co is the nocn amplitude of phase oscillation in

a bunch. Further discussion of these factors will be

postponed until after the presentation of spectral dnta.












CHAPTER III


Spectral Studies Using Multi-Channel Data


A. E.quipient utilized
During the 1960 observing season, the Florida station
employed an 18 Mo/ Yagi-type antenna, an 8-element 18
Mo/s broadside array, a 22.2 Mc/s polariameter consisting
of two crossed four-element Yagis, a 22.2 Mo/s 32-olo-
ment broadside array, and two 27.6 Mo/s arrays, consis-
ting of a Yagi type and a corner reflector with a di"
pole cut to this frequency.
The 18 and 27.6 Mc/s Yagi antennas are steerable,
so that they provide coverage over the entire possible
listening period.
The Chile observatory used two 16-element 18 Mo/s
broadside arrays, a l-element 10 Mc/s array, a corner
reflector which was operated at 16.7 Mc/s, an 8-element
20 Mc/s broadside array, and a square 22.2 Mo/s array
consisting of l. folded dipoles mounted one-quartor
wavelength above a reflecting plane,. This last antenna
provided the polarimeter measurements from Chiles
In the 1961 observing season, the corner reflector
in Florida was used at 15 MCc/ while a 15 Mo/s steerable







array replaced the 16.7 Mc/s channel in Chilo. In n:dition,

a 5 Mc/s broadside antenna was constructed and is now in

use in Chile.

Each antenna at these sites is connected to a com-

mercially-built short-wave receiver when in use. The
audio output from the receiver is rectified and recorded

on pen recorders running at the rate of six inches per

hour. During Jupiter noise storms, selected channels arp
also recorded at five millinoters per second.

The rhombic wide-band antenna used at the Florida

site will be described in Chapter IV.


B. Observing procedure

An observer is always present during watch time.
It is the duty of the observer first of all to differen-

tiate between noise produced by Jupiter and that produced

by terrestrial interference, which takes the form of
static, station interference, ignition interference,
busses produced by faulty neon signs and faulty trans-

mission lines, as well as other less easily identified

sources. It has been found that Jupiter-produced noise

has a characteristic "swishin~f sound, which makes it

easily distinguishable from most types of interference.
An unmodulated station fading in and out most nearly
duplicates this type of behavior, and e::ccpt in cases
of extreme station interference, this type is easily

identified by tuning through a few kilocycles; a sta-

tion will "tune out", while radiation from Jupiter is







broad enough in frequency to be nearly unaffected. Radi-

ation from Saturn may have been observed, but it is very

much weaker than most Jupiter stones. During the periods

th:In Saturn and Jupiter are nearly in conjunction, the

type of antenna most used at this observatory will not

differontiate between the two sources* This was the

case during much of the present observing season.

At the onset of a Jupiter storm, the obsorvor

chooses the channels to be recorded at high -speed, vhich

may include the polarimeter channel. The polarimeter

records are run separately at throes 'inches per minute

during noise events. The small froquency-difference

spectral data were obtained by tuning two channels to
about the same frequency near 18 Mc/s. During this ox-

porinmnt, a single antenna was used to feed both receivers.

Various differences in frequency woro used during Jupiter

storms, these differences bcing typically of hbe order

of a few tenths of a megacycle. Of course, wide frequoncy-
difference records were also obtpinod from the several

different antennas depending on the channcls on which

Jupiter was being received.


C. The small frequency-differonco records

On a nuraber of occasions curing the 1960 urntchos
in Chile, very satisfactory multiple-froeuency records

were obtnincd. Using either of the two 18 Mo/s broad-

side arrays described above, records were obtained w-ith

frequency differences up to about one Mc/s. An effort







was made to use these records to obtain an impression

of the bandw'idth of a pulse of the radiation. Even at

smnll frequency separations, as shown in FiCuro 4, the

correlation between pulses can be quite poor at times.

By judiciously picking time intervals so that a narimum

correlation could be expected, a number of readings of

the linear correlation coefficient were rondo.

The purpose of these moasurcraonts was to dotormine

at how wide a frequency difference good correlation could

be obtained, This gives an indication of the maximum

with of the received pulse. A choice of correlation

samples whore there-was reception at both frequencies

provides information as to whether or not energy re*

ceived at both frequencies was part of a pulse of this

frequency ridth. Another possibility is that pulsec at

two different frequencies are not part of a continuum

through those frequencies, but are two csparate tine-

correlated pulses. A discussion of this possibility is

given in Chapter V.

Given two variables, FP and F2, the linolr correla-

tion coefficient between them is given by




Here, and F2 are the two samplod frequencies, F2

being the lower frequency in each case. A one-to-one

correlation between pulses would yield a value of r

of one, bhilo no correlation would produce a zero.

Random correlation is always present, but its effect
























RECORDER MARK
SF.J: .18.100oMC fEc.


*>W '*^ >Mj'^UUj;,lJ.->^1 .AU. .Al-A ^ l -t


Pig. 4.---Frequoncy-Separated Record Oi- Mc/s.







decreases as sample size increases. A resume of a num-
ber of correlation meacurcments made for tines when thcro

was Jovian noise at both monitored frequencies s given

in Table 2, v:ith the column headings corresponding to the

quantities defined above, The value of P, the probability

that a correlation of this size would come about in an

uncorrelated population is obtained from Fiheor (25),

As can be seen, with a proper choice of sample, there

exists a large correlation between pulson separated by
as .much as one Mc/s, Conversely, the correlation can
be small for frequency sopp.rations of 0422 ITc/s. The

probability column indicates that the chance of these

correlations being random is very small in most cases,
even whore the calculated cort:olation coefficient is
rather small. Since correlated pulses in this sa-mpling

have been observed up to a frequency difference of one

Mc/s, this indicates the occurrence of pulses of at
least this width. Activity during the snmo time period

at this small frequency separation seems to indicate

the reception of a single pulse of this frequency spread

in the majority of caeno. Similar work was not attemp-

ted for larger frequency separations, because visual

inspection indicated a generally poor correlation. On

several records of coincident 10 and 20 !c/s runs in

Chile, no instance could be found of simultaneous pulses.
Rever.l records of the type shown in Figure 4 were
analysed for relative amplitude at the two frequencies*

The maximum amplitude of each pulse was measured from









TABLE 2
CORRELATION BETWEEN FREQUENCY-SEPARATED PULSES


Date Channel 1 Channel 2 F rpF k p
(Me/s) (Mo/s) (M/s) 2


3-30-60
3-30-60
3-30-60
3-24-60
3-24-60

3-24-60
3-24-60
3-24-60
3-24-60
3-24-60
3-29-60
3-29-60
3-29-60
3-29-60
3-29-60
3-29-60
3-29-60
3-29-60


18.148
18,148
18.148
18,600
18,600
18.600
18.600
18,600

18.402
18.395
18.700
18.500
18.300
18.100
18.700
18.700
18.700
18.700


17.744
17.744
17.744
17.980
17.980
17.980
17.980
17.980
17.980
17.980
18.000
18.000
18.000
18.000
17.800
17.800
17.800
17.700


0.404
0.4o04
0.4o4
0.620
0.620
0.620
0.620
0.620
0.422

0.415
0.700
0.500
0.300
0,100
0.900
0.900
0.900
1.000


0.60
0*66

0.41
0.19

0.50
0.48
0.47

0.24
0.34
0.74
0.27
0.70
0.73
0.75
0.25
0.62
0.34
0.60


<0.01
<0.01
<0.01
>0.10

<0.01
<0.01
<0.01
(0.01


<0.01
>0.05
<0.01
<0.01



<0.01

>0.10
<0.01
<0.02
<0.01


rr- ------ -------- -- --- ---- .~..~.. 1.1 .. ~ _-._..__1.1..1..._.~... ~







the level of galactic background .When the ratios of these

amplitudeo at each sampled point uoro takon, most occurrence

periods. shoved no chanGo in the avoraeo amplitude ratio

from boGinninC to end of the period. This period.was

about a minute long in each case. Many of the periods

used in the correlation measurements also constituted

soapling periods for these measurements. On some occur-

roncos, the average of the amplitude ratio taken over

several consecutive pulses socned to change over the sam-

pled period. Figure 5 illustrates an apparent change in
amplitude ratio rbore A represents the amplitude of the

pulse at higher frequency and A represents the ampli-

tude of the pulse at lovror frequency. Figure 6 illus-

trates an apparent shift in the opposite direction*

The data points are indicated to bho7: the wide ranCo of

fluctuation in relative amplitude which occurs in a

grouping.
Because of the wide soattor of data points, only a

qualitative measurement could be made here. The measure-

mont indicates the possibility of an observable effect

occurring during this time period, It will be shonv later

that the change in rclative amplitude fits into the ob-

served spectral data quite well, and could have been

anticipated from it,


Do Observations of onset and termination of storms
A feature of the spectrum vrhich is readily observable

is the time of first occurrence of radiation at each












RELATIVE AMPLITUDE SHIFT AT 0.3 Mc/s FREQUENCY SPACING
MARCH 29, 1960



o
3



2 0 0


o
SI- 0 00
0 0

O
51 ) 0 0
_1O 0



10 20 30 40 50
TIME (sec)







Fig. 5,--Example of Incroase in. Ratio vith Time of Amplitudes


of Froquency-Separated Pulses.



















RELATIVE AMPLITUDE SHIFT AT 0.620 Mc/s FREQUENCY SPACING

MARCH 24, 1960


0
0



0


0 0





10 20 30 40 50 60 70
TIME (sec)


Pig. 6--Eamnple of Decrease in Ratio vnith Time of Amplitudes


of Frequency-Separated Pulsos.


3







0
'--


0
n-



0-
I-
0.







!::oiiored frequency, and the time at vbhich the radiation

stops at each frequency. Sometimes difficulties in making
those measurements are presented by. local interference or
by weak or sporadic noolon emitted from Jupiter, Thoro is
also the possibility thnit the rndlintion is in progress
when the planet co-mes into the beam of the ontonnas.c

The occurrences were divided into three. separate
groups. The first group consists of those storms which
started at a higher .frequency and then appeared at

successively 'lowder frequencies, The second group con-
sists of storms in which no drift was apparent, or the

time of first.occurrence at each frequency did not pro-

coed in any one direction. The third group is the re-,
verse of the first group, consisting of stor.-s showing
a.drift upward in frequency. The terminations were

likowviso divided into these three groups, That is,

activity stopped at the higher frequencies first for
the first group, and so on* Table 3 indicates the
results of this grouping with three columns being

listed; occurrences in Florida, occurrences in Chile,
and occurrences picked from these in which the fro-
quency drift was in the same direction at both sta-
tions on the sane night. Also listed is the case hero

the onset and ending drifted in opposite direction,
Typical rates of drift are of the order of a few 9ia.ea-
cycles per second per minute. An analysis of drift

rate versus position of Jupiter in relation to the an-
tennas indicates that this drift is not caused by









TABLE 3
DRIFT RATES OF ONSET AND ENDING OF JUPITER NOISE STORMS


Year No. That Drift to No. That Drift to Erratic or
Higher Frequency Lower Frequency No Drift

Fla* Chile C Fla. Chile C Fla, Chile C

1960
Onset 3 16 1 27 2 24
Ending 2 11 1 27 2 30
Both 6 16
1961
Onset 3 4 1 3 14 2 8 8 3
Ending 3 7 1 5 6 2 6 13 5
Both 3 3 1 2 5 1





Drift Opposite at Onset and Ending

Fla, Chile C

1960 0 15
1961 0 5







Jupiter drifting into the beam of different antennas at

different times.

The measurements sho'!n in Table 3 and labeled as

"C" indicate those events which both Florida and Chile

reported as possessing the given drift direction. A

comparison of the times of initiation and termination

of the storms shows that 3 of the 8 "C" events in the

"driftup" or "drift down" columns did not correspond

in time, so that other factors than those at the source

must have entered to give the observed drift* The

smnll number of these measurements in which the drift

rates are in the same direction at the same time is

partially caused by lack of a large number of multi-

frequency records from Florida. The more severe

interference at the Florida station and its less favor-

able position with respect to Jupiter has resulted in

far fewer records being made of Jupiter bursts in

Florida than in Chile. Many of the Florida occur-

rences were on a single frequency, while Chile re-

ported a multi-frequency occurrence, further indica-

tinr- that this lack of correlation in drift rates is

to some extent due to local conditions. Table 4 lists

196t) zucurrences when the drift rates were in opposite

ditecions, or where one drift rate of the pair was

negligible. The list includes only unquestionable

occurrences, omitting occurrences where radiation may

have been interrupted and then resumed at a later time

on the same night in those cases where this effect









TABLE 4
OCCURRENCES IN 1960 AND 1961 IN WHICH ONSET AD EIIDIIIG
OF STORM VIERE IN OPPOSITE DIRECTION


Date Onset Onset Drift Ending Ending Drift


6-9-60

4-25-60
8-5-60o
7-29-60
5-6-60
L-1l4-60
3-24-60
6-7"60

7-30-60
4-17-60

5-4-60
6-30-60

4-20-61
3-25-61
4-6-61
3-27-61

4-21-61


120

125
136


160
196
214
221
228
230

235
239

120
238
252
260

317


130
130
150
150

170-270
202

235
227
246
250
250
239

127
267
258
271

336


----- ^- -- ~ -1-I -- -- -- ~- -







confused the possible drift nsciGnmont. The symbols

used are D-dovn, N-no drift, and U-up. The occur-

rences are given in terms of median longitude, with one
exception, the occurrence of May 6, 1960* Here the
toerination tines at different frequencies were so far

apart that the rrnce of longitude is given. It should

be noted that in all the occurrences listed, the 10

tc/s charnnl was either not in operation or no recep-

tion at that frequency occurred* When records which

othorwiso might have been included s1hovod 10 Mc/s

occurrences, the 10 Mc/s onset and ending times often

destroyed a drift tendency noted at the other fre-

quencies. When such was the case, the event was not
included. The drift tendency for the 1960 season as

a whole is plotted in Figures 7 and 8. Thoso figures

show number of occurrences versus System III longi-
tude. It is apparent that a drift downward is far more
provalont for both onset and termination than is an

upward drift in frequency. The grouping of these is, of

course, due to the probability cf occurrence of noise

storms as a function of the longitude, although the up-

drift occurrences are well spread out. Thoroe sees to

be a slight tendency for the dovn-drift group to be

displaced to higher longitudes for the greatest density

of occurrences on the onset figure.

The multi-channel records present a number of baf-

fling problems, with a number of curious phenomena

observed. The greater probability of drift down in













DRIFT hATES FOR 1960-STORM ONSET


I I I I I I I I I


11iII, I I


UP







S.-
ui.

o- ERRATIC
OR
z NO DRIFT
5-
o


1111


I I I I I I


200
SYSTEM III LONGITUDE


Fig. 7.--Drift Tendency of Onset of Noise Storms.


I I I


I=ONE OCCURRENCE


DOWN












DRIFT RATES FOR 1960-STORM TERMINATION


( I I I I I III I


I I


1 ONE OCCURRENCE


I I I II


, ,I I 11 ,


100 200 300
SYSTEM III LONGITUDE


Fig. 8.-Drift Tendency of Termination of Noise Storms.


UP







I-
cr
C
- ERRATIC
OR
z NODRIFT
i--
w
a


I I I


II I


I IS S Ii


DOWN-


I I







frequency, and the poor correlation (at times) of noise

picked up on channels separated by a mall frequency
difference, while correlation measurements mndo at much
wider frequency separations at times show Jood correla-
tion have been noted. From this analysis, pulse widths
up to one Mo/s have been deduced. The spectral analysis

system described next clears up some of the ambiguity

so far introduced, and produces results which tend to
dispel many of the unanswered questions brought up

here,












CHAPTER IV


A Swept-Frequency Decameter-Wavelength Receiving System


Much of the work connected with analysis of the
spectrum of the decameter-wavelength radiation from
Jupiter has been done using data collected from a
swept-frequency receiving system. The system, shown
schematically in Pigure 9, consists of a rhombic-type
antenna, a balanced input wide-band preamplifier, a
commercially built balun coil to go from balanced to
single-ended feed, a high-pass filter network, a coaxial
line to the observing point, a wide-band commercially
built preamplifier, a spectrum analyser, and recording
apparatus* The spectral distribution was photographed
from the cathode ray tube display face of the spectrum

analyser by a Bolex 16mm. motion picture camera and
Eastman Tri-X film, This recording set-up is illus-
trated in Figure 10.
The rhombic antenna was designed, using published
data (26), for a center frequency of 20 Mc/s, a aide
length L of 1*. wavelengths, an angle of 900 between
sides, and an input impedance of approrzinately 600 ohms.
A schematic representation of the antenna is presented
in Figure 11. The terminal impedance of the antenna was






















PG-1] Line


Rhorbic
Antenna


*,?IDE PAND TETFCTION AND PIECCRDINr- SVSTFM


Fig. 9.--Roprosnntation of Wide-Band Roceiving System.





































Pig. 10.--Spectral Distribution Recording Configuratica.








checked at three different frequencies using a Ieathkit

impedance meter* Since the impedance meter is a, single-
ended instrument, three baluns were necessary to make
these measurements. The baluns were constructed of RG-63

coaxial cable, two half-wavelength lines being used to
provide a balanded-to-unbalancod match at the desired
frequency, following standard constructional procedures

(27)i The values obtained with the final configuration
are shown n Figure 11. Figure 12 is a view of the
termination and one side of the final configuration@
The following factors were adjusted to obtain these
figures the distance between the antenna ends at the
resistor termination end, the slack on the wires which
make up the antenna, snd the separation between the two
wires which make up each side. The separation was the
most critical factor in adjusting reactance. As a whole
the impedance changes extremely slightly with a change
in the tautness of the wires or the termination sepa-
ration.

The design resulted in the following theoretical
figures: the maximum sensitivity is in the plane of the
antenna, the half-power points are 15$ off-axis in the plane
of the antenna, .the half-power points off the plane of
the antenna at the axis are n!Tout 150 from the plane of
the antenna, and the antenna is most sensitive to radi-
ation polarized in the plane of the antenna. On the
basis of these figures, the antenna is useful over a
period of about two hours during each watch, asaming





















RHOMBIC WIDE-BAND ANTENNA


T o -
I D


DESIGN PARAMETERS
L-70.7' h85
Rx66O *2900
K:65' os37
0 lod


FREQ(Mc/I)
212
182
16.9


IMP.bhmfn
560
600
650


Fig*, l1.--ihomble Antenna Design Parameters,






































Fig, 12.-.View of Pinal Conf'iguration of Rhombic Antenna

bowing One Side and Output Termination.







thar.t Jupiter passes all the way through the boamn. The

dircctivity Cain, compared to a frco-space dipoloe is

9.85 decibels. The signal gain of the rhombic antenna

is decreased from this value, because of .the losses ex-

perienced in the termination reciztors. The termninntion

resistors absorb approximately 3 decibels of the input

power, so that the signal gain is about 6,85 decibels

over a free space dipole. In addition, the coaxial

transmission line has a loss greater than 2 decibels,

since its length is in oeccsc of 250 feet.
SBecause of the low nal in d e lw iCnl n d losses in

the matching and lead-in section, a low-noise preampli-

fier v.as necessary to increase the signal to a level

sufficient for the panoramic receiver* In 1960, a

proanplifior using a sinr-lo 6BQ7 input tube as a cas-

code proniplifior followed by additional sta~;cs of

amplification was employed at the receiver end of the

coaxial input. The cascode amplifier was induc'tively

coupled to the .trsnomission line,a uing e single tapped

coil to provide strong coupling. The strong coupling

produced a frequency response that was, sufficiently

flat in the frequency range of interest, The preampli-

fier had a rather large noise level, 'so that only the

strongest radiation from Jupiter could be roccivod.

For this reason, this pronmpliflor will not be dis-

cussed in detail.

Before the 1961 season, it was decided to build a

balanced preamplifier to be mounted at the antenna. The







bnlLun coil used the previous year was discarded in favor

of the Hoathkit balun, since the old coil had rosistive

olcionts in its matching system, and it was novw apparent

that there was not any gain to spare.

The balanced pron!rplifior is abown in Figure 13.

Only those component values which are unique are indi-

cated. One may see that the preamplifier consists of

six ossc.ntially identical sections. The design is

based on the original paper on this type of amplifier

listed in Reference 28. The signal travels down the

artificial line and is amplified by each tube in turn,

If the velocity of propagation is the same in both grid

and plate circuits, then for n tubes in the circuit, the

amiplification is n times that of a single tube. In

the present circuit, the gC.d and plate lines woro

mado identical, vith an inpedanco of 330 o1hno The

Slrinos of capacitance and inductance used are givon in
Appendix C. The input and output impedincoa of the

con!iletod proam:plifior proved to be close to the dcoign

values. The irnpcdcnco is relatively constant up to

about 24. ic/s, after which it varies radically* The

total signal gain to the commercial proe'nplificr

which was mentioned in the discussion on page 4.,,

is approximately 16.5 decibels over a freo-space

half-wave dipole. This figure checks well with ob-
served signal gain of the 18 Mc/s Yagi-typo antenna.

The antenna was connected to the sweep analyser and















Ot1 I


Out 2


Fig. 13,--Balanced Input Wide-Band Single Stage Preamplifier.







the galactic level compared to the galactic level with
the rhombic system connected. Using a signal generator,

the noise level of the rhombic antenna balanced preampli-

fier was monoured to be about 0.25 microvolt across 300

ohms* This noise level is about the same as that mea-
sured for the commercial preamplifier across 50 ohms.

The galactic noise level, as observed on tbh cathode

ray tube face, is well above the noise level of the

total system.

The high-pass filter is also discussed in Appendix
C. Its lower frequency cut-off is 15 Mc/s, and it
determines the lower odre of the received band. It
was placed at the antenna end of the receiving system

during the present season to minimize.line reflections
which tworc found to be present.

SThe commercial preamplifier was built by Ceco Com-

pany, College Park, Pennsylvania. It has a frequency
response of 10 to 90 Mc/o, an input and output iupocdhnce
of 50 ohms, and a gain of 0.Q docibcls. The panoramic
receiver is slown in block din-rim representation in

Picuro 14 (29). The swoop width was operated at soveral
different values, running from 1 Mc/s up to about 4. Mc/s.
The recurrence rate was usually set in the neighborhood

of 30 cycles per second. The 16n=m notion picture
camorn was triggered at the rate of two frames por
second during toh first season of use. The appearance
of the records indicated a loss of information, so that
in the 1961 observing season tho rate was increased to













Input


32 m. 32 me. xixe 2.7 me.
FilterIn~t Int er.
Freq. Freq.
Amap. h g







29.3 nc. Detector
08C.








Sawtooth H orizontal ' Vertical
Cenerator output o utput
1-60 cps

Oscilloscope
Tube
SWEPT FR~EQUENCY SPECThUL4 ANALYSES


Pig #1*--Block Dingramn of Panoramic Tiecoiver.







four franos per second. The persistence of the phosphor

on the tube face provides information as to more rapid

changes. To provide a measure of the amplitude of the

envolopo of the noise, the video filter in the instru-

ment was employed. The effect of this control is to

average out sharp changes in level with frequency.

Various levels of video filtering were tried, and the

most successful from the point of view of providing a

clear picture of the radiation was chosen. In the

photographs to be presented, some variation in the

filterin can be noted as a variation in the glactic

noise signal width from series to series*

The exposed 16mm. Tri-X film was developed in DK60-A

developer, following the recommondations of the manu-

facturer, Eastman Kodak Company. It was found that one

gallon of developer was sufficient for four 100-foot

rolls of film. Further use produced poor results. A

Morse G-3 rewind-type developing tank was used to de-

vclope all fil. The film is tightly rolled on two

reels which are iL-erscd in the processing liquid in a

licht-tlght container. It is usually necessary with this

typo of tank to roll the film between reels by hand, using

cranks oxtcnding outside the tank. Because the film is

in free contact with the solutions for a short period

during each re-rollinG operation, the time in each

solution is extended beyond What it would be if the

film were in constant contact This factor is taken into

account in the published developing data supplied with







the film., Because of the large quantity of film to be

processcd, a mechanical rowinding mechlnnicn was constructed

in the physics shop. This mnc-chnicm provides for. auto-

matic cranking of film between reels, and by use of a
r-..r.'tltr nrr nrrnccent, a reversal of crrnn:ing direction

at the end of the film. The time to run through 100

feet of film once is about twominutoos The procedure

continues for the time required in each solution. The

operators required to put to t no the tank, through a

light-trap, the solutions in the sequence hardener,

developer, and fixer. After the last solution, a

washing tine of at lost 45 ninuton must be usnod. The

film is then rolled on a drying rack, cnuloion side

out, to dry*

The film was viewed on a l6nmu movie film editor,

The poa: frequency data were obtained front the viewing

screen of the editor. It was found necessary to use

care in centering each frame so that it was the sawno
size on the viceor as the other framor from which data

had been tnl:on.












CHAPTER V


Spectrum of Bursts


A. Description of noise pulses

An analysis of the photographic records indicates

that somo periodic gain irregularities still ronain in

the system. They are noticable only on occasional
frames and soon to be spaced about 0.8 Hc/s apart.

Two sets of measurements were found to be neces-

sary to delineate the nature of the noise bursts.
Vith a swoop width of three to four Mo/s, the various

bandwidths of noise bursts could be observed, while

the structure of the individual bursts could best be
oraminod with a one Mc/s sweep width. The proscnta-
tion of photographs of pulse samples is necessarily

bulky, so these are included at the end of the proesnt
chapter. The following order was established to pre-

sent, in some kind of ordered mannor, the rance of

pulse typos which were observed. Each series that is
dcscribod was picked as showing some particular fea-

ture that will be pointed out. Figures 17 through

57 depict those oorles of spectral photographs.
Series 1 is at wide swoop width and illustrates

a smooth pulse developed over a wide bandwidth. Series







2 through Series 4 consist of several wide pulses with
various charactoristics apparent in each. In Series 3,
in particular, there exist several frames in which ex-
tremely sharp spikes appear. Series 5 and 6, while
broad-band, indicate sharp frequency differences in

amplitude. The sharp rise in amplitude noted in Series
6 is what one would expect in plasma oscillations -
that is, a sharp lower frequency cut-off. Series 7

illustrator a sharp pulse of rather large amplitude.
On weak pulses, a variety of features may be noticed.

Nost of such pulses are erratic in nature, and in many
a bifurcation is seen. Series 8 through 11 depict
these conditions. In Series 9, a pulse appears at
about 18.8 Me/s and about one-half second later a
pulse appears at about 17*4 Mc/s. Series 10 differs
in that a sharp pulse appears in the first frx~: at
a hiCher frequency, followed by a wider pulse centered
at about 17*6 uc/s. Series 12 and 13 consist of one
frame apiece, the first illustrating a nearly sym-
metrical pulse and the latter one with a gradual rise
in amplitude with froqucncy, followed by a stooper
doclino.

The movement of the noise 'energy either up or down
in frequency has been an object of discussion before.
MVhile a drift rate measured in megacycles per minute
would not be too readily apparent in those high-speed
photographs, a few series soem to represent a motion,
although no single pulse uws seen to travel in frequency*







This may be due to the short time duration of individual

pulses. In Series 14, the pulse appears rather sharply

at 17.5 Hc/s and then broadens out to higher frequency*

Series 15 and 16 indicate, respectively, two pulses,

one after the other, moving downward in frequency, and

one pulso starting at higher frequency and broadening

to lower frequency. Finally in this width group, Series

17 and 18 show two series of pulses developing. Series
18 appears to consist of pulses all near the higdh-fre-

quency (right) .ond of the trace, and it should be noted,

as a characteristic of the pnnoranic receiver, that the
traco begins at the left at a high level if it ends at

the right at that level. (That is, a signal coning in

at the higher-frequency end of the trace produces a
deflection which persists at the lower-frequency end

of the trnce. This effect is apparently caused by the

use of the video filter. The filtering also produces

an increase in the general baseline level if a strong

signal comes in at one frequency.) The width of the
"rLass" or noise trace vertically gives a bettor indi-

cation of amplitude of reception in this case

Pulses, ranging from the broadest, of 3 Mc/s

width, to so-:to of the narrowest, of 0~5 Mc/s width,

have been observed using the broader swoop rancos.
The pulse amplitude at both these extremes can be quite
larog. The wide pulses appear to be smoothly shaped for

the most part, with occasional exceptions, as sho:m in

the e.camoples







The appearance of broad-bond pulses sometimes ap*

pearin; to have erratic frequency versus amplitude
characteristics prompted the investigation of the noise

bursts at one Mc/s sweep width* The characteristics

revealed by observations at this swccp width will. be

discussed later. oi:.1plen of the pulses appear in the

series of figures at the end of this chapter. eorios 19

dopicts a sharp pulse at 17.5 'c/s. The apparent bifur-

cation amounts to about 0.1 Mo/s. In Series 20 the

activity is predominrcntly at the loowr end of the band.,

Some tendency to peak formation can be noted., In Series

21 the activity is across the band, and again pcaks

0.1 Mc/s apart appear to be present* In Series 22,

peaking continues, but some frames seem to have moderate

activity with no peaking visible. Series 23 likewise

exhibits cambiguious characteristics, with some frImeS

indicating poeklng, other frames having s trong localized
radiation at a particular frequency. Finally, Series 24

presents about 22 seconds of continuous activity, some

of full-scale amplitude and showing peaking t'e-ndncy.
Since the apparatus, as described above, is knoVwn to have

some minor gain irregularities at larger frequency sepa-

rations, the possibility of peaking in the apparatus
nuat be explored thoroughly before this feature can be

ascribed to the source* Some of the features of Jupiter

noise observable at one Mc/s sweep width are considered

below,







B. Analyssi of peaking

On three occasions to dato, sufficiently long recep-
tion vns obtained at 1 Nc/s vuccp width so that a tiLc.

analysis of pulse position could be made. Thcco dates

are April 18, April.25, and April 30, 1961, Of those
dates, the most.extensive records were obtained on April
30, but upon c Mmanation, they proved the least suscep-

tible to analysis, first bocaucs of station interference,
and second, because of a lack of regularity of peaking

that was not observed on the two previous occasions.

The characteristics of the pulses soecned to follow the
anro pattern on each of the first two ni"htz. The peak

separation is of the order of 0,1 Mc/s and the pen:ks
appear to drift dInvward in frequency, taking the order
of 10 minutes to drift 0.1 Mc/s. Tables 5 and 6 give
the pertinent information concerning these pen!:s for the
two nights. The least-equarcs line through each of the

peak lines has been calculated for the first two nights.
This was not done for the third night because of
definition problems as Stated above. The valueo of fre-
quency difference arc obtained by taking frequency dif-

ferences between sets of data points, and are not ob-
tained by differences in frequencies of the lonct-squaros

lincs. Figure 15 illustrate the peak drift appearance

for April 25, 1961. hote that a peak crosses the 18.00

Mc/s line at about 5.65 minutes. Figure 16 is a photo-
,rrph of the Brush high speed pen record for this fro-

quoncy, nonr this time. The lower channel is 18.00 Ic/a.







TABLE 5


PEAK SEPARATION AND DRIFT APRIL 18, 1961
PERIOD: o005 0510 E.S.T.
SYSTEM III LONGITUDE: 2670 2700


Initial Peak Position (Mc/s) Peak Separation (Mc/s) Least Squares Slope (Mc/s/mnin)

17.949 0.094 0.019 (s.d.) -0.063
17.864 0.108 0.015 -0.159
17.755 0.111 0.026 -0.059
17.650 0.094 0.022 -0.021
17.564 -0.00o







TABLE 6
PEAK SEPARATION AND PRIFT APRIL 25, 1961
PERIOD: 0440 0448 E.S.T.
SYSTEM III LONGITUDE: 2250 2300


Initial Peak Position (Mc/s) Peak Separation (Me/s) Loast Squares Slope (Mc/s/min)

18.214 0.090 0.014 (s.d.) -0.077
18.134 0.087 0.010 -0.094
18.073 0.093 0.012 -0.130
17.967 0.099 t 0.011 -0.112
17.873 0.075 0.012 -0.116
17.802 0.084 0.016 -0.130
17.723 0.089 0.016 -0.139
17.627 0.077 0.017 -0.119
17.536 -0.079













FREQUENCY SHIFT OF PEAKING APRIL 25,1961



18.20 = -----------. ___





18.00






Z
z


S17.60




17.40
TIME 2l4 6 7
TIME (minutes)


Fig, 15.--IwPeuenoy iDeift of Amplitude Peaks of Noise Pulases



















CHAfRT NO. BL 009 BRUSH ELECTRONIC COMPANY Jm..I CHAfRT NO. tL 909 BRUa








-T




1



.::'"1 EICTRONICS COMPANY umI'r , us CHART NO. BL 009 BRUSH ELECTRONICS COMPANY

/L iit I I /I .


Fig. 16.-Bi'ush Rooowd of Timse of Peak Crossing at 18.00 No/s&







A large amount of activity may be noted near this time.

Of course, this does not definitely establish the e:is-

tance of the peaks in the received poor, but it is a

strong indication in that direction. Many other records

were examined at points where the sweep record indicated

a crossing of the fixed-frequency record, and no gross

discrepancies could be found. It soeas highly probable

that the peaks do represent an actual characteristic of

the radiantion.

Further evidence for the reality of this peaking

is at hand. In the two channel frequency separated

records, the high proportion of time when there was

small correlation between channels was noted (See page

35). At times when the correlations did occur, a
tendency of the amplitudes of the pulses to change in

relationship to orch other was shown to o:ist, under

some circumstances This shifting of amplitudes fits
into the picture dravn above. If the two frequencies

are separated by a distance slightly more than the

distance between peaks, then the ratio of the amplitude

at the lower frequency to that at the higher frequency
should increase during the time of concurrence. Con*

versely, a frequency separation slightly less than the

distance between peaks should lead to the opposite

effect, a decrease in the ratio,

The discussion in Chapter I (See pngo 1 indicated
that work done in Florida and Chile simultaneously showed







a scintillation offset occurring on the incoming radin
tion from Jupiter (12). This scintillation seemed to
have two components; the cooponcnt of long duration,
of the order of a minute, is of interest here. It

appears from the above data that part of this grouping
of pulses observed in Jupiter noise storm records is
due to a frequency shift of the peaks of radiation
intensity. The single-channel record gives an indi-
cation of the passage of this pon through the monitored
frequency by sho'vinC a Croup or burst of relatively
high intensity pulses.


C, A working model
It is now necessary to oramine the spectral charac-
teristics of the radiation brought forth in this paper
and see if they fit a model based on synchrotron radia-
tion.
Pick as the frequency of max.inltr~ radiation 18 Mc/s.
It is seen from Equation 2-23 and Appondix B that this
corresponds to a value of




From Equation 2-22, then

n,, S^ ^
(5-2)

In the previous discussion of the nodclj it was pointed
out that the angle oc goes to 90 at the mirror point*







Also, because a largo amount of radiation must be o:.ittod
in one direction, it is asznmcd that the radiation re-
ceived is that cn.itted by the electrons near the mirror
point, 1horo their velocity paralllto the field is
saall* The radiation at other points would be veakh in
comarison, because' of the larCor forward velocity of
tho o:mittinG particles. The value
S51 Ine
vwillJ be used on this basis, The cyclotron frequency,
together with Equation 5-2 will yield a value for rmag-
notic field stronghb and a value for e
The two equations are

n o (5-3)




Numerically, take

^z.Z^ ^ ( )-5)

The observed data give two indications of harmonic
content. The bifurcation noted first in 1960 for strong
pulses, and again in 1961 on occasion, has pcaks sepa-
rated by about 0.4 M.!c/s. The records taken at one lic/S
sweep width in 1961 have peaks with a separation of
about 0.09 Ic/s as a lower limit. This yields for
values of n18 value of

Idf 0.o01= "oo
fjl f 0 Y ,5







hbich in turn loed to


Fi.-o, = O.i


and values of Bo of



S~3 _oF 0 .= auss


The poor in each harmonic is proportional to

F(n/nc),. For 27 Hc/s radiation, the harmonic number
ratio to n. is

_n_- = o0.3 X 20- ,qS


The ratio of power at.27 and 18 Me/s should bq from
Appendix B

7_ 0, q96

This ratio is dependent only on the assumption that
the radio noiso is synchrotron radiation from highly
relativistic particles.
Figure 1 shops that the ratio of radiation in-
tensities must be very much smaller than this in the
majority of noise storms* This is indicated by the
relatively low probability of occurrence averaged on
a yearly basis at 27,6 ;.!c/s as coparsod to 18 :lc/s.
The observed data do not fit the theoretical expec-
tations at all in this respect,







If it is assumed that the radiation is, indeed,
from highly relativistic particles, additional assump-
tions must be made concerning its characteristics. Two
factors brought up so -fr could alleviate the problem.
Onter chows that low harmonic number radiation takes
place over largo angles, The radiation reaching the
earth could be that emitted at a largo anglo to the
plane of gyration of the electron. If this were the
case, as the region of high field intcnsity rotated to
the control meridian, one vould oxpoct the radiation
received to move up in frequency; then, as rotation
continues, to move down in frequency. As shown in
Table 4, this type of drift has been noted, but the
longitude correlation is not good. Also, it is not
the predominant tendency shown in the t'b-lo. The num-
ber of occurrences of up and dovn drift Should be about
equal, with more up drifts at onset and more down drifts
at termination* This also is not noted. The only other
orplnnation brought forward here is that at these low
froquencion, the radiation is coherent, and, as pointed
out earlier (pago 30) interference offocts can give a
sharp tailing off of radiation with frequency.

The range of /3= involved in the two limits set
forth here is




1 0-. 0.07= 0.?2







Tho first v.luc corresponds to a sliCghtly larger value

than the lower limit place on S in Chapter II, whilo
the upper limit would obviously better fit the theory

outlined in that chapter. Because of the reduced field

strength necessary to fit the nar-.inu frequency, the

electron donasty remains about the same. The charac-

teristic most often pointed out in the literature as

suggosting a synchrotron radiation model is the re-

duced peak width at larger frequencies that is noted

in the probability histograms. The rapid change in

this width with frequency seems to indicate a small

bari-onic nuiiber for the r-diation. Because of the

ultra-sharpness of the radiation at high harmonic

numbers, however,, one may take the othior stand, and

say that the rapid chance is due to a nranification

of the small-anglo differences in these harmonics by

the rotation of the planet.

The polarization of the radiation may be compared

to Equation 2-19. The ratio of the polarization aeoo

is a strong function of the anglo between the plane

of motion and the direction of the observora Hloto

that for an observer in tho plane of motion, the radi-

ation is plane polarized.

The polarization of individual pulses during a

single storm fluctuates over wide values, but by
taking a running average of the axiol ratios, this

fluctuation can be smoothed out (11) Slower, more

regular fluctuations are brought out by this process,







It may be supposed that over a short period of time, the

radiation is omitted with a certain mo.n polarization.

This would indicate that tho radiation is either omitted
by electrons of given ( at a given angle, or that the

values of e and Y for the electrons contributing to the

rocoived radiation chance so as to keep.the mean value

of the axial ratio almost constant. If the potai! noted
earlier arc a result of synchrotron radiation, then the
first case must hold, and an e::trcm!ly strong velocity
separation nust occur at the point of maximum radiation.

That is, if the electrons are moving along a curved
field lino, those of a particularvelocity ran-o must
gyrate in a plane sufficiently roeovod from the.planes

of Gyration of electrons of other velocity ranges to
prevent soaring out of tho harmonic content in fre-

quency, Because the angular spread of radiation is
smallor at higher harm-onics, this would be more likely

to show up at the 200th harmonic than at the 4Sth
harmonic, so, again, the hiChor enorgies are preferred.

On the basis of Equation 2-19 and the first assumrption,
lower-frequency radiation should show incronsed axial
ratio, by the rcasonin.g gven below
The argument of the modified Bessel functions Kxy/
and Ky, is



which for g=o.w&, and -=o.o gives a value of
-A-. ( fY) t oz
3sind







This results, from Appendix B, in an axial ratio of 6/1.

Substituting a value of =o0.1 gives an an-ial ratio of

4/1i At an;les of about 0.1 radiant tho axial ratio for
the hi,;hor-onorgC particles is about 1.f/l. At these

larger harmonic numbers, angles greater than that men-

tioned are less likely because of the smallness of the

radiation cone,

If, in a typical storni, the radiation picked up
on earth is radiated from electrons whose plano of

gyrntion is tilted 0.1 radians away from the line of
sight of the observer, the axial ratios of the ob-

servod polarization ellipses would bo in the raneo

meosurod. Two effects that should be noticeable in

this model are the increase in axial ratio with de-

creasing frequency, but at an exceedingly slow rate

because of the cmTll chanSos in the ratio of the modi-

ficd Bosscl functions; and the drift to lower frequency
of the energy peaks-across the polarimeter frequency

should produce an observable effect on the axial ratio,
The peaks of radiation should have a rather con-
sistent polarization n::ial ratio, since they rcprcoont the

radiation harbnonics of a group of electrons of similar

energy gyrating in the same region of the magnetic
field. The axial ratio of the polariscd radiation

emitted by these particles should be within a small
range of values corresponding to the enorgy range

within the grouping. The axial ratios of the radiation

emitted by particles outside these groupings should







differ from thobe within it. ,Thus a typical polariza-

tion ratio should occur as those oner-:; roupings cross
the polnriuoter frequency, and different values, por-

haps rnndon, should occur betwoen those groupings as

the result of radiationfrom sporadic particles not

rlo:-ibers of a definite group.

One more point should be brought in concerning

tho poking effect* Table 7 lists the frequencies and

tiro3 of radio noise reception during the periods covered

by the peaking data, If the separations noted in the

peoks correspond to the cyclotron frequency, and the

slope indicates a decrease in enorgy, thbi indicates

that theoelectrons penetrate less far into the high-

field rozion in such proportion to their energies that

the cyclotron frequency decreases, instead of what

would be observed in a constant field, where a loss of
onorGy would lead to an increase in the cyclotron fre-

quency. The peaks appear to be slightly less far apart

on April 25 than they did on April 18, indicating on

the basis above, that the electrons wore sliChtly less

energetic. The figures in the table indicate that on

the IatLer day, the lower froqucncios were active,

while the 22.2 Mc/s antenna did not pick up anything.

On the first day, the 10 Mc/s antenna did not pick up
anything.


D. Conjectural note

To complete the picture presented above, some effects






TABLE 7


RADIO IDISE RECEPTIOIT DURING THE PEAK ANALYSIS PERIODS
TIMES IN E.S.T.


Preq. April 18, 1961 April 25, 1961 April 30, 1961
Fla. Cbile Fla. Chile Fla. Chile

10 0232-0251 0317-0640
044-0501
0650-0655
15 0454-0517 0458-0519 0340 406o-0416 0436-051i o0400504
0444-o515 0418-0428 0530-0545 0535-0605
o040-0453
18 o454-o517 0453-0521 0242;o413 0419-0447 0400-0417 0400-05o00
0415;0430 0527-0545 0635-0645
o434-o45o os56;o610
20 0251-0253 o446-045o o047-o455
0450-0515. 0647-0654 0635-0647
22.2 0454-0517 0448-0514 0506;0513
051 o04o8-o452 04o0-o045







may be postulated. The downward drift in frequency,

which seems to predominate, may be explained in terms

of the electron oner:,y. If unusually onergeticelectrons

enter the radiation belt of Jupiter so that they go into

the outer fringes of the atmosphere, the result would be

as observed. The more highly cnergotic electrons would

emit radicntion with a naximt~ r intensity at a higher har-

monic numbor, Electrons emitting at a peak cnorgy at

higher frequency would either lose energy or be lost in

collision. They would penetrate further into the region

of stronger magnotic field and at some point, enter the

upper fringes of the atmosphere. Those electrons that

did penetrate this far should be lost from the radiating

group*
If the electrons are accelerated by the mechanism

proposed by Forii (23) in the inhomogcncous magnetic

field between the orbits of the earth and Jupiter (30),
this would lead to an explanation of the decrease in

activity with increased solar activity. This inhomo-

geneous magnetic field is pushed further out during

periods of high solar activity, so that there would be

less accelerating space available to the electrons.

They would be less likely to attain high energies at the

orbit of Ju-pitcr under such conditions. Lastly, the

fact that very little left-handed polarized rnlintion

has been observed suggests the possibility that just

one source is responsible for all radiation, and that




83


radiation is somotirncn picked up after 3rfectlon at

lareo anr-lcs fromi Jupitert's onoslilbcrc.































I I I I 1 i
17 18 19
FREQUENCY (Mc/s)


17 IB 19
17 18 19


Fig. 17.--Series 1, April 9, 1961. 0435:30 E, S. T.
(Time progresses reading down the columns and to the right
at the rate of one-fourth second per frame* First frame
shows galactic background no Jupiter radiation. Level
change indicates radio noise from Jupiter, Third column
has strongest Jupiter noise present, with a maximum at
17.5 Mo/s)


W


17 18 19


ProM


PM













I"
I"
wBW'^B


r^ n


I I I I I
17 18 19
FREQUENCY (Mc/s)


Pig. 18.--Setries 1, page 2,


"J U


-= 1


17 1 19
17 fa 19


17 1 I 1
17 18 '9


- 1


]Cna


^ -I



















Sa m -,. .- ,


17 18 19
FREQUENCY (Mc/s)


17 1 1 9
17 18 19


Pig. 19.-.Series 2, April 27, 1961. 0439100 E. S. T.


17 18 19












2 m


U m


U


m I


I-


U U


I I I i I I I
17 18 19
FREQUENCY (Mc/s)


I I 1 1I
17 18 19


Fig. 20o.-Series 3. April 25, 1961l 0439t20 E. S. Ti


17 1 9I
17 18 19


pM! I


I P B "


6mm









A m


", N


i 1 I I iiL l I. I Ii l l1 I
17 18 19 17 18 19 17 18 19
FREQUENCY (Mc/s)


Pig* 21.--Series 3, page 2.


11


F~a31


I


IPow


-I 1 111


601mid I


io 60 Wl





















































17 18 19
FREQUENCY (Mc/s)


Fig. 22.*.Series 4. April 27, 1961. 040S20 E. S. T.


WON.




92






U-- -- RK^














17 18 19 19 17 19 17 18 19
FREQUENCY (Mc/s)


Pig. 2)..-Series 6. April 27, 1961, 041i140 E. S. To




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs