Analysis of the decameter-wavelength radio emission from the planet Jupiter.

MISSING IMAGE

Material Information

Title:
Analysis of the decameter-wavelength radio emission from the planet Jupiter.
Physical Description:
xv, 272 leaves : ill. ; 28 cm.
Language:
English
Creator:
Six, N. Frank ( Norman Frank ), 1935-
Publication Date:

Subjects

Subjects / Keywords:
Radio astronomy   ( lcsh )
Observations -- Jupiter (Planet)   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 262-271.
Statement of Responsibility:
By Norman Frank Six, JR..
General Note:
Manuscript copy.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000561601
notis - ACY7535
oclc - 13549055
sobekcm - AA00004951_00001
System ID:
AA00004951:00001

Full Text










ANALYSIS OF THE

DECAMETER-WAVELENGTH RADIO

EMISSION FROM THE PLANET JUPITER










By
NORMAN FRANK SIX, JR.


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
April, 1963










ACKNOWLEDGMENTS


The author is deeply indebted to the Chairman of his Graduate

Committee, Dr. Alex G. Smith, for his guidance and encouragement

throughout this entire research project. He also wishes to thank

Dr. Thomas D. Carr for helpful discussions and suggestions, and

Drs. D. C. Swanson, F. E. Dunnam, and J. T. Moore for serving on

his committee.

The analysis described in this thesis would not have been

possible had it not been for the assistance of the staff members of

the Florida and Chile radio astronomy stations in carrying out a

routine observing program. These colleagues are:

W. F. Block R. J. Leacock

H. Bollhagen J. Levy

T. D. Carr J. May

N. E. Chatterton D. M. Newlands

R. S. Flagg A. G. Smith

T. Hlaing J. E. White

Acknowledgment is due the Central Scientific Industrial and

Research Organization of Australia for permitting T. D. Carr to use

their equipment for Jupiter observations, C. L. Seeger of the Stan-

ford Radio Astronomy Institute for supplying L. Cunningham's ephemeris

of Comet Seki in advance of the regular notices, W. F. Block for his

help with the comet program, and G. W. Brown for pre-publication infor-

mation concerning flux density studies of the Jovian emission.









The programming efforts of W. L. Howell and F. D. Vickers

have been an invaluable aid, and the help of W. L. Cain, M. L. Fagerlin,

E. J. Lindsey, and W. Mock in preparing the illustrations is greatly

appreciated. H. W. Schrader lent photographic assistance and Mrs. T.

Larrick has done a marvelous job in editing and typing the manuscript.

The writer is grateful for the financial support, in the form

of research assistantships, supplied by the Office of Naval Research,

the U. S. Army Research Office Durham, and the National Science

Foundation during the progress of this study.

The encouragement provided by the writer's parents has con-

tributed significantly to the success of this work.

The author's wife deserves much more than gratitude for her

patience, understanding, and sacrifices throughout eight years of

college life. It is to her that this thesis is dedicated.













TABLE OF CONTENTS


ACKNOWLEDGMENTS ....... .......... ..... .

LIST OF TABLES .................. .. .

LIST OF FIGURES .......... ...... .

CHAPTER
I. INTRODUCTION ... .............. .

Planetary Observing Program with Emphasis on Jupiter .

The Decameter-Wavelength Range . .
Kinds of Radiation . .....
Types of Studies . . .
Instrumentation .. ....... ........ ..

II. JUPITER DATA IN 1961 AND COMPARISON WITH PREVIOUS YEARS .


Page
ii

vii


Analysis ........ .


Raw Data . . .
Jupiter Program for the IBM 709 . .
Merge Program for the IBM 709 ............

Gross Statistics 1961 . .

Source Studies . . .

1961 Probability Histograms . .
Merging the Data from Different Stations .
Search for Long-Term Changes . .
Search for Station Effects . .
Combining the Data of Different Apparitions
and Different Stations . .
Conclusions Regarding Number, Locations,
Separation, and Width of the Sources .
Intensity Histograms . . .

Activity Studies . . ...

1961 Activity Plots ..... ..
Activity Plots for 1957 1960 . ...
The Angular Rate Effect . .


* *o









TABLE OF CONTENTS--Continued


CHAPTER


The Distance Effect ... .
The Elongation Effect . .
Evidence of Particle Stream Deviation
by the Earth's Magnetosphere .


III. SOLAR CORRELATION IN 1961 AND COMPARISON WITH
PREVIOUS YEARS . .


Long-Term Inverse Correlation with the Sunspot Cycle
Short-Term Correlations .....


Sunspot Number . .
Solar Flare Program for the IBM 709 .
Results of the Chree Analysis of the
1961 and 1960 Data . .
Geomagnetic Activity .....
Polar Cap Absorptions ...........
Overall Solar Activity . .
Conclusions Regarding Short-Term Correlations


IV. ORIGIN OF THE NON-THERMAL JOVIAN RADIO EMISSION .

Characteristics of the Radiation . .


Observed Frequency Range .
Temporal Behavior . ...
Dynamic Spectrum . .
Polarization . ..
Correlation with Rotation of Jupiter .
Overall Spectrum . .
Long-Term Variability . .
Source Dimensions . .

Theories of Origin .............

Jupiter's Microwave Emission .
Jupiter's Decameter Emission .


* e o

eeoee
eeee


V. A SEARCH FOR DECAMETER RADIATION FROM COMET 1961 f ..

Nature of Cometary Activity . .

History of Radio Observations of Comets ..


.Radio Observations of the Close Approach
of Comet Seki . ....


9 9 0 9








Page


@


100
102

108


127

127
136

136
139

152
172
176
177
178

180

180

185
187
196
200
204
207
210
215

213

214
215

227

228

230


232



,







TABLE OF CONTENTS--Continued

CHAPTER Page

Description of Comet Seki .......... 232
Orbital Elements ............ . 232
Ephemeris . . 235
Observing Program ............... 236
Evaluation of the Data .......... ... 247
Conclusions . . .. 256

VI. SUMMARY . . . 257

LIST OF REFERENCES . . .. 262

BIOGRAPHICAL SKETCH . . . 272













LIST OF TABLES


Table Page
1. Types of Radio Emission from Jupiter . 3

2. Gross Statistics 1961 . ..... 18

35. Circumstances of Events Which Were Recorded at One
Station Only, When the Other Station Was
Listening Effectively ................. 19

4. Location and Width of the Decameter Sources from the
Merged Histograms . . .. 50

5. 1961 Data: Location and Width of the Decameter Sources 57

6. Location and Width of Source A: 1957 1960 Data .... 58

7. Zenocentric Inferior Conjunctions of Mercury .. 120

8. Heliocentric Latitude Differences at Zenocentric
Inferior Conjunctions . ....... 125

9. Lag Times for the Transit of Solar Particles 124

10. Heliocentric Coordinates of Jupiter . 130

11. Constants Used in the Determination of the Activity
Index of a Solar Flare ................. 150

12. Chree Analysis Table ................ 151

15. Peak Days of Jupiter Emission during the 1961 Apparition 153

14. Peak Days of Jupiter Emission during the 1960 Apparition 154

15. Solar Particles and Jupiter Emission around Opposition
in 1961 ................ 175

16. Polar Cap Absorptions in 1961 . 177

17. Characteristics of the Non-Thermal Radio Emission
from Jupiter . . 181


vii








LIST OF TABLES--Continued


Table Page
18. Ephemeris for Comet 1961 f by L. Cunningham ..... 235

19. Results of the Analysis of Comet Seki Observations
Made from Florida ............... ... 248


viii














LIST OF FIGURES


Histograms of the Florida 1961 data .....

Histograms of the Chile 1961 data .


5. 1961 histograms of the 18 Mc/s Florida data .


10 Mc/s histograms 1961 .

15 Mc/s histograms 1961 .

18 Mc/s histograms 1961 .

22.2 Mc/s histograms 1961 .

27.6 Mc/s histograms 1961 .

20.0 and 19.7 Mc/s histograms

18 Mc/s histograms 1960 .

22.2 Mc/s histograms 1960 .

10 Mc/s histograms .

18 Mc/s histograms .

20 Mc/s histograms .

22.2 Mc/s histograms ..

27.6 Mc/s histograms .

10 Mc/s histograms, data of d:

18 Mc/s histograms, data of d:


1961 ...........

. . .

0 . 0




Lfferent

Lfferent


years combined .

years combined .
years combined ...


22.2 Mc/s histograms, data of different years combined .

27.6 Mc/s histograms, data of different years combined .

10 Mo/s histogram of Chile and Australian data 1960-1961

15 Mc/s histogram of Florida and Chile data 1961 ..


Figure
1.

2.







Page
22

23

25

27

28

50

31

32

33

34

35

57

39

41

42

43

45

47




.









LIST OF FIGURES--Continued


Figure
23. 18 Mc/s histogram of Florida and Chile data
1957-1961 . . .

24. 20 Mc/s histogram of Chile data 1960-1961 . .

25. 22.2 Mc/s histogram of Florida and Chile data
1958-1961 . . .

26. 27.6 Mc/s histogram of Florida and Chile data
1958-1961 . . .

27. Location of the decameter sources as a function of
frequency . . .

28. Coordinate system attached to the Jovian surface .

29. Variation of source width with frequency .

30. Geometry of radio reception of Jovian outbursts assuming
a Jovian ionosphere and latitude separation of the
frequencies emitted from a single source .


31. 18 Mc/s intensity histograms . .

32. 1961 intensity histograms . .

33. Jovian flux density histogram calculated using the
Chile, 18 Mc/s data of 1961 . .

34. 15 Mc/s activity plots of 1961 data ...

35. 18 Mc/s activity plots of 1961 data .

36. 22.2 Mc/s activity plots of 1961 data .

37. 27.6 Mc/s activity plots of 1961 data .

38. 10 Mc/s and 20 Mc/s activity plots of 1961 data .

39. 18 Mc/s activity plot, 1961 data of Florida and
Chile combined ... . .

40. 22.2 Mc/s activity plot, 1961 data of Florida and
Chile combined . . .

41. Activity plot of Florida 18 Mc/s 1960 data .

42. Activity plot of Chile 18 Mc/s 1960 data .


Page


53

54


55


56


59

61

66




69








r













r













LIST OF FIGURES--Continued


Figure Page
43. 22.2 Mc/s and 27.6 Mc/s activity plots of Florida
1960 data . ......... 91

44. 20 Mc/s and 22.2 Mc/s activity plots of Chile 1960 data 92

45. 10 Mc/s and 16.7 Mc/s activity plots of Chile 1960 data 953

46. Activity plots of Florida 1959 data . 94

47. Activity plots of Florida 1957 and 1958 data 95

48. Angular rate geometry . . 97

49. The apparent angular rate of Jupiter as seen from the
Earth in seconds of arc per day (top), and the dur-
ation of Jupiter's decameter storms at 18 Mc/s
(below), during the 1961 apparition . 99

50. 1961 Florida data showing the elongation effect. The
activity values have been adjusted to correspond to an
Earth-Jupiter distance of 5 A.U. ........ 104

51. 1961 Chile data showing the elongation effect. The
activity values have been adjusted to correspond to an
Earth-Jupiter distance of 5 A.U .. ... 106

52. Elongation effect during 1957 1961. All curves are
from the Florida data except the 1961 22.2 Mc/s curve.
Activity values have been standardized to 5 A.U. 107

53. The earth's magnetosphere (taken from reference 36) 109

54. Cavity carved out of the solar stream by the earth .. ill

55. Variations in Jupiter activity at the Chile station
on 18 Mc/s during the period from 30 days before
to 30 days after opposition ... .... 114

56. Variation in Jupiter activity around zenocentric
inferior conjunction of Venus . 118

57. Variation in Uupiter activity around zenocentric
inferior conjunction of Mercury . ... 119

58. Ten-day activity plots of the 18 Mc/s Florida data of
1961 showing the dates the planets were at zenooentric
inferior conjunctions ................ 122








LIST OF FIGURES--Continued


Figure Page
59. The long-term inverse correlation of the occurrence
of decameter radiation with the sunspot cycle 129

60. Comparison of the variation in decameter source
width at several frequencies with the sunspot cycle 131

61. Jupiter-Sun geometry and the sunspot belts . 1533

62. Apparition activity index rate and magnitude of the
solar latitude of the sub-Jovian point plotted as
a function of the mean epoch of the observing season 134

63. Daily activity index rate (from Figure 39) and sunspot
number variation during the 1961 apparition of
Jupiter . . .. 157

64. Monthly average sunspot number and monthly activity
index rate for the 1961 apparition .. .. 138

65. The heliographic coordinate system ........... 140

66. e before and after opposition . 143

67. Geometry involved in the determination of 9 (the angle
between the Sun's meridian as viewed from Jupiter
and the Sun's meridian as viewed from Earth) for
December 26, 1959 ................... 145

68. Geometry involved in the determination of 9 (the angle
between the Sun's meridian as viewed from Jupiter
and the Sun's meridian as viewed from Earth) for
August 6, 1960 ........ ....... .. 146

69. Regions on the solar disk as viewed from Jupiter
determining the designation of flares as belonging
to groups 1, 2, or 3 ........... 147

70. Chree analysis of Jupiter activity, geomagnetic index
Ap, and sunspot number Rz, using the 20 peak days
of 18 Mc/s Jupiter emission monitored at the Florida
station in 1961 .... ..... 156

71. Chree analysis of solar flare activity index in groups
1, 2, and 3, using the 20 peak days of 18 Mc/s
Jupiter emission monitored at the Florida station
in 1961 . o . 157









LIST OF FIGURES--Continued


Figure Page
72. Chree analysis of solar flare number in groups 1, 2,
and 3, using the 20 peak days of 18 Mc/s emission
monitored at the Florida station in 1961 ....... 158

73. Chree analysis of Jupiter activity, geomagnetic index
Ap, and sunspot number Rz, using the 20 peak days
of 27.6 Mc/s Jupiter emission monitored at the
Florida station in 1961 . .. 160

74. Chree analysis of solar flare activity index in
groups 1, 2, and 3, using the 20 peak days of
27.6 Mc/s Jupiter emission monitored at the Florida
station in 1961 .. ........ 161

75. Chree analysis of solar flare number in groups 1, 2,
and 3, using the 20 peak days of 27.6 Mc/s Jupiter
emission monitored at the Florida station in 1961 162

76. Chree analysis of Jupiter activity and geomagnetic
index Ap using the 20 peak days of 18 Mc/s Jupiter
emission monitored at the Florida station during
the 3 months around opposition in 1961 ...... .. 163

77. Chree analysis of Jupiter activity, geomagnetic index
Ap, and sunspot number Rz, using the 20 peak days
of 18 Mc/s Jupiter emission monitored at the Chile
station in 1960 ..... ........... 165

78. Chree analysis of solar flare activity index in
groups 1, 2, and 5, using the 20 peak days of
18 Mc/s Jupiter emission monitored at the Chile
station in 1960 ............ 166

79. Chree analysis of solar flare number in groups 1, 2,
and 3, using the 20 peak days of 18 Mc/s Jupiter
emission monitored at the Chile station in 1960 167

80. Chree analysis of Jupiter activity and geomagnetic
index Ap using the 20 peak days of 18 Mc/s Jupiter
emission monitored at the Chile station during the
3 months around opposition in 1960 .... .. 168

81. Geomagnetic index Ap, Jupiter activity index rate,
geomagnetic storms, polar cap absorptions, and
overall solar activity in 1961 .. 174

82. The frequency of occurrence of Jovian radio noise as a
function of frequency ............. 185


xiii








LIST OF FIGURES--Continued


Figure Page
85. Typical low speed recording of a Jupiter noise
storm on 18 Mc/s ....... ........ 188

84. Comparison of short and normal pulses of Jovian radiation at
at 18 Mc/s .................. ... 190

85. Poor time correlation of pulses on the Florida and Chile
records of March 24, 1960 ............. 192

86. Period of good correlation from the records of March 29,
1960, taken at the Florida and Chile stations. 193,

87. Alternate fading or out-of-phase scintillations from
the records of March 23, 1960 ........ 195

88. Build-up and decay of an ordinary pulse of Jupiter noise 197

89. Spectrograms of two Jovian noise pulses . 198

90. Dynamic spectrum of a Jovian pulse exhibiting fine
structure ........ .. ..... 199

91. Individual frames from the spectra of two bursts of
short pulses which occurred several minutes apart 201

92. Average axial-ratios of polarization ellipses at the
two stations on different dates in 1961 ....... 205

93. The Jovian microwave halo and the circulation of the
plane of polarization ................ 206

94. Spectral distribution of averaged decameter peak flux
densities from Jupiter in 1961 ............. 209

95. Spectral distribution of Jovian microwave flux
densities (4). ....... ..... ... .... .. 211

96. Coordinate system of gyrating electron ......... 223

97. U. S. Navy photograph of Comet Seki on October 18, 1961,
taken by Dr. E. Roemer .......... 233

98. Elements of the orbit of Comet Seki 1961 f ....... 234

99. North polar view of Comet Seki's path ........ 237

100. 18 Mc/s yagi antenna at Gainesville, Florida ..... 238


xiv











LIST OF FIGURES--Continued


Figure


101. East section of the 22.2 Mc/s interferometer at
Gainesville, Florida . .

102. West section of the 22.2 Mc/s interferometer at
Gainesville, Florida . .

1035. 22.2 Mc/s interferometer pattern . .

104. Alt-azimuth system showing Comet Seki's position in
the sky as seen from Gainesville, Florida ..

105. Zenith view from Chile showing Comet Seki's position
in the sky . . .

106. 18 Mc/s scan record of Comet Seki taken in Florida

107. 22.2 Mc/s interferometer records taken in Florida
on November 14 and 16, 1961 .

108. 22.2 Mc/s interferometer records taken in Florida
on November 17, 18, and 19, 1961 ....


r







r














Page


240


241

242


244


246

251


252


254













CHAPTER I


INTRODUCTION


The discovery that Jupiter is a strong, intermittent source

of non-thermal radio-frequency energy was made in 1955 by Burke and

Franklin (1). The following year the University of Florida Radio

Observatory began systematic observations of Jupiter at a frequency

of 18 Mo/s. Since that time, the program has been expanded to include

radio-frequency monitoring in the range between 5 and 51 Mc/s. Jupiter

has not been the only subject under investigation. Observations have

been made of Saturn, Uranus, Venus, and Mars, and in 1959 a southern

hemisphere field station was established at Maipu, Chile.



Planetary Observing Program with Emphasis on Jupiter


The Decameter-Wavelength Range

The radio astronomy group at the University of Florida has

concerned itself principally with planetary radiation in the "tens-

of-meters" wavelength range. Observations of Jupiter, Saturn, Uranus,

Mars, and Venus have been made. The results are negative for Uranus,

Venus, and Mars. (From records taken in 1960, we concluded that Venus

did not emit non-thermal decameter-wavelength radiation of flux density

greater than 5 x 10-22 watts/m2/cps (2).)









The results for Saturn are inconclusive. In 1960 there were

seven weak events possibly of Saturnian origin. Matters were compli-

cated by the fact that Jupiter and Saturn were separated throughout

the apparition by approximately one hour in right ascension. Again

during the 1961 apparition there were several occasions of possible

Saturn radiation, none of which was conclusive. With the greater sep-

aration of Jupiter and Saturn relaxing the resolution requirement, it

is hoped that the 1962 observing season will establish the existence

or nonexistence of decameter-wavelength radiation from the ringed

planet.

Jupiter appears to be unique among the planets. As far as we

know, none of the other bodies which circle our sun (with the possible

exception of Saturn) are strong emitters of non-thermal radio noise.

Considering extraterrestrial sources, the powerful Jovian outbursts in

the decameter-wavelength range are exceeded in intensity only by the

sun. Thus, it is not surprising that this new avenue to information

about the giant planet has been the main concern of the University of

Florida group.


Kinds of Radiation

The radio emission from Jupiter is caused by at least three

distinct mechanisms. Table 1 divides the radiation into three com-

ponents according to wavelength and presents empirical evidence relat-

ing to the origin of each.

The apparent black body disk temperature is the temperature

that a hypothetical black body, -which subtends the same solid angle









03


03J
-_~-- 0-- -H



V d






I Ep i i o


I o 1 0 0




r- X rg M X !
F4 0 D


r-' o o o


SI I
-cc

S0 0
*i c3 j
W ;- ,' I (

0 2 a o 0 o o0
"i a) U i I
tdo oD o .


H A0 0 0 0 0 to 1
0 0 0 HH 0 -1
Cu 4 p o 0 0 CO
134 LO 0
"i w "<





00 o -

b- 0l KL 1( O0 |









| D C
F4 P g
H -14 to cn(



00 1
aP <(D < 0(D i




P 4a)4









as the visible disk of the planet, would have if it emitted thermal

radiation of the same intensity as the observed radiation. The values

for the apparent black body disk temperatures in the first two wave-

length regions in the table were obtained from reference (4). This

reference also contains a graph showing the variation in apparent disk

temperature with wavelength. The Rayleigh-Jeans approximation to the

Planck black body curve well describes the energy distribution of the

thermal radiation at radio wavelengths. The temperature 1450 K agrees

well with infrared measurements (6). The spectral distribution of the

two non-thermal components of radiation will be discussed later as

will the mechanisms involved.


Types of Studies

Since 1956, five major problem areas have shown up in the anal-

ysis of the decameter records and consequently these have received the

most attention. They are as follows:

1. localized sources

2. polarization

5. spectrum analysis

4. influence of the terrestrial ionosphere

5. solar correlations

The decameter radiation comes from "hot spots" which are local-

ized or fixed in longitude System III (2). These sources rotate with

a period of 9h 55m 29s35. If we plot the probability of receiving

Jovian radiation when a 50 increment of longitude is on the central

meridian of the illuminated disk versus System III longitude, the








localization of the sources is quite evident. There appear to be three

main sources and some of these show a tendency to split. The radiation

at different frequencies comes from slightly different longitudes and

at very low frequencies the source structure is lost altogether.

Polarization studies of the radiation have revealed that the

polarization does not occur in the terrestrial ionosphere, but is

characteristic of the Jovian signals. These studies provide evidence

for the existence of a magnetic field on Jupiter and lay a foundation

for testing parameters such as the magnetic field strength and the ion

density in Jupiter' s ionosphere, both of which are of fundamental con-

cern in trying to discover a "model" which explains the origin of the

decameter radiation.

Spectrum analysis yields information about the structure of the

individual pulses and bursts of radiation, and how they drift up or

down the frequency spectrum. Here again, knowledge of how the energy

is distributed with frequency is a must for determining what mechanism

is involved.

The influence of the terrestrial ionosphere must be classified

as one of the unsolved questions. From high speed recordings taken at

the stations in Florida and Chile at the same frequencies, it is obvious

that the ionosphere is altering the pulse structure. It is only on

rare occasions that we get pulse to pulse correlation in time. We

believe that inhomogeneities and clouds of ions in the ionosphere are

responsible. The differences in the axial ratios of the polarization

ellipses at the two stations are too great to be accounted for by the

measured ion densities in our ionosphere (7).









Solar correlations have been attempted to find if there is

any connection between the Jovian decameter radiation and the bursts

of particles and radiation from the sun. In some cases, correlations

have been found. The solar related variables under inspection have

been the following: sunspot number, solar flares, geomagnetic storms,

solar m-regions, and terrestrial polar cap absorptions.


Instrumentation

The microwave observations are carried out with large parabolic

antennas and low-noise amplifiers, since receiver noise is the prin-

cipal limitation. The decameter work is conducted with conventional

antenna arrays and standard communications receivers, the principal

limitation being "atmospherics" and man-made noise. The background

signal from the galaxy at the longer wavelengths is greater than the

receiver noise, so there is not much to be gained by going to low noise

devices.

The major equipr.nt at the Florida observatory includes two

18 Me/as Yagi antennas on equatorial tracking mounts to provide discrim-

ination between Jupiter and Saturn; a 27 Mc/s Yagi on an equatorial

tracking mount; an 8-element 18 Mc/s broadside array (a description of

a similar antenna is found in reference (8)); two 52-element 22.2 Mc/s

broadside arrays, which can be used as a lobe-switching interfero-

meter (9, 10, 11); a broad-band rhombic array for use with a swept-

frequency receiver, which displays any 4 Mc/s segment of the spectrum

up to 25 Mc/s on a cathode-ray tube (12); and a 22.2 Mc/s polarimeter

consisting of crossed Yagis on a steerable boom (15, 14). There is








also a 900 corner reflector, steerable in altitude and operable in the

frequency range from 14 to 31 Mc/s through the use of appropriate

dipoles (15).

The major instrumentation in Chile includes two steerable

18 'Mc/s 8-element broadside arrays, which track a planet automatically

by means of a motorized phasing device; 5, 10, 15, and 20 Mc/s broad-

side arrays; a 22.2 Mc/s polarimeter consisting of four folded dipoles

over a tilted ground plane; a 16 Mc/s crossed Yagi polarimeter on an

alt-azimuth mount; and a mechanically steerable corner reflector iden-

tical with the one in Florida.

Signals from all the antennas are amplified by commercial com-

munications receivers and recorded on pen recorders. A standard

recording speed of 6 inches/hour is used, with high speed recorders at

each station making auxiliary recordings at 5 mm./sec. during intervals

of planetary activity. Motion-picture recording of the scope display

of the swept-frequency receiver is employed during Jupiter outbursts.

The receiver-recorder systems are calibrated for absolute flux levels

by means of noise diodes. Time marks are placed on all records by

timing systems calibrated against the signals from WWV. Short-wave

transmitting equipment is used at both observatories for the inter-

change of data and instructions. An observer is on duty at each observ-

atory during all periods of recording to monitor the signals, eliminate

interference, and perform special experiments.

In addition to the observing programs conducted in Florida and

Chile, Dr. T. D. Carr of the Gainesville group spent the period of





8



June 17 August 30, 1961, in Australia monitoring Jupiter radiation

at the Fleurs field station of the Radiophysics Division, Commonwealth

Scientific and Industrial Research Organization. Instrumentation

included a 10 Mc/s fixed beam array of 2 full-wave dipoles and a 19.7

Mc/s linear array of 43 half-wave dipoles, a portion of the north-south

arm of the Mills Cross built by Shain (16). Two such arrays phased for

different directions were used to permit longer observing.












CHAPTER II


JUPITER DATA IN 1961 AND COMPARISON
WITH PREVIOUS YEARS


Analysis


Raw Data

Pen recordings of the radiation monitored in Florida and Chile

are taken on Texas dual channel recorders at a chart speed of 6 inches/

hour. These records are analyzed, and the following information is

taken from them: date, frequency monitored, observing station, begin-

ning and end of the listening period in Universal Time (hours and min-

utes, e.g., 0745), beginning and end of the Jovian activity period,

intensity of the activity, and pertinent remarks concerning such things

as listening conditions, degree of certainty in identifying the radia-

tion as Jovian, structure of the signals, and equipment troubles.

The procedures involved in obtaining some of the above informa-

tion need to be explained. Since quite frequently the observing periods

span midnight, Eastern Standard Time, more than one date would be

involved. This would make it difficult to correlate daily indices such

as sunspot number with the activity during a watch period. To avoid

the difficulty and others, the following convention regarding the date

of a watch period has been adopted. The 24-hour period beginning at

12 noon Eastern Standard Time, date x, and ending at 12 noon Eastern

Standard Time, date y (where y x + i), is designated the period of









date y. An observing period occurring any time in this interval is

called the watch of date y. It is extremely unlikely that the Jupiter

watches at either the Chile or Florida station will ever straddle 12

noon Eastern Standard Time.

The listening period is that part of the watch during which

receiving conditions are good enough to permit detection of Jupiter

radiation. Limits on the listening periods are also imposed by the

antenna patternsof the non-steerable arrays. During 1961 the effective

listening periods of the systems using the following arrays were

limited to + 4 hours from Jupiter transit of the station meridian:

5 Mc/s broadside, 22.2 Mc/s polarimeter, and the 27.6 Mc/s corner

reflector in Chile; the 15 and 31 Mc/s corner reflector in Florida.

The beam limits of the 22.2 Mc/s polarimeter at the Florida station

were + 3 hours from the direction of pointing. From February 25, 1961,

through September 19, 1961, it was pointed due south. During Septem-

ber 20 30, 1961, it was pointed 400.west of south.

The listening periods of the systems utilizing steerable arrays

were restricted to altitudes of the planet greater than 5. Conversion

of right ascension and declination to altitude and azimuth was facil-

itated by use of a nomogram. The antenna pattern limits and the alti-

tude limits were determined experimentally by checking to see how often

radiation was received outside of these regions.

The Australian data collected during June 17 August 30, 1961,

were handled specially. The convention explained above regarding the

date of an observing period would not work; so, for this data a 24-hour








period beginning at Oh U.T. was used. Listening period limits due to

antenna patterns were not applied.

The relative intensity of a Jupiter radio storm is arrived at

by taking the average deflection above the galactic level of the three

highest peaks during the activity period and dividing by the galactic

level (17). If there is a break in the activity of more than 8.3 min-

utes, corresponding to a 5 rotation of Jupiter, then we consider the

separate parts as two Jovian storms. The comments concerning listen-

ing conditions and description of the radiation provide a basis for

deciding upon the authenticity of the Jovian signals.


Jupiter Program for the IBM 709

A yearly analysis program for the IBM 650 computer was developed

in 1960 (18). Since that time the University of Florida has replaced

the 650 with an IBM 709 computer. The original yearly analysis pro-

gram has been rewritten with many additions, deletions, and corrections.

The input to the 709 is put on punched cards and consists of

the previously mentioned raw data:

a. station

b. date (month, day, year)

c. frequency

d,e. listening period beginning and end in U.T.

f,g. activity period beginning and end in U.T.

h. intensity of the Jovian radiation

plus the following information:









1. normalization constant: This compensates for the change

in galactic level between the 1957 position of Jupiter

and subsequent positions.

j. daily sunspot number: Obtained from N.B.S. Solar-Geophysical

Data Part B.

k. `XI Oh U.T.: This is the longitude in System II (19) of

the central meridian of the illuminated disk of Jupiter

at Oh U.T. on the date in question. See the American

Ephemeris and Nautical Almanac.

1. Julian day number: Found in the American Ephemeris and

Nautical Almanac.


The last three of the above items are daily parameters and must

be selected according to date of the watch period as determined by the

previously mentioned procedure. In putting the data on punched cards,

the listening periods must be split in many cases so that there is only

one activity period in each listening period. Thus, there will be a

card for each activity period.

Seven major computations are performed by the 709 in the Jupiter

program. We shall consider each separately.

Computation of the beginning and end of the listening and activ-

ity periods in System III longitude. The rotation period of the radio

sources is 9h 55m 29.35 (2). The designation "System III" has been

given to the longitude system which rotates with this period and was

coincident with System II at Oh U.T. on January 1, 1957. The System III

longitude of the central meridian at Oh U.T. on Julian date J is found








by means of the equation


XOh X0 + 0.2747 (J 2435839.5). [1]
III II


hI is input item k; J is input item 1; the constant 0.2747 is the

drift in degrees/day between systems II and III; and 2435839.5 is the

Julian day number of January 1, 1957, the date when the two systems

coincided. The System III longitude at H hours M minutes before or

after Oh U.T. is obtained from the relationship


X = + 36.27 (H) + 0.6045 (M). [2]
III III

The minus signs are used if we want the XIII corresponding to a Uni-

versal Time H hours and M minutes before Oh U.T. The plus signs are

used for a U.T. after Oh U.T. The computer checks each U.T. and makes

the correct choice of signs. This method of performing the computa-

tion is necessary because of the convention adopted regarding the date

of an observing period. The constants 36.27 and 0.6045 are degrees of

rotation per hour and degrees of rotation per minute, respectively,

for System III.

Computation of the activity index for each activity period. The

activity index is calculated as follows (20):

activity index = (intensity)(duration of the activity period

in minutes).(normalization constant). [3]

Remember that the intensity of a storm is the average height above the

galactic level of the three highest peaks, taken as a ratio to the

galactic noise level. This is input h; the normalization constant is

input i; and the duration of the activity period is obtained from









inputs f and g. The activity index, as its name implies, is a measure

of how active Jupiter is and contains the factors intensity and duration.

Computation of the daily activity index rate. The computer

uses the equation

Z (activity index)a
daily activity index rate a a [4]
Z (duration of listening
a period in minutes),


to compute the daily activity index rate. "a" runs over all the listen-

ing periods on a day. Remember that in input form the listening periods

were split so that there was never more than one activity period in each

listening period. Notice the division by the duration of the listening

period. This is done to obtain a measure of the rate of activity.

Computation of the monthly activity index rate. This is found

using the above equation, where a runs over all the listening periods

in a month.

Computation of the apparition activity index rate. Again the

same equation is used and a runs over all the listening periods in the

apparition.

Computation of the monthly average sunspot number. Using the

equation



monthly average sunspot number= daily sunspot- [5]
number of days in the
month


the computer determines the average sunspot number for each month.








Generating the probability histogram table. For each 50

increment of System III longitude, the computer counts the number of

times that Jovian radiation was received and the number of times the

station was listening effectively, while any part of that 5 increment

of longitude was on the central meridian of the illuminated disk. The

probability of getting radiation from Jupiter when a particular 50 incre-

ment is on the central meridian is the quotient obtained by dividing the

former by the latter. Seventy-two such values, one for each 5 of lon-

gitude, comprise the histogram table. The count can be taken over an

entire apparition, several apparitions, or just part of an apparition,

depending on what cards are fed to the 709.

If listening or activity occurs in any small part of a 5 incre-

ment of longitude, the program considers that the complete 50 was

listened to or had activity in it. This results in a slight smearing

of the data. When a listening period is split so that there will be

no more than one activity period in each listening period, then the

5 increment of System III longitude in which the split is made is

counted as being listened to twice. Because of the random positions

of the splits, it is felt that this state of affairs does not appre-

ciably alter the histogram tables.

The printout sheets from the 709 Jupiter program contain all

of the input data except the Julian day number, and in addition the

following:

listening period beginning and end in System III longitude

activity period beginning and end in System III longitude

activity index for each activity period









daily activity index rate

monthly activity index rate

apparition activity index rate

monthly average sunspot number

probability histogram table


Merge Program for the IBM 709

The purpose of the merge program is to combine the data taken

at the observatories in Florida and Chile in order to get the best

station-wide activity period and listening period. By merging the data

taken on a particular day at the two stations, we obtain a more real-

istic picture of how Jupiter was behaving. With two-station coverage,

conditions at a single station, such as equipment failures, interfer-

ence, or bad atmospherics do not result in no data for that period.

As an example of how the merge program works, suppose that on June 10

the effective listening at the Florida station was from 0600 to 1030 U.T.

and in Chile from 0415 to 0545 and from 0900 to 1200 U.T. After merging,

the effective listening period would be 0415 to 0545 and 0600 to 1200 U.T.

The input to the merge program consists of the same cards used

in the Jupiter program except that the Chile and Florida cards must be

collated before putting into the 709 so that the dates are in order.

The output of the merge program consists of the date, frequency, merged

listening period beginning and end in U.T., merged activity period

beginning and end in U.T., XII for 0h U.T., and the Julian day number.

These data are punched into cards by the 709 so that they can be fed

into the Jupiter program. The printout of the Jupiter program will








then contain the output of the merge program, minus the Julian day

number, and in addition the merged listening period beginning and end

in System III longitude, the merged activity period beginning and end

in System III longitude, and a merged probability histogram table of

Florida and Chile data combined.



Gross Statistics 1961


The 1961 apparition began February 1, 1961, and ended Feb-

ruary 1, 1962. This is the first time that observations of the deca-

meter radiation from Jupiter have been carried through a complete cal-

endar year. Table 2 summarizes the gross statistics of the 1961

apparition.

The observing period, the number of nights of effective lis-

tening, and the number of nights that Jupiter radiation was received

were arrived at by considering that monitoring at one station consti-

tuted coverage. Notice that the percentage of nights that Jupiter

radiation was received tends to decrease with increasing frequency.

There are some exceptions to this trend. The 19.7 Mc/s system in

Australia employed an extremely high-gain antenna, which may account

for the figure of 85 per cent. The 49 per cent value at 15 Mc/s is

somewhat low. This is probably due to the fact that the array used

with this system is non-steerable; thus, Jupiter is in its beam a

relatively short time during each night. The 50 per cent value for

5 Mc/s is doubtful. At this wavelength the listening conditions are

extremely bad due to absorption in the earth's ionosphere and















ri0




CO) to + LO OD o 0
0 0 (


&


Sr.
+.3 .H
-P -
ID CUr3
*H -H -0
(0 .d






I-H Fi.
1 ^ 4 1E1






41 0- H)
K H


Kr r1 c

00












(Dtoi05
i( S^ 1








+) a















[gI
*^ d
i -
Iooa?




! D o

S <




I ^


(D





I~


"4 0


OI r rt o
t-, I) LIE ,- tt
H- -~ to CV Co to HO




CV V



cvI
C12 CV

0) LO C)
rU R
-4 H H. I
H~ H H1 "' 02





HeN) 0 H ~ C- 3 H

4b a (


R S h fa- 'a S ?+ '1 x>

O ) 0 0 0



L 0 Llz 03C. 0 H
H H i-lH to V C H


H
co


E-t
0n







0
CO


o(D
0:) (








interference from stations and static, and the few cases susceptible

to analysis give at best poor statistics.

During the 1961 apparition the periods of two-station (Chile

and Florida) coverage at the different frequencies were: 15 Mc/s -

3 months, 18 Mc/s 9 months, 22.2 Mc/s 7 months, and 27.6 Mc/s -

3 months. During these periods the nightly watches were as nearly

simultaneous as ionospheric conditions and the 115 difference in

longitude would permit. The majority of the Jovian events were re-

corded at both stations. Two noise storms are considered as different

events if they are separated by more than 8.3 minutes of time, or 50

rotation of Jupiter. Table 3 summarizes the circumstances of the

Jovian events which were recorded at one station only when the other

station was listening effectively. "Effective listening" means periods

during which the equipment is operating, Jupiter is in the antenna

pattern concerned, and the station is not blanketed by interference

at that particular frequency.


TABLE 3

Circumstances of Events Which Were Recorded at One
Station Only, When the Other Station Was
Listening Effectively

15 Mc/s 18 Mc/s 22.2 Mc/s 27.6 Mc/s,


;Events recorded in Florida only:
Event very weak and/or brief
Unexplained

Events recorded in Chile only:
Event very weak and/or brief
Unexplained


2 28 10 4
0 1 1 0


16 6 7 1
0 0 0 0









The two unexplained events require comment. On July 31

at the Florida station a strong Jovian storm was detected on 18 Mc/s

lasting from 0620 to 0659 U.T. Chile was listening under good condi-

tions and detected nothing. On August 10 at the Florida station a

Jupiter event lasting 50 minutes (0330 0400 U.T.) was recorded on

22.2 Mc/s, yet there were only two pulses on the Chile pen recording.

It seems reasonable to assume that these two events were not detected

in Chile due to unusual ionospheric conditions. The values in Table 5

reflect the superior listening conditions in Chile at frequencies of

15 Mc/s and below. Any attempt made to explain the distribution of

weak events must take into consideration the characteristics (gain,

directivity, steerability) of the antenna systems at the different

stations.


Source Studies


1961 Probability Histograms

In contrast to the optical features, which show considerable

irregularity in their motions (21), the decameter radio sources on or

near Jupiter maintain a constant period of rotation. This period was

recently determined to be 9h 55m 29355 (2), which agrees very well

with the value 9h 55m 29?57, independently arrived at by Douglas (22).

The existence of localized zones of activity on or around

Jupiter is demonstrated by the peaks on the histograms in Fig-

ures 1 and 2. To construct these histograms, we imagine the planet

divided into 5 zones of longitude. The "probability of occurrence"








for each zone is the fraction of observing time during which radiation

was received while any part of that zone was on the central meridian of

the planet.

Figure 1 is a composite of the histograms constructed from

data taken in Florida during the 1961 apparition of Jupiter. The

letters A, B, and C on the 18.0 Mc/s plot are source designations.

The fact that the main peaks on the histograms are less than 1800 wide

is interpreted as evidence for the existence of directional sources

located at the longitudes of the peaks. Notice that the peaks become

narrower as the frequency increases. Because of the scarcity of radia-

tion at the higher frequencies, the statistics are weak in the 27.6 and

31.0 Mc/s histograms. Notice also that source B always appears to be

broadened. The sources appear to shift position with a change in fre-

quency. As an example of this drift, note how source A is displaced

towards lower longitudes as the frequency increases.

Figure 2 is a composite of the histograms constructed from

Chile data taken in 1961. Again A, B, and C are source designations

and the data used in arriving at the 27.6 Mc/s histogram are sparse.

There were cases in 1961 of reception of Jovian radiation at 5 Mc/s in

Chile; however, the data were not sufficient to justify plotting a

histogram. Again the same features are found by examining the Chile

histograms: the peak widths are less than 1800, the peak width decreases

as the frequency increases; source B appears broadened; and as the fre-

quency increases, the peaks are displaced towards lower longitudes.

The 10.0 Mc/s histogram is difficult to interpret because of the broad-

ening of the sources at the lower frequency. It is reasonable to




























































900 1800 2700
LONGITUDE, SYSTEM III


Figure l.--Histograms of the Florida 1961 data.


3600








.1 4i-- L


0-

.4n


. 1


.1 -_


0-

.3-


.2-1





o.-[-
0o


900 1800
LONGITUDE, SYSTEM III


Figure 2.--Histograms of the Chile 1961 data.


2700


3600









assume that the main source A extends from 2660 to 360, as will be

pointed out later in the text.

Three degrees of certainty are used by observers on watch in

identifying Jovian radiation: "positive," "possible," and "dubious."

In constructing our histograms, only the "positive" and "possible"

Jupiter radiation is used. As a check on the authenticity of the

"possible" identifications, a histogram of the Florida 18 Mc/s "posi-

tive data" has been generated and, in Figure 3, it is compared with

the regular 18 Mc/s Florida histogram. All features of importance on

the "positive data" histogram are plainly evident on the regular

histogram; hence, it is concluded that inclusion of the "possible"

identifications is not diluting the good data.


Merging the Data from Different Stations

During 1960 and 1961 Jupiter radiation was monitored on sev-

eral frequencies which were used at two observing stations. This not

only provides a check on our data, but also makes it possible to deter-

mine whether or not there are any lasting station effects such as

better resolution with the antenna system at one station, shifts in

the peaks on our histograms, or better atmospheric conditions charac-

terized by superior reception at one of the stations. All of the histo-

grams have been "smoothed" by taking a 3-point running average, i.e.,

averaging the probability value for each 50 increment of longitude with

the value preceding it and the value following it. This "smoothing,"

as the name implies, removes much of the bumpiness evident in the

histograms of Figures 1 and 2. The following figures contain






































0

0i .4- -















LONGITUDE, SYSTEM III

Figure 3.--1961 histograms of the 18 Nc/s Florida data.
The "possible" identifications of Jovian noise were not in-
cluded in the data used to construct the upper histogram.









single-station histograms above and a merged histogram, which is the

result of combining the data of the two stations, below. See the

explanation earlier in this chapter of the merge program for the 709.

Figure 4 contains 10 Mc/s histograms of the 1961 data. The

Australian histogram comes from data taken by Professor T. D. Carr

during June August, 1961, at an observing station near Sidney.

The 10 Mc/s antenna systems at the Chile and Australian stations are

identical, hence the slight splitting of the peak representing source B

is assumed due to the statistical fluctuations in the data. Consider-

ing the broadening of the peaks at the lower frequencies, it is felt

that the upper two histograms are very well correlated. The higher

probability values associated with the Australian histogram reflect the

better listening conditions at that station due to less atmospheric

noise and a generally less dense night-time ionosphere (25). The

merged histogram in Figure 4 results from a combination of the Chile

and Australia data, and it is felt that this plot gives the best pic-

ture yet obtained of the source structure at 10 Mc/s.

Figure 5 contains 15 Mc/s histograms of the 1961 data. The

fact that source B shows more splitting on the Florida histogram must

be due to scatter in the data. Again the higher probability values

reflect better listening conditions in Chile. It is noteworthy that

Jupiter was more nearly overhead in Chile during the 1961 apparition.

More data were available from which to construct the Chile histogram

because the antenna system is electrically steerable, providing longer

observing periods. The merged histogram gives the best representation

of the 1961 15 Mc/s data.






.4-- -
I Chile


H


K-


Merge


/i


90


1800
IDONGITUDE, SYSTEM III


Figure 4.--10 Mc/s histograms 1961.


K


.5 3


2700


3600





28




.5 .... -- ........ .. .- I

Florida
.4-


.5-


.2


.1


O-

Chile
4


A C
o K




.21

















.1
0.- _




00 900 1800 2700 3600
LONGITUDE, SYSTEM III


Figure 5.--15 Mc/s histograms 1961.








Figure 6 contains the 18 Mc/s histograms constructed from the

1961 data. The correlation between Florida and Chile is excellent.

Figure 7 and Figure 8 represent the 22.2 and 27.6 Mc/s data respec-

tively. In all cases the histogram correlation between stations is

good, and it does not appear that any station effects are being dis-

guised by merging the data. (Notice that the small bump on source C

in Figure 7 disappears in the merge. This is due to the round-off

procedure of the 709 program, which calculates the probability values

to the nearest 0.01.)

Figure 9 compares the Chile 20.0 Mc/s and the Australian 19.7

Mc/s histograms, both constructed from the 1961 data. There is no

merged histogram in this case because the frequencies are not the same.

Any effects such as the longitude shift of the peaks with frequency

and the narrowing of the peaks with increasing frequency would have

been hidden by merging the data. The frequencies are sufficiently

close that a comparison seems worthwhile, and the correlation is quite

good. It should be remarked that the 19.7 Mc/s system in Australia

possesses very high gain.

The first year that the Jovian decameter emission was monitored

simultaneously in Chile and Florida was 1960; thus, for the sake of

completeness, the 1960 station merged histograms are included here.

They have been constructed by the 709 computer using the corrected

System III rotation period. Figures 10 and 11 contain the 18 Mc/s

and 22.2 Mc/s histograms of 1960. The Florida and Chile histograms of

both figures are well correlated. There is one feature, however, which













Florida


Chile


Merge


900 1800 2700
LONGITUDE, SYSTEM III

Figure 6.--18 Mc/s histograms 1961.


3600































I Chile
S.4-
- -

u

(x,
o



B A C






Merge
.4


.3


.2






0-
00 90 1800 270 3600
LONGITUDE, SYSTEM III


Figure 7.--22.2 Mc/s histograms 1961.














Florida


Chile


Merge


isoo
WLNGITUDE, SYSTEM III


Figure 8.--27.6 Mc/s histograms 1961.


.3J

.22


0-



.4-


.3-i


.2-









.1-






0 -


.1-
0i-




33




.5 1 I .. I .

Chile 20.0 Mc/s
.4-





.2




CD


Australia 19.7 Mc/s



S.3- \ -



A C




04---i-~-i~'r---------------------------i---------
00 900 1800 2700 3600
3IIJGITUDL, SY.TM N III


Figure 9.--20.0 and 19.7 Mc/s histograms 1961.











I I


.5-


.4-


.3-


.2


.1-


0-



.4-






.2


.1-






.4












0
.2






o-4
C


B A C

/


Merge


90 180 ?7&

LONGITUDE, SYSTEM III

Figure 10.--18 Mc/s histograms 1960.


I1
360


L-.


I


Florida

















Chile


)










- -- -_ i-


.2








Chile

I












Merge
.34





.2 --
c
S i









.3-j





0 -= = --- -- ^ -- ^ -- --

00 90 180 2700 5600
JDNGITUDE, SYSTEM III


Figure 11.--22.2 Mc/s histograms 1960.









deserves mentioning. On both the 18 Mc/s and 22.2 Mc/s Florida

histograms, source B appears split, having three maxima on the 18 Mc/s

curve and two on the 22.2 Mc/s curve. This splitting does not show

up on either of the Chile histograms. Considering antenna patterns

and observing conditions at the different stations, I cannot think of

an explanation for this feature; hence, it is concluded that the effect

is not a real character of the Jovian radiation, but rather a conse-

quence of scatter in the data. Again the merged histograms provide the

best picture of the source distribution.


Search for Long-Term Changes

Having concluded that there are no significant station effects,

a comparison is made in this section of single frequencies over a span

of years using the smoothed histograms and the smoothed merged histo-

grams to determine whether or not there is evidence of any long-term

change in the activity of the different sources.

Figure 12 shows the 10 Mc/s histograms for 1960 and 1261.

The 1960 apparition in Chile was our first attempt to obtain Jovian

signals at a frequency as low as 10 Mc/s, and not much data was ob-

tained. With the sunspot activity on the decline and the earth's iono-

sphere becoming more transparent to the lower frequencies, a large

amount of data was collected on 10 Mc/s during 1961. Notice that

Jupiter seems much more active at 10 Mc/s in 1961 than in 1960.

Comparison of the two histograms in Figure 12 reveals the same

general source structure. The 1961 merged histogram shows sources B





37













1961
Merge: Chile and Australia
.4


.3-


.2
C B A








.3


.2-


.1


0
0 9 0 180 270 3600


Figure 12.--10 Mc/s histograms.









and C more developed. It is interesting to note that source A is

bifurcated on the 1960 plot and also slightly on the 1961 plot.

Figure 13 contains 18 Mc/s histograms covering a six-year span.

The data used to generate the 1956 histogram are those of R. M. Gallet

(24), taken at the National Bureau of Standards, Boulder, Colorado, dur-

ing the period January March 1956. The shape of this histogram con-

forms well with that of a histogram constructed from data collected by

K. L. Franklin and B. F. Burke (25) in 1956 on 18.5 Mc/s. This latter

histogram is not included here because of the difference in frequency.

The remaining five histograms in Figure 13 are those of the Florida

group. Most noticeable is the decrease in activity during 1958 and

1959 accompanying sunspot maximum. Source A is the most prominent

peak on each year's histogram. B appears to be the broadest peak,

except in 1958 and 1959 when there was little activity. Source C

started out in 1956 more active than B. In 1960, C was surpassed by B,

but during the 1961 apparition C regained the lead. The breadth of

source B may be an indication that there are two closely spaced sources,

and in some cases the separation on the histograms may be evidence of

resolution. Notice this on the curves of 1956, 1960, and 1961. This

feature will be commented upon later in the text.

The center of source A is marked on each of the histograms at

one-half maximum height. No consistent drift is apparent; however, a

certain amount of randomness shows. On the 1961 histogram a shift to

the right is evident. A plausible explanation of this effect will be

discussed later on. Again, be reminded that the data taken in 1958

and 1959 were thinly scattered.




39



.4 L _. .._
1961 M
.1







S 960

S.i-4 1959 F \


C)1958
O- / _\





-a L
0 /\\





.1



*1 i / \ -


00 90 180 2700 360
LONGITUDE, SYSTEM III
Figure 13.--18 Mc/s histograms. Legend: F Florida,
G Gallet, M Merge.









Figure 14 contains three 20 Mc/s histograms. Gallet's data

were taken from reference 24. There being no continuity from 1956 to

1960, it is only possible to infer from the data at other frequencies

that the 20 Nc/s activity declined during the period of sunspot maximum.

A is the most prominent source and B again appears broadened. Notice

that source C was less active than B at 20 Mc/s in 1956. The reverse

held true at 18 Mc/s. The 1961 histogram shows C more active than B.

Again the peaks in the 1961 histogram are shifted to higher longitudes.

Figure 15 contains 22.2 Mc/s histograms covering 4 consecutive

years. The activity is at a low in 1959 as signified by the smaller

values of probability. Once more, A is the most prominent; B appears

broad, suggesting a combination of sources on the 1958 and 1960 curves;

and source A is shifted to the right on the 1961 histogram. At 22.2

Mc/s, source C is very insignificant.

Figure 16 contains 27.6 Mc/s histograms for 1959, 1960, and

1961. The probability of receiving decameter wavelength radiation from

Jupiter falls off with increasing frequency. This is well illustrated

by the small ordinate values in Figure 16. The 1959 data were very

meager. Notice that source C disappeared in 1960. There is some evi-

dence in 1960 and 1961 of splitting of source B. Any conclusions based

upon the 27.6 Mc/s curves must be regarded with caution because of the

small amount of activity they were derived from.

In order to see if there were any short-term changes in the

activity of the different sources, several sets of monthly histograms

were generated on the computer. At none of the frequencies investigated




41




.5 I
1961
Chile
.4


.3


.2






0--
m 1960
Chile
.4-i

3.3
0
o
2 B A C
S.2-







1956
Gallet




3H









0 900 1800 2700 3600
LIrJGITUDE, SYSTEM III


Figure 14.--20 Mc/s histograms.






42




i I- \ --- II

.4- 1961
Merge: Florida and Chile

.3-


.2






0
S.5- 1960
Merge: Florida and Chile

.2-
3 B A C



H
S.1-




0 1959
( .2- Florida


.1-


0- y

1958
.3- Florida


.2


.1




0 90 130 2700 360
LOTNGITUDE, SYSTEM III


Figure 15.--22.2 Mc/s histograms.











I I 4-


.5 -



.4-



.3



.2


.1 -



0



.4 -



.3



.2


.1 -



0



.4


.3



.2



.1 -


0 -


1961
Merge: Florida and Chile










B


1960
Florida


1959
Florida


900 180 2700
LOrIIGITU1JDE, SYSTEM III

Figure 16.--27.6 Mc/s histograms.


6 I
360


A C


i ,I









did the activity show any pattern of shifts from one source to another

on these monthly histograms.

In conclusion, the following can be said about the search for

changes in the source structure from year to year. Most of the prom-

inent features on the histograms remain unchanged. Source A is the

most prominent. Source B is broad and appears split in many cases.

But, two effects have been discovered.

1. The activity of source C varies from year to year in

comparison with B.

2. Source A is shifted towards higher longitudes in 1961.


Search for Station Effects

It was concluded earlier that by merging the data obtained on

a single frequency but at two different stations, no effects that char-

acterized the decameter sources were being covered up. In order to

subject this conclusion to further test, a comparison is made of his-

tograms obtained by combining the data of different apparitions but

not merging the data from separate stations. In so doing, the two

effects discovered in the previous section will be masked, but we will

remember that they do exist. The prominent features of the source

structure will not be altered by this treatment of the data. One

question posed earlier should be answered by this examination. Does

source B show more splitting on the Florida histograms?

Figure 17 contains two 10 Mc/s histograms, one a combination

of data taken during two apparitions in Chile, and the other constructed

from data obtained during a 2 1/2-month period in Australia. The


















.4- I -
Combined 1960-1961
Chile
.3


.2 \ C AB \






0 -
S 1961
SAustralia
o .6-


.5-


S.4


.3


.2


.1


0 I

0o00 180 2700 3600
LONGITU7E-, SYSTEM III

Figure 17.--10 Mc/s histograms, data of different
years combined.









general source structure is the same on the two curves. The slight

splitting of source B on the Australian histogram has been noticed on

the histograms in preceding sections.

Figures 18, 19, and 20 compare Florida histograms representing

the data of several apparitions with Chile histograms of data taken in

1960 and 1961. The correlation in each figure is extremely good. The

probability values are about equal at the different stations. The

amount of radiation received from the separate sources is the same in

Chile and Florida. There appears to be no displacement of peaks

between stations. And in answer to the question posed, source B has

the same structure on the Florida and Chile histograms at each of the

three frequencies at which a station comparison is made. Hence we con-

clude once again that there are no station effects in the data which

alter the histogram structure.


Combining the Data of Different Apparitions and Different Stations

In order to obtain the most reliable picture possible of the

source structure at each frequency, all the data that have been col-

lected by the Florida group since 1957, including that taken in Chile

and Australia, have been combined to produce the histograms in Figures

21 through 26. The only effect we are hiding by lumping together the

data of different apparitions is the shift of source A to slightly higher

longitudes on the 1961 curves at 18 Mc/s, 20 Mc/s, and 22.2 Mc/s. The

result of combining the 1961 data with that of the other years will be

a broadening of source A at these three frequencies and a slight shift

of the center of A toward higher longitudes.





























.2- B A C


b .1
0

0 o-


H Combined 1960-1961
.4- Chile


.3-



0
.2








0 900 1800 2700 3600
LS'IJGITUDF, SYSTEM III

Figure 18.--18 Mc/s histograms, data of different
years combined.





48











.5
Combined 1958-1961
Florida
.4 -


.3-




B A C



o-
0 0 .i __-r _.- "-------------Z I -- -

Combirne.' 1960-1961
H Chile
.4 -


P .3


.2







00 900 1800 2700 3Ci
LONGITUDE, SY3TEll III

Figure 19.--22.2 Mc/s histograms, data of differ-
ent years combined.





















Combined 1958-1961
Florida













B A C



Combined 1960-1961
Chile


-4\


1800
LOIrCIT UDE, ,-Y'rTE III


2700 3600


Figure 20.--27.6 Mc/s histograms, data of different
years combined.


.4



.3



.2



.1


.4



.3



.2



.1


900


I I


I I


I I









Conclusions Regarding Number, Location, Separation,
and Width of the Sources

By inspection of Figures 21 through 26, three main sources,

A, B, and C, can be clearly identified, and there is a suggestion of

splitting of source B on the 18 Mc/s, 20 Mc/s, and 22.2 Mc/s histograms.

It is possible that this broad peak represents two closely spaced

sources which are only partially resolved. The existence of four main

sources is easier to explain than three. If the sources are connected

to the solid core of the planet through its magnetic field, it is pos-

sible that two magnetic dipoles are involved.

The source structure does not vary much from frequency to fre-

quency. The most change occurs in source C, which shrinks away to

nothing at 27.6 Mc/s, yet at 18 Mc/s it is more active than source B,

and at 15 Mc/s it is on a par with source A.

Table 4 gives the location of sources A, B, and C in System III

longitude and the widths of the histogram peaks at one-half maximum

height. These measurements were taken on Figures 21 through 26.

TABLE 4

Location and Width of the Decameter Sources
from the Merged Histograms
Location of the source
Frequency center (System III) Width (degrees)
(Mc/s) A B C A B C
10 330 222 80 135 90. 65
15 246 142 312 67 70 46
18 235 120 510 66 100 50
20 228 135 299 52 70 45

22.2 225 127 505 45 90 55
27.6 225 125 302 45 25 20




























ri
CD
0)


0
CD
0)
H
00 u




r3
H
% O




0






% U


N 9 0 lI~O0 0
0




I 1 9
4-)

ri







0)

























SONS{8fODO do lniavaoIHd



























0
0
c~







--4




W
















o
0
0C


3't3YDnk1jCi A o rIIIOaraod




























0










H
H
H


0
oC
cv















0 m
















0
C)I


L'flZ9lIc'id A0 u LIrIIvSoadd




























o H
00 1


CD
l)
r-1

+P



o-4 H-




0 0





-S m






0



0 c
0 0
E t< l
1 Fx K


ONaMiniooo .dO .L TIIvsOcud























,o


co
0)

S H
01


4-




HH

10





H





aJ

H .
0








o

0 C,





I
ru

i)




.re


j 0 0 nIgC'vd



























o )

0 H
0
0>

C- V
JD

r1





rl





o o
z .4-I



o
Q0





d












.0.








0 0
0 Q





0I
*H




I I I
(Q M .-!
*~ *
3:ii'3yao~o ~o .Lnidaoy









The 1961 smoothed, merged histograms were analyzed in the same

manner as Figures 21 through 26, and Table 5 contains the results.


Frequency
(Mc/s)

10

15

18

20

22.2

27.6


TABLE 5

1961 Data: Location and Width
of the Decameter Sources

Location of the source
center (System III) Width (degrees)
A B C A B C

331 219 82 130 85 68

246 145 314 69 70 45

239 121 312 62 98 41

232 139 306 49 68 31

227 128 304 42 88 56

224 120 303 56 26 46


Since our data covering the years 1957 to 1960 have not before

been analyzed using the corrected System III rotation period, Table 6

contains the location and width of principal source A for these years.

The other sources were not sufficiently defined during the years of

maximum sunspot activity.

The shift of source A towards higher longitudes which was

evident on the 18, 20, and 22.2 Mc/s histograms of 1961 is probably

not a real effect amenable to explanation in terms of a physical event

happening at Jupiter. It is the conclusion of the writer that this

shift is due to statistical fluctuation in the data. Assuming a random

distribution of the System III longitudes of source A at 18 Mc/s for

the four apparitions preceding 1961, the longitude of source A at









18 Mc/s on the 1961 histogram is well within 2r of the mean, -here c

is the standard deviation of the sample. The deviation in the longitude

of source A on the 1961 18 ic/s histogram is not statistically signif-

icant.


Frequency
(Mc/s)


10

16.7"

18

20

22.2

27.6


196
Location V


316

231

232

219

219

222


TABLE 6

Location and Width of Source A:
1957 1960 Data

50 1959 19
Jidth Location Widthi Location '


124

55

64

58

32

20


*This histogram


58
Width


214 55 234 54



222 3355 219 28

209 13

not included in the text.


1957
Location Width


We noticed earlier

A towards lower longitudes


in the text that there was a shift of source

as we inspected histograms of higher and


higher frequency. Evidence of this trend is found by inspection of

Figures 1 and 2. It is assumed that the same drift is occurring for

B and C, but, since these sources are not as well defined, the shift is

not so obvious. The values in Table 4 indicate the longitude displace-

ment with frequency, and Figure 27 is a plot of these values showing the

location of the sources as a function of frequency. The combined Flor-

ida and Chile data of 1957 1961 were used to generate these curves.


.~--~-1-------~~-





I I I I I i I I I


-P 0C 0
r-4 r-i r-l


(S/own) 'M2l'i


0
om


I 1 I I I I




























<1e


o0 C CD 1 C2 0 O CD
to CV c Cv C C0 W -H r-1


1


0
0
I-M


H
E-4

ClO





0

0o c











o
0


S I | i | | B I .-- I


!


I


1f- 0










The longitude shift was first noticed by Gardner and Shain (26),

and was pointed out again in 1960 data published by the Florida group

(2). It is the conclusion of the writer that this effect is real.

What could cause such a shift with frequency? One possible explanation

involving the mechanism of the decameter radiation requires radial

movement of the source away from the surface of the planet. If, for

example, the source is a plasma cloud which has its genesis near the

surface of Jupiter and then drifts to higher altitudes, it will neces-

sarily fall behind the rotating surface in order to conserve angular

momentum. If, as it rises, it emits radiation of lower and lower fre-

quency, characteristic of cyclotron emission in a magnetic field which

is becoming weaker with increasing distance from the surface, then the

lower frequencies will come from regions whose sub-Jovian longitudes

are greater. The high-frequency decameter radiation, occurring nearer

the planet, will be associated with smaller values of longitude. This

hypothesis has additional advantages in connection with the directional

characteristics of the sources, as we shall see later. One of the dif-

ficult points in the argument is accounting for the production of such

a source near the surface.

If such a model is a description of the mechanism responsible

for the decameter emission, then a calculation of the radial gradient

of Jupiter' s magnetic field should be possible. An outline of such a

calculation, attempted by the writer, follows. If it is assumed that

a plasma cloud is ejected radially from the Jovian surface at lati-

tude 8, and that the only force acting on the cloud after ejection is








the force of gravity, then the equations of motion of the cloud rela-

tive to axes rotating with the planet are


x W vy 2 cos 9 -E t3 cos 8 [



y y t gt2. [



Figure 28 shows the orientation of the coordinate axes.


Figure 28.--Coordinate system attached to the Jovian surface.



x is the westerly displacement of the plasma; y is its altitude above

Jupiter's surface; vy is the velocity of ejection; w is the angular

velocity of Jupiter; t is the time elapsed since ejection; 9 is the

Jovian latitude; and g is the acceleration due to Jupiter's gravity.

If a latitude is chosen, then we are left with four unknowns: x, y,

vyo, and t. Next it is necessary to introduce empirical information

in order to reduce the number of unknown quantities. From Figure 27









we can obtain the change in the longitude of the decameter sources
AX
with a change in frequency: ".


Since x rAX [8]


and r = ro + y, [9]


where ro is the radius of Jupiter, we arrive at


x (ro + y) AX. [10]


If we know y, we can find x. The number of unknowns has been reduced

to three. Spectral data tell us that the frequency drift rates of the

Jovian noise bursts range from .01 to 2 Mc/s per minute (12). If we

assume that the 27.6 radiation is emitted near the surface at approx-

imately t O, then it is possible to find the time at which other

frequencies will be emitted by choosing a drift rate in the observed

range. Now we are left with two unknown quantities y and v Equations
Yo
[6] and [7] can be solved simultaneously for the velocity of ejection

of the plasma and the altitude of emission of the lower frequency. If

it is assumed that the radiation is emitted at the cyclotron frequency,

then the magnetic induction can be found at the altitude of emission

using

f 2.8 B [11]

where f is in Mc/s and B is in gauss. Carrying out the above procedure

to find the altitude of emission of several frequencies, one derives

the change in B with altitude or the radial gradient of B.








Although the method seems plausible, trouble was encountered

due to the large range of observed frequency drift rates. Different

values of drift rate gave positive, negative, and infinite velocities

of ejection. The drift rates which yielded positive ejection velocities

also led to altitudes of emission of the 10 Mc/s radiation below the

altitudes where the 18 Mc/s emission occurred. Assuming emission at

the cyclotron frequency, this leads one to believe that the Jovian

magnetic field strength increases with altitude--an oddity to be sure.

Since the empirical frequency drift rates do not give reason-

able velocities of ejection, it is concluded that the equations of

motion must be in error. It would seem worthwhile to repeat the calcu-

lations including in the equations of motion the effects of electro-

magnetic forces on the plasma. Such calculations are not within the

scope of the present endeavor.

If the sources are connected to the solid disk of Jupiter

through its magnetic field, then the separation of the peaks on the

histograms should give us a clue to the geometry of this field. By

taking the differences between the longitudes locating sources B and C

in Table 4, it is found that, excluding the 10 Mc/s values, these

sources are separated by approximately 1800. The average separation is

176, and the greatest deviation from a 1800 displacement is only 160.

Considering the arbitrariness involved in determining the center of the

sources, these values are quite close together. The next question is:

What feature appears on the histograms at the longitude 1800 from

source A? The answer is--nothing. On all but one of the histograms









in Figures 1 through 26, the feature opposite source A is a null or

minimum value of probability. If the location of the sources is an

indication of the magnetic field geometry, it appears that Jupiter has

a three-pole field. More reasonable would be the assumption of a

dipole field with at least one strong inhomogeneity or perhaps two

magnetic dipoles oriented in some strange fashion to give the indicated

geometry. If source B is indeed a double source, then the latter sug-

gestion does not appear so unreasonable.

The fact that the widths of the peaks on the histograms are

less than 1800 implies that the radiation has a directional character-

istic. If the sources radiated isotropically, then we would receive

the signals for the half-revolution of Jupiter during which the source

was on the hemisphere facing the earth. That this is not the case is

demonstrated by the values in Tables 4, 5, and 6.

The directional property of the sources was noticed by C. A.

Shain (27) in data obtained during August September 1951. K. L.

Franklin and B. F. Burke (25) arrived at the same conclusion by analy-

sis of their 22.2 Mc/s data of 1956, and the same year Gardner and

Shain (26) found the effect at 19.7 Mc/s. In 1958, Carr et. al. (28),

pointed out that the Florida data supported the previous findings,

and it was suggested that the existence of a Jovian ionosphere was

plausible.

Tables 4, 5, and 6 show that the width of source A decreases

with increasing frequency. This is better illustrated in Figure 29,

which contains separate curves for each of the years 1958, 1959, 1960,








and 1961, along with a curve derived from measurements of the combina-

tion histograms. Again the widths of the peaks on the histograms were

measured at one-half the maximum height. All of the curves in Figure 29

indicate that peak width decreases as the frequency goes up. J. W.

Warwick has written that beam width is independent of frequency (29).

Our data contradict this.

The decrease in the apparent width of the principal decameter

source with increasing frequency cannot be explained in terms of a

fixed point source lying under an ionosphere. Let us examine the rea-

soning behind this statement. If a source of electromagnetic radiation

lies below an ionized layer in which the maximum electron density is N,

then the lowest frequency capable of penetrating this layer at normal

incidence is given by the expression


f2 Ne2 [12
o nm

fo is the critical or plasma frequency for a layer with N electrons/cm.3,

e is the electronic charge in electrostatic units, and m is the mass of

the electron in grams. All radiation of frequency less than fo which

is incident on the layer will be reflected. Now suppose that the radi-

ation is not normally incident, but makes an angle i with the normal to

the layer. The lowest frequency f that will penetrate the layer at

this angle of incidence is related to fo by


f fo sec i. [13]


All frequencies less than fo sec i are now reflected. If the decameter








































600






400



300 -KA
t. \

200-


100-

00o I I II I

10 15 18 20 22.2 27.6
FRQnUEHCY (Mc/s)

Figure 29.--Variation of source width with frequency.









radiation from Jupiter came from a point source located under an ion-

ized layer whose critical frequency fo was less than 10 Mc/s, then

10 Mc/s radiation would have a cone of escape with half angle i smaller

than that for the cone of escape of 20 Mc/s radiation. As these cones

of radiation swept past the earth, the source of the 20 Mc/s radiation

would appear broader than the 10 Mc/s. This is not what we observe.

The peaks on the histograms become narrower at high frequencies. In

the previous discussion, we have neglected the probable presence of a

magnetic field on Jupiter. To be exact, we should consider the propa-

gation of electromagnetic waves in a magneto-ionic medium. Such a

medium is doubly refracting and two modes of propagation must be con-

sidered. A concise explanation is found in reference (50).

How, then, can we explain the two observations: histogram

peak widths less than 1800 implying directional sources, and the de-

crease in peak width with increasing frequency? There are two possible

explanations, both doing away with the idea of a fixed point source

from which all the decameter radiation spews forth.

If the high-frequency radiation came from well beneath the

layer of maximum ionization in a Jovian ionosphere and the lower fre-

quencies originated further out from the surface of the planet, then

the high frequencies would be limited to narrower cones of escape,

explaining the peak width variation on the histograms. The idea of

radial movement of the source was mentioned previously in connection

with the apparent shift of the sources to lower longitudes on the

higher frequency curves.









Another possible explanation is that the higher frequencies

come from greater latitudes on the planet. The emission cones for

the high and low frequencies could intersect so that as the sources

rotate around the planet, an observer on earth would be exposed to a

wider segment of the low frequency beam. Figure 30a shows the geo-

metry of radio reception at the earth. The half angle of the emission

cones for different frequencies is determined by the electron density

in the Jovian ionosphere. Figure 30b shows the regions of radiation

that would pass over the earth as Jupiter rotates. Segment BC corre-

sponds to 20 Mc/s reception and AD corresponds to 10 Mc/s reception.

Notice that the 10 Mc/s cone has been displaced in the direction of

greater System III longitude to conform with the data.

Both of the above explanations retain the idea of a Jovian ion-

osphere and separate the points of origin of the different decameter

wavelengths. Further evidence in favor of a Jovian ionosphere will be

brought up in Chapter 3 in connection with the long-term inverse corre-

lation with the sunspot cycle.


Intensity Histograms

So far, the conclusions regarding source structure have been

based on the probability histograms of Figures 1 through 26. A nat-

ural question to ask is, "How does the intensity of the decameter

radiation vary with Jovian longitude?" In order to answer this ques-

tion, intensity histograms have been constructed and are presented in

Figures 31 and 32. A few words of explanation about the method used

in developing these histograms seem appropriate. In reducing the slow




69



20

/ 10










-_ ~__ Earth
20

Jupiter



10

a









20 Mc/s




10 Mc/s /







D C B A Earth's motion with
respect to radiation
cones



b

Figure 50.--Geometry of radio reception of Jovian outbursts
assuming a Jovian ionosphere and latitude separation of the frequen-
cies emitted from a single source.









speed pen recordings of Jupiter noise, the average height above the

galactic level of the three highest pulses in a storm is determined,

and the ratio of this average height to the level of galactic noise

at the position of Jupiter is found. This ratio is a measure of the

relative intensity of the Jovian noise storm. To make a histogram

table, it is also necessary to know what longitudes (System III) were

on Jupiter's central meridian while the storm was in progress. Know-

ing these things, we can construct a histogram table by recording the

relative intensity value in each column representing longitudes that

the storm covered. Having done this for a particular frequency and

apparition, we add up the values in each column and divide by the

number of entries to obtain an average intensity for each 5 increment

of Jovian longitude. These values are then smoothed by three-point

averaging and plotted. Of course, a better method would have been to

measure the intensity of the storm every 8.5 minutes, corresponding to

5 rotation of the planet; however, the amount of time required would

have been considerable. A certain amount of smearing results from

using the average-three-high-peak ratio as a measure of the intensity

of the whole storm. Very often the three high peaks are clustered close

together while the complete storm might last as long as two hours,

corresponding to more than 700 of rotation.

Figure 51 contains 18 Mc/s intensity histograms for the five

years, 1957 through 1961. The only prominent feature common to all the

curves is the general dip in the average intensity centered at about 400.

This lines up quite well with the null in the probability histograms.












1961
Chile








1960
Chile


1959
Florid a








1958
Florida








1957
Florida /


4-


3600


Figure 31.--18 Mc/s intensity histograms.


2-


1-


0-


00 90 180 2700
LONGITUDE, SYSTEM III


0









The following points are of interest. The peak at 120 on the 1961

curve corresponds to the location of source B. On the 1960 curve

there is a small peak at 400 in the center of the general null extend-

ing from 200 to 600. The 1959 curve shows two dips, one at 400 and

another at 532. There are also two peaks, a broad one centered at

1800 and a narrow one at 3550. The probability histogram of 1959

shows source A at 2140. The 1958 plot is quite jagged and the null

at 400 is not conspicuous to say the least. The broad peak extend-

ing from 200 to 2600 corresponds to the position of source A. The dip

on the 1957 plot occurs at 250, and none of the other features line up

with sources on the probability histogram for that year. There does

not appear to be any regular pattern in the intensity behavior from

year to year.

Figure 52 contains four more 1961 intensity histograms. The

upper two curves are for 10 Mc/s in Chile and Australia. It appears

from the flatness of these histograms that there is almost no intensity

variation with longitude at 10 Mc/s. The 19.7 Mc/s curve of the

Australia data shows the prominent dip at about 400. Notice that the

ordinate scale is compressed. The 27.6 Mc/s intensity histogram is

the only one that looks like its corresponding probability histogram.

The three peaks lie at 2180, 1160, and 295, and they match the longi-

tudes of sources A, B, and C respectively. This is probably due to

the fact that there was such a small amount of 27.6 Mc/s radiation

received from Jupiter. What was received came only when the sources

were near the central meridian.


























































90o 180 270
LONGITUDE, SYSTEM III
Figure 32.--1961 intensity histograms.









Except for the general null at about 40 corresponding to the

longitude from which we very seldom get radiation, there is little

correlation between the intensity histograms and the probability his-

tograms. The flat appearance at 10 Mc/s and the random nature of the

other plots, where there is enough data to give a fair representation,

lead to the conclusion that the most intense radiation does not neces-

sarily come from those longitudes which are the most frequent sources

of radiation. In other words, there does not appear to be any well--

defined variation of intensity with longitude. Certainly, there is

room for more investigation along these lines.

As mentioned earlier, a truer picture of the intensity var-

iation would have been obtained by breaking up the periods of activity

into shorter segments, say 10-minute intervals, and taking measure-

ments of the maximum pulse height in each. This was done for the 1961

Chile data at 18 Mc/s. The measurements and theory involved are

described below.

Let D deflection due to Jupiter plus the galaxy

G deflection due to the galaxy alone

Pj = power from Jupiter alone

PG = power from the galaxy alone.

D and 0 are measured on the pen recordings. The power ratio is given

by the expression


PJ D2- 2 -1 14]

TheG flux density is the power per unit area per cycle per second

The flux density is the power per unit area per cycle per second








hence, the flux ratio is

Fj D2 [15]
F- -- 1. [is]
FG G2

Solving for the Jovian flux density we get


Fj F D2 1 in watts/m2/cps. [16]
J G 2


The flux density from the galaxy is determined in the following

manner.

Let e m charge on the electron (1.6 x 10-19 coulombs)

A = effective area of the antenna

T = transmission coefficient of the transmission

line (percentage of power getting through)

R = output resistance of the calibrator


IG = product of the average calibration current in

amperes giving the same deflection as the

galactic signal and the scale factor 6.5.



Then F e IGA [17]
G AT


The Jovian flux density is given by


F = I R D2 [18]
J AT G2









The constants for the 18 Mc/s Chile data of 1961 are the following:

iG (.014)(6.5) 0.091 amp

R 75 ohms

T 0.7

A 0.15 gX2 = (0.13)(15)(16.7m)2 543m2

where g is the average power gain of the interferometer antenna, and

X is the wavelength in meters. The average value of the calibration

current IG used was that determined for the three-month period June -

August, 1961. By multiplying these factors together, we get a value

of 2.88 x 10-21 watts/m2/cps for the galactic flux density at 18 Mc/s.

Equation [18] now takes the special form


Fj 2.88 x 10-21 1 [19]


Since the Jovian flux is proportional to and D is a

measure of the highest pulse in a ten-minute interval, a plot of

-- 1 versus longitude will reveal the variation of Jovian peak
G2
flux density with longitude. As before, a histogram table was compiled,

this time using values of (2- 1) The average value for each 50

increment of longitude was calculated. The result is the flux density

histogram in Figure 33. The most prominent feature is the dip centered

at about 60. The dashed and solid lines at the bottom of the figure

designate longitude intervals in which less than 10 or more than 20

Jupiter noise storms occurred respectively. The sections of the curve

below which there is no line correspond to longitude segments in








77





o
0
Sto






a,
II- -


--1 o .-
co
a)


0 s



-o
a)




0 r-
----, o_ Q





0 -0
Cv H



0a E^
H '




,-I i









o, ______-_ 0


tO

to
0















Cr -) DoaAV









which 10 to 20 decameter storms occurred. It is interesting to note

that the features which stand out on the histogram (the dip at 600 and

the peak at 1100) occur where the statistics are the poorest. Where

the statistics are the best, from 1250 to 3300, there are no prominent

maxima or minima. There is almost no correlation between this curve

and the 18 Mc/s Chile probability histogram of 1961. Thus, the flux

density histogram supports our previous conclusion that there is no

well-defined correlation between the intensity of the decameter radia-

tion and Jovian longitude.

To convert the ordinate scale in Figure 33 to values which

correspond to peak Jovian flux densities, multiply by 2.88 x 10-

watts/m2/cps. If we assume that the average flux is proportional to

the peak flux, i.e.,

Fj K (Fj) max, [20]


then all we need to know in order to convert the ordinate scale to

values which correspond to average Jovian Flux densities is the

constant K. G. W. Brown, a member of the radio astronomy group at the

University of Florida, has recently determined K to be approximately

0.11. If the ordinate values in Figure 33 are multiplied by the two

factors 2.88 x 10-21 and 0.11, then they become average Jovian flux

densities in watts/m2/cps.









Activity Studies


1961 Activity Plots

Up to this point we have been examining probability and inten-

sity histograms to obtain information about the longitude distribution

of the decameter radiation from Jupiter. These investigations have

been referred to as "source studies." Now we wish to take another

point of view and find out how the Jovian activity varies from day to

day. Such information is important in searching for suspected corre-

lations with other daily indices.

A measure of the Jovian decameter activity on a particular day

is the "daily activity index rate." This quantity is computed on the

IBM 709 and is explained at the beginning of this chapter. Briefly,

it is the sum of the activity indices of all Jovian storms on a given

date, divided by the total duration in minutes of the listening period

on that date. Remember that the activity index of a Jovian noise storm

was defined as the product of the intensity, the duration in minutes,

and a normalization factor. The duration of the listening period

appears in the denominator of the daily activity index rate calcula-

tion in order to remove the effects of variation in the amount of good

receiving time from night to night. If a watch period lasted for six

hours, one would expect to get more Jovian noise than during a night

when listening conditions were such that there was only one hour of

good reception.

Figures 34 through 38 contain daily activity plots of the

1961 data. Those frequencies which were covered at two stations










appear in a single figure with data from one station plotted above

the other, corresponding dates lying on a vertical line. The starting

and ending dates of the watch, as well as periods of no monitoring, are

marked on the plots. Dates when listening conditions were poor due to

atmospheric or man-made interference are interference are signified by

"x." Dates when there was no watch due to equipment failure or absence

of. the observer are marked "0." The plot of the Florida data in

Figure 35 shows the dates of zenocentric inferior conjunction of Mer-

cury, Venus, Earth, and Mars. This will be discussed later in the text.

The 15 Mc/s, 18 Mc/s, 22.2 Mc/s, 27.6 Mc/s, and 10 20 Mc/s data are

found in Figures 34 through 38, respectively. During the 1961 appari-

tion, there were only eight nights when the ionosphere at the Chile

station permitted reception at 5 Mc/s. Although a constant vigil was

maintained in Florida at 31 Mc/s for a five-month period, only five

pulses of possible Jovian origin were received. For these reasons, no

activity plots were constructed for 5 Mc/s and 31 Mc/s.

Figure 39 was obtained by combining the two activity plots of

Figure 35. This was accomplished by recording the larger of the two

values of daily activity index rate for each day. By doing this, we

consider only the best Jovian reception and eliminate many of the

"no-data-days" when atmospheric conditions blanked out reception on one

station or the other. Figure 359 gives the best indication of Jupiter's

activity from day to day at 18 Mc/s during the 1961 apparition. A sim-

ilar combination of data was performed on the activity plots in Fig-

ure 56, and the resulting graph is found in Figure 40, which shows

Jupiter's 22.2 Mc/s activity during 1961.











No watch---------- -------


I Watch starts


10 20 50 10 20
2.5 Mar. Apr.
2.3,_


-I L.. .[,- --J-- -T j- --L--r 2 r.-
30 10 20 30 10 20
May June


30 10 20
July


30 10 20 30
Aug.


10 20
Sept.


-II.---I-- -----I-- .... -- 1>
S0 10' 20 30 10 20
Oct. Ilov.


10 20
Dec.


30 10 20
Jan.


'Watch starts
I


I, 1


I I I


1 2.95


S-----No watch--


Legend: x Interference
o No watch


- --.I I Y'L I--111h-r'" I .1 j |111| iII4[ I I-I.- 1jJJ^4lJl- IL4J11 11. 1. ll,1 v1 ..d.^ ^


10 20 50 10 20
Mar. Apr.


30 10


20 50


10 2
June


30 10


20
July


30 10


20
Aug.


30 10 20
Sept.


30 10 20 50 10 2
Oct. Nov.


Figure 34.--15 Mc/s activity plots of 1961 data.


- Florida


2.0 L Chile


1.5 L


End I


End


__~


)l- _.-.AH -.. ,- b&.0 ,x x A. MI"tIInlnm


~---


1


--T


I


I '














I Watch starts


I i I,.


1.0




.5




I,

m2.0



E-4
15




1.0


30 10
May


20 30


10 ;
June


Watch starts


II


30 10 20
July


Legend: Interference
o No watch


I L.I.. .- ;, .... .


Aug.


ulo watch


20 30


10 20
Sept.


30S 10
Oct.


20 30 10


I I I


EndI


20
Nov.


Legend: ? Mercury
? Venus
s Earth


End I


I -I. l, Ii IHi .I III II I' 1L l l lll' III l I lllA.H 1. .11..! L .. I II J 11.J.I I I Ikl 1 1111 11 [l 2 I ILI If. I I 1_ II J ,


50 10 20
1---- Apr.



30 10
May


20 50 10 20
1 June


30 10 20
I-- July


30


10 20 30 10 20
Aug. Sept.


30 10
1----


20
Oct.


10 20
Nov.


30


10 20
Dec.


30 10 20
1 -i Jan. '62


Figure 55.--18 Mc/s activity plots of 1961 data.


Chile


10 20
Apr.


--
10 20
Feb.


,I


10 20
Mar.


30


Florida


I


10 20
*Feb. '61


10 20
Mar.


101 11. 4 -1 H 1. -t --)llL 11 3 Y


i II !. I !I 1 1 I 1 1 1 .1 ,-L -rL


I


I,


Al I


I


II I










Legend: Interference
o- No watch


Chile


I Watch starts


INo watch


EndI


11----- 1 ,1 -L, -1-1- I -I .-1 --[,- -,- -,I-tlI-J L ._t r L., i.I.,l ,-l .4 I, 11, ,--, I -- T --r -- -- -\ > -
10 20 30 10 20 30 10 20 30 10 20 50 10 20 30 10 20 30 10 20 50 10 20 30 10
Mar. Apr. May June July Aug. Sept. Oct. Nov.


Florida


Watch starts Feb. 25


1.OL


- --- -. r--L, -_ -,
10 20 30 10 20
Mar. Apr.


30 10 20 30 10 20 30 10


June


July


-- No watch -- Endl


20


---.-_ 1--- A -f -T r ,- _
30 10 20 30 10 20 30
Aug. Sept.


10 20 30 10 20 30
Oct. Nov.


Figure 56.--22.2 Mc/s activity plots of 1961 data.


.$5



























Chile


pWatch starts


INo watch


End


Legend: x Interference
o No watch


I I I I .1 .1 1. 1 1.. 6


Florida



.0 IWatch starts





.5







10 20 10 20 30 10 20 30 10 20 30 10 20 30 10 20 30
Feb. Mar. Apr. May June July


3D 20 30
Aug.


10 20 30 10 20 30
Sept. Oct.


10 20
Nov.


Erndi


1- 114 -1,1 -, --1 41 Sf2


V j


I A XI f 0 I -. .' 0- .


10 20 30 10 20 30 10 20 30 10 20- 30
Sept. Oct. Nov. Dec.


10 2c 30
Jan. '62


Figure 37.--27.6 Mc/s activity plots of 1961 data.


I


I


.5 L








Chile
10 Mc/s


lWatch starts


. .,L,, U. i-. ,- .1t ,
10 20 30
Mar.


10 20
Apr.


30 10


1.0 Chile
20 Mc/s


I Watch starts


II
20 30 C10 20 30 10 20
y June July


I,







3il
30


.5 I


10 20 30 10
Mar. 1


20 30 10 20 30 10 20 30 10


June


July


1i0 20 30
Aug.

t1.8


10 i
h 'g.


End


10
Sept.


INo watch


30 10
Sept.


End I


-c--i7-
20 30 10 20 30 10 20 30
Oct. Ilov.


Figure 3&.--10 Mc/s and 20 Mc/s activity plots of 1961 data.


Legend: Interference
o lo watch