Title: Analytical studies of selected Jovian decametric phenomena
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00097735/00001
 Material Information
Title: Analytical studies of selected Jovian decametric phenomena
Alternate Title: Jovian decametric phenomena
Physical Description: xii, 120 leaves. : illus. ; 28 cm.
Language: English
Creator: Olsson, Carl Niels, 1930-
Publication Date: 1970
Copyright Date: 1970
 Subjects
Subject: Jupiter (Planet) -- Observations   ( lcsh )
Physics and Astronomy thesis Ph. D
Dissertations, Academic -- Physics and Astronomy -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis - University of Florida.
Bibliography: Bibliography: leaves 117-119.
Additional Physical Form: Also available on World Wide Web
General Note: Manuscript copy.
General Note: Vita.
 Record Information
Bibliographic ID: UF00097735
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 001148183
oclc - 20397132
notis - AFP7872

Downloads

This item has the following downloads:

PDF ( 4 MBs ) ( PDF )


Full Text








ANALYTICAL STUDIES OF SELECTED JOVIAN

DECAMETRIC PHENOMENA














By
CARL NIELS OLSSON












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





































To

Jean, Kris, and Kay













ACKNOWLEDGEMENTS


I wish to thank the members of my committee, Drs. Alex G. Smith,

Guy C. Omer, F. Bradshaw Wood, and Daniel C. Swanson. I am especially

indebted to Dr. Alex G. Smith, the chairman of the committee, for his

help and encouragement throughout my association with the radio astronomy

group.

Many others have been of great help through their suggestions

and discussions. They are: Drs. Thomas D. Carr, George R. Lebo,

Henry I. Register, Jorma J. Riihimaa, Izzy Shever, David J. Kennedy,

Messrs. Robert J. Leacock, Jorge May, Jorge Levy, Ray Sisler and Mrs.

Dolores Krausche.

I am thankful for the help I received from Mrs. Jan Schultz and

Mrs. Judy Lipofsky with my computer programs.

As all others before me, I owe a great deal to all of the members

of the Florida and Chile groups who have served in the nightly observations.

I am especially thankful to Mr. W. W. Richardson. He always found

time to help me with the many aggravating computer problems which arose,

and he is also responsible for the excellent figures used in this

dissertation.

A special thanks is due Dr. Stanley S. Ballard, Chairman of the

Department of Physics and Astronomy, for his support, both moral and

financial, throughout this work.

Finally I wish to thank my wife, Jean, for her encouragement

and for typing this dissertation.
iii












TABLE OF CONTENTS


ACKNOWLEDGEMENTS .

LIST OF TABLES .

LIST OF FIGURES . .

ABSTRACT . . .

Chapter


I. INTRODUCTION .. ...

II. THE DATA RESERVOIR . .

Observatories ...

Instrumentation and Raw

Other Data . . . .

III. LONGITUDE DISTRIBUTION C

Types of Pulses .

Io Influence .. ...

IV. VARIATIONS OF ROTATION I

Introduction .. ...


Data . . . .




)F THE MILLISECOND






PERIODD BY SOURCE AN


PULSES . .






ID BY FREQUENCY


Method of Analysis . . .....

Frequency and Source Dependence of the Drift . . .

V. A STUDY OF THE ROTATION PERIOD BASED ON STORM
COMMENCEMENT TIMES .. . . . .. ......

Introduction . . . . . . . . ..

Duncan's Hypothesis .........


Page

iii

vi

vii

xi







Chapter Page

V. A STUDY OF THE ROTATION PERIOD BASED ON STORM
COMMENCEMENT TIMES (Continued)

Analysis of Florida and Chile Observations .. ... 61

Storm Durations . . . . . . . . . 77

VI. SATELLITE INFLUENCES ON THE RADIATION . . . ... 93

Hypothetical Satellites . . . . . . . 93

The Florida Analysis. .. . . . . . . . 94

Amalthea . . . . . . . . . ... . 99

VII. SUMMARY AND CONCLUSIONS .. . . . . . . 100

APPENDIX . . . . . . . . ... . . . . . 103

LIST OF REFERENCES . . . . . . ... ...... 117

BIOGRAPHICAL SKETCH . . . . . . . . . . 120











LIST OF TABLES


ble Page

1. ANTENNAS AT THE FLORIDA AND MAIPU OBSERVATORIES. ... 11

2. DATA COLLECTED ON SLOW-SPEED RECORDER. . . .... . 13

3. LOCATION OF THE SOURCES. . . . . . . . ... 41

4. YEARLY AVERAGE DRIFT RATE OF SOURCES . . . . .. 46

5. LONGITUDE DIFFERENCES IN SOURCE LOCATION
AT DIFFERENT FREQUENCIES .... . . . .. 54

6. DIFFERENCES IN LONGITUDES OF THE SOURCES
AT FIXED FREQUENCY. . . . . . . . ... 55

7. STORM STATISTICS . . . . . . . . . . 90













LIST OF FIGURES


figure Page

1. Typical reception probability histogram as a function
of System III (1957.0) CML, showing the approximate
locations of the four sources. . . . . . . . 4

2. Location of the University of Florida Radio
Observatory and Chilean field stations . . .... ... 10

3. Average pulse character as a function of System III
CML for 18 MHz during 1952 . . . . . . .... 16

4. Number of 30-second periods containing character 2
and millisecond pulses in 1961.2 at 18 MHz . . ... 18

5. Number of 30-second periods containing character 2
and millisecond pulses in 1963.8 at 18 MHz . . ... 19

6. Number of 30-second periods containing character 2
and millisecond pulses in 1964.8 at 18 MHz . . ... 20

7. Number of 30-second periods containing character 2
and millisecond pulses in 1963.8 at 22.2 MHz ...... 21

8. Relative number of intervals of 18 MHz type 2 pulses
as a function of longitude for various positions of Io 24

9. Relative number of intervals of 18 MHz millisecond
pulses as a function of longitude for various positions
of Io ....... . . . . . . . . .. 26

10. Contour plot of 18 MHz character 2 activity as a
function of longitude and the position of Io for
the 1963 Florida-Maipu apparition. . . . . . ... 29

11. Contour plot of 18 MHz millisecond activity as a
function of longitude and the position of Io for
the 1963 Florida-Maipu apparition. . . . . . . 31

12. Number of intervals including 18 MHz millisecond
pulses versus the longitude of lo. . . . .... . 34




.gure Page

13. Polar plot of 18 MHz millisecond pulse activity
for the indicated range of Io positions centered
on 2100. . . . . . . . . . ...... 35

14. Polar plot of 18 MHz millisecond pulse activity
for the indicated range of Io positions centered
on 270. .. . . . . . . ...... . 36

15. Apparent longitude drift of the sources at 15
and 18 MHz . . . . . . . . ... .. .... 43

16. Apparent longitude drift of the sources at 18
and 22.2 MHz . . . . . . . . . .. . . 44

17. Apparent longitude drift of the sources at 18
and 27.6 MHz . . . . . . . . ... . . 45

18. Source drift rate versus frequency for 1964.9. . . ... 48

19. Source drift rate versus frequency for 1963.9. . . ... 49

20. Drift rate of source B as a function of time ...... 50

21. Drift rate of source A as a function of time ...... 51

22. Drift rate of source C as a function of time ...... 52

23. Positions of sources A, B, and C at various
frequencies in 1962. . . . . . . . . ... 57

24. Commencement time of storms histograms of source A
for 1965.9 to 1968.1 at 18 MHz . . . . . .. 63

25. Commencement time of storms histograms of source A
for 1958.1 to 1964.8 at 18 MHz . . . . . ... 64

26. Commencement time of storms histograms of source A
for 1960.2 to 1967.0 at 18 MHz . . . . . ... 65

27. Commencement time of storms histograms of source A
for 1958.1 to 1968.1 at 22.2 MHz . . . . . ... 66

28. Commencement time of storms histograms of source A
for 1960.2 to 1967.0 at 22.2 MHz . . . . . ... 67

29. Merged histogramsof commencement time of storms. ... . 68

30. Histograms showing commencement fronts with dashed
lines for 1964.8 to 1968.1 at 18 MHz . . . . ... 69






Figure

31. Histograms showing commencement fronts with dashed
lines for 1958.1 to 1963.8 at 18 MHz . . . . . .

32. Histograms showing commencement fronts with dashed
lines for 1964.8 to 1968.1 at 22.2 MHz .. . .....

33. Histograms showing commencement fronts with dashed
lines for 1958.1 to 1963.8 at 22.2 MHz .. . .....

34. Location of the commencement front of source A
during ten apparitions of Jupiter .. . . .....

35. Motion of the center of source A in the storm commencement
histograms for 18 and 22.2 MHz .. . . .......


36. A conventional probability histogram .


37. Distribution
to 1961.6 at

38. Distribution
at 18 MHz. .

39. Distribution
at 18 MHz. .

40. Distribution
at 18 MHz. .

41. Distribution
at 18 MHz..

42. Distribution
at 18 MHz. .

43. Distribution
to 1962.7 at

44. Distribution
at 22.2 MHz.

45. Distribution
at 22.2 MHz.

46. Distribution


of storms within source A


18 MHz.

of storms


of storms


of storms


of storms


of storms


of storms


within source A


within source A


within source A


within source A


within source A


within source A


22.2 MHz . . . .

of storms within source A


of storms within source A


of storms within source A
of storms within source A


1959.1


1962.7


1963.8


1964.8


1965.9


1967.0


1959.1


1963.8


1964.8


1965.9


to 1967.0 at 22.2 MHz .. . ..


Over-all distribution of storm duration for source A .

A history of mean storm duration .. . .......








49. Reception probability versus the position of
satellite A . . . . . ... .

50. Reception probability versus the position of
satellite A with axis shifted .. . ...

51. Reception probability versus the position of
satellite A with To-related data omitted .

52. Florida-Chile merged histogram at 18 MHz for


Florida-Chile

Florida-Chile

Florida-Chile

Florida-Chile

Florida-Chile

Florida-Chile

Florida-Chile

Florida-Chile

Florida-Chile

Florida-Chile

Florida-Chile

Florida-Chile


merged histogram at

merged histogram at

merged histogram at

merged histogram at

merged histogram at

merged histogram at

merged histogram at

merged histogram at

merged histogram at

merged histogram at

merged histogram at

merged histogram at


1960.2. .

1961.6. .

1962.7.

1963.8. .

1964.8. .

1965.9.


18 MHz for 1967.0. .


1960.2. .

1961.5. .

1963.8. .

1964.8. .

1965.9. .

1967.0. .


65. Florida-Chile merged histogram at 27.6 MHz for 1963.7. .


Page





Abstract of Dissertation Presented to the Graduate Council
in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy


ANALYTICAL STUDIES OF SELECTED JOVIAN
DECAMETRIC PHENOMENA

By

Carl Niels Olsson

December, 1970


Chairman: Alex G. Smith
Major Department: Astronomy

The Jovian 18 and 22.2 MHz radiation data, which were collected

at the University of Florida Radio Observatory and the Maipu Radio-

astronomical Observatory on high-speed recorders, were separated into

L-bursts and S-bursts. These two groups of data were then independently

studied to determine their distributions in longitude, and also examine

to what extent the bursts were influenced by the satellite Io. The

noise storms from sources B and C contained a greater percentage of

S-bursts than those from source A. The L-bursts from source A seemed

to be insensitive to the position of lo, whereas the source A S-bursts

showed an Io effect. Both types of bursts from sources B and C were

influenced by lo.

All of the Florida and Maipu data at the frequencies of 15, 18,

22.2, and 27.6 MHz were investigated in an attempt to determine any

possible frequency or source dependence of the sinusodial drift of the

sources. The results of this study were negative, possibly due to the

difficulty in locating sources B and C at the lower frequencies and for

apparitions of low activity.

The 18 and 22.2 MHz data were analyzed in order to theck the

xi






hypotheses of Duncan that (1) the radio rotation period is different

from the System III (1957.0) value when it is determined by considering

the commencement times of source A storms for each apparition, and (2)

that the drift is due to a variation in the duration of the storms.

From this study it was concluded that this method of determining the

rotation period is questionable and that the drift cannot be accounted

for by a variation in the storm durations.

Finally, the question of possible satellite influence on the

radiation by Amalthea and by two hypothetical satellites was considered.

No significant influence was detected in this study.


xii















CHAPTER I

INTRODUCTION


The Jovian decametric emission has been the subject of intensive

study for some fifteen years, but as yet no wholly acceptable theory has

resulted.

As a preliminary to the study covered in the following chapters

of this volume, one should briefly review the present state of our

knowledge concerning the decametric radiation from Jupiter.

Although the Jovian radio signals are highly sporadic, they tend

to occur as noise storms of durations measured in minutes up to very long

ones lasting several hours. Early attempts to correlate these storms

with System II central meridian longitude (hearafter abbreviated CML),

an optically derived system, suggested that the storms occurred only

when certain longitudes were on the central meridian; that is, there

seemed to be three or four source regions associated with the planet.

Further study indicated that the radio sources drifted in longitude

in System II.

As more data were collected and studied, it became apparent to

many groups of investigators (Shain 1956; Gallet 1961; Burke 1957;

Garnder and Shain 1958; Carr et al. 1958, 1961; Smith and Carr 1959;

and Douglas 1960, 1960b, 1962) that the sources could be used to

define a radio rotation period -- that is, a longitude system that

would minimize any apparent drift of the sources. Based on the above

studies, a System III (1957.0) longitude with a period of 9h55m29s37








(slightly shorter than System II) was tentatively adopted by the Inter-

national Astronomical Union (IAU 1962) in 1962 for specifying the

CML of radio features of Jupiter.

At present there are three strong sources and one rather weak

source identified; they are shown in Figure 1. Their names and approximate

locations are as follows:

Source A or Main source between longitudes 200'-3000

Source B or Early source between 100-200

Source C or Late source between 3000-20

Source D or Fourth source between 200-1000

In 1962 (Douglas & Smith 1963a Smith et al. 1965a) evidence

appeared that the sources were drifting in System III, implying a rather

sudden change in the radio rotation period. The sources drifted steadily

to higher longitudes at about 100/year until about 1965, when the

sense of drift reversed. This apparent source drift will be considered

in detail in following chapters.

The decametric noise storms consist of relatively short bursts

of radiation. Most of the bursts fall into two categories -- the

L-bursts, by far the most common, with durations from a few tenths

of a second to several seconds and the more rapid S-bursts, whose

durations are measured in milliseconds. On rare occasions a third

type has been observed, lasting tens of seconds. It now appears that

the fine structure of the L-bursts is due to scintillations produced

in the interplanetary medium (Douglas and Smith 1967), whereas there

is strong evidence that the S-burst structure is generated at Jupiter

(Smith et al. 1966).

Several groups have studied the morphology of the fine structure































r4 0
C

0 -.4
4J
Cu
0 C

U
ci

H 0c


x

CO-



o 0




4 0


-Hl M




0 m
o-


-co
C-u




o ~


-4



U 0








OB
ccu
*l4 J 4





















0







So
0-
T -------- ,~ o -






"J












N -- J
LL


-0

uo O

Q










C N -'- o






0 10


All!118V0 OO-d








of the bursts using both fixed-frequency receivers and wide-band receivers.

These results are reviewed in the recent paper of Carr and Gulkis (1969)

and also in one by Riihimaa et al. (1970).

The longitudinal distribution of the L- and S-bursts has been

studied by the Florida State University group (Baart, Barrow, and Lee

1966; Barrow and Baart 1967) and the University of Florida group (Smith

et al. 1966; Olsson and Smith 1966 and 1967) and will be discussed in

detail in Chapter 3 of this dissertation.

One of the most exciting discoveries concerning Jupiter came in

1964, when the Australian E. K. Bigg (1964) showed that the inner most

Galilean satellite, lo, has a very marked influence on the probability

of receiving decametric radiation from the planet. Bigg found that

the probability of receiving emissions from source B was greatly enhanced

when Io was in a position in its orbit of about 90 from geocentric

superior conjunction, and from sources A and C when Io was at 2400.

The Io effect is more pronounced for sources B and C than for source A.

It has also been shown that the Io influence is strongest at the higher

frequencies (27-30 MHz) and becomes progressively weaker at lower

frequencies (Lebo et al. 1965; Register 1968).

Bigg's discovery of the Io effect inspired others (Lebo et al.

1965; Dulk 1967; Tiainen 1967; Register 1968) to investigate possible

influences on the emissions by the three other Galilean satellites.

However, to date any effect they may have is questionable. Moon V or

Amalthea, the satellite closest to Jupiter, and two as yet hypothetical

moons (Bigg 1969; Douglas and Bozyan 1970) have also been studied for

possible influences, with negative results.

Soon after the discovery of Jupiter's radio emission the co-








discoverers, Burke and Franklin (1956), also showed that the bursts

are most often circularly or elliptically polarized in the right-hand

sense, which strongly suggested that Jupiter is surrounded by a magne-

toionic medium. The existence of a Jovian magnetosphere was soon

confirmed by the many observers studying the decimeter radiation from

the planet. The analysis of the decimeter radiation is basically

responsible for our present-day model of the Jovian magnetic field,

giving such information as the orientation of the field with respect

to the rotational axis, direction of the field, and an approximate

idea of its strength.

Polarization studies in the decameter range have been carried

out by several observers (Carr et al. 1965b; Barrow and Morrow 1968;

Sherrill 1965; Kennedy 1969). One finding of particular importance

is that while at the higher frequencies the polarization is predominately

right-hand, it becomes more and more left-hand with a decrease in

frequency. This fact has been useful in studying lo-related and non-

lo-related emission,in deducing which hemisphere the emission comes

from, and even in helping to locate the fourth or D source. A complete

and up-to-date review of the polarization phemonema can be found in

the dissertation of Kennedy (1969) and the paper by Carr and Gulkis (1969).

Recent refinements of technique in very long baseline (VLB)

interferometry (Carr et al. 1965a; Slee and Higgins 1966; Brown et al.

1968, Block et al. 1970; Dulk 1970) have made it possible to set an

upper limit on the sizes of the sources. Dulk (1970) has recently

quoted an upper limit for an incoherent source as 0''10 or 400 km at

the planet, using an interferometer with a baseline of 487000X operating

at 34 MHz. Hopefully, VLB interferometry can also determine the positions









of the sources relative to the disc of the planet if one can indepen-

dently determine the location of the central fringe of the interfero-

meter pattern.

Any acceptable model of Jupiter's decametric radiation must be

able to account for all of the above observational results. The two

discoveries which have caused most theories the greatest difficulty

are the apparent drifting of the sources in longitude and the Io effect.

The recent model of Goldreich and Lynden-Bell (1969) fits most of the

data and seems to be the most promising at present.

Of the above-mentioned observations and analyses, the present

study is concentrated in the three broad areas of the longitudinal

distribution of the pulses, source drifts, and satellite influences.

In particular, the following will be considered:

1. The longitude distribution of L- and S-bursts

2. Influence of Io and L- and S-bursts

3. Frequency dependence of source drift

4. Variation of drift among sources

5. Duncan's (1967) hypothesis of source drift

6. Durations of the storms

7. Influence of Amalthea and the two "new" satellites

hypothesized by Bigg.

The investigations listed above, and to be described in later

chapters, have led to several publications in the open literature. In

particular, item 1 is summarized in the following papers:

"Jovian Rotation Periods and the Origin of the Decametric
Burst Structure," A. G Smith, G. R. Lebo, C. N. Olsson,
W. F. Block, N. F. Six, and T. D. Carr, Proceedings of
CALTECH-JPL Lunar and Planetary Conference, California
Institute of Technology (1966), pp. 128-133








"Decametric Radio Pulses from Jupiter: Characteristics,"
C. N. Olsson and A. G. Smith, Science 153, 289-290 (1966)

"Influence of the Terrestrial Environment on the Temporal
and Statistical Characteristics of Jovian Decametric Radiation,"
A. G. Smith, W. F. Block, W. A. Morton, G. R. Lebo, T. D. Carr,
and C. N. Olsson, Radio Science, 1 (new series), 1167-1171
(1966)

"Delineation of Jovian Decametric Sources by Millisecond
Pulses," C. N. Olsson and A. G. Smith, Nature, 214, 999-1001
(1967).

Item 2 is summarized in the paper "Influence of lo on Millisecond

Pulses of Jovian Decametric Radiation," C. N. Olsson and A. G. Smith,

Bull. Am. Phys. Soc., 11, 853 (1966). Items 3 and 4 are summarized in

the papers "Jovian Rotation Periods," C. N. Olsson, A. G. Smith, and

G. R. Lebo, Bull. Am. Phys. Soc., 11, 513 (1966) and "Apparent Drift of

the Radio Rotational Period of Jupiter," C. N. Olsson and A. G. Smith,

Astron. J., 73, S30 (1968). Items 5 and 6 are summarized in the following

papers:

"Analysis of the Jovian Decametric Radiation Based on the
Beginning of Storms," C. N. Olsson, H. I. Register, and
A. G. Smith, Bull. Am. Phys. Soc., 13, 1714 (1968)

"Analysis of the Jovian Decametric Source Drift Based on
the Beginning of Storms," C. N. Olsson, H. I. Register, and
A. G. S mith, Bull. Am. Ast. Soc., 1, 202 (1969)

"Commencement Times and Durations of Jupiter's Radio Noise
Storms," C. N. Olsson, A. G. Smith, H. I. Register, and
J. May, Icarus. 11, 212 (1969)

"Jupiter's Decametric Rotation Period as Defined by the
Commencements of Source A Storms," C. N. Olsson, A. G. Smith,
H. I. Register, J. May, Bull. Am. Phys. Soc., 15, 203 (1970).















CHAPTER II

THE DATA RESERVOIR

Observatories


The University of Florida maintains two major radio observatories;

one is located in Florida and another in Chile. The Florida observatory

was situated in Gainesville from December of 1956 until the summer of

1967, when it was moved to a more remote site in Dixie County. In 1959

the University of Florida, with the cooperation of the University of

Chile, established a southern hemisphere radio observatory near Maipu,

Chile, a suburb of Santiago. These two observatories, located approxi-

mately the same distance from the equator but in opposite hemispheres,

assure the best possible monitoring of Jupiter year after year as the

planet changes declination. In 1964 a field station was established

in an Andean valley about 300 miles north of Santiago near the town of

Huanta. The locations of these observatories are indicated in Figure 2.

All of the data used in the following analyses were collected at

either the University of Florida Radio Observatory or the Maipu Radio-

astronomical Observatory.


Instrumentation and Raw Data


The radio-astronomical instrumentation at the observatories includes

the antenna systems listed in Table 1. The signal from each antenna is

amplified by a high quality communications receiver and recorded on pen
































































Figure 2. Location of the University of Florida Radio
Observatory and Chilean field stations.













TABLE 1

ANTENNAS AT THE FLORIDA AND MAIPU OBSERVATORIES


Location

Florida

Florida

Florida

Florida

Florida

Maipu

Maipu

Maipu

Maipu

Maipu

Maipu

Maipu

Maipu

Maipu

Maipu

Maipu


Frequency

12.5 MHz

15 MHz

18 MHz

22.2 MHz

27.6 MHz

12 MHz

15.8 MHz

18 MHz

22.2 MHz

27.6 MHz

30 MHz

32 MHz

34 MHz

36 MHz

38 MHz

40 MHz


Type

H Broadside

Yagi

2 Yagis

Yagi Polarimeter

Yagi

Yagi Polarimeter

Yagi Polarimeter

Yagi

Yagi Polarimeter

Yagi Polarimeter

Broadside

Broadside

Broadside

Broadside

Broadside

Broadside


No. of Elements

4

5

5

5

7

4

3

5

5

7

15

15

15

15

15

15









recorders. All of the data except those used in Chapter III were

recorded on Texas Instruments Dual Rectiriter recording oscillographs,

which are slow-speed recorders. The data in Chapter III were recorded

on high-speed pen recorders, Brush Mark II oscillographs. The slow-

speed records were reduced and punched on computer cards using the

method described by Register (1968). The method used in reducing the

high-speed records is outlined in Chapter III. Table 2 lists information

concerning these data.


Other Data


Throughout this dissertation the Jovian longitude system used

is the System III (1957.0) as defined by the International Astronomical

Union (1962). This system rotates with a constant period of 9h55m29s37.

The periods of the Jovian satellites used in this study were obtained from

the American Ephemeris and Nautical Almanac.














TABLE 2

DATA COLLECTED ON SLOW-SPEED RECORDER


Frequency Mean Epoch Station

12 MHz 1966.0 Florida


15 MHz







18 MHz



















22.2 MHz


1961.5
1962.7
1963.8
1964.8
1965.8
1967.0

1957.1
1958.1
1959.1
1960.2
1960.4
1961.6
1961.6
1962.6
1962.7
1963.7
1963.8
1964.8
1964.9
1965.9
1966.0
1967.0
1967.0

1958.1
1959.2
1960.2
1960.4
1961.5
1961.5
1962.8
1963.7
1963.8
1964.8
1964.9
1965.9
1965.9
1967.0


Maipu
Florida
Florida
Florida
Florida
Florida

Florida
Florida
Florida
Florida
Maipu
Florida
Maipu
Maipu
Florida
Maipu
Florida
Florida
Maipu
Florida
Maipu
Florida
Maipu

Florida
Florida
Florida
Maipu
Florida
Maipu
Florida
Maipu
Florida
Florida
Maipu
Florida
Maipu
Florida


Listening Hours

590

934
610
830
1530
540
830

195
278
376
324
779
1145
1190
1200
1115
1130
1570
1630
770
1550
735
1100
635

366
320
290
964
610
1190
1030
1000
1590
1610
675
1580
620
1250


Activity Hours

102

176
49
102
107
45
109

42
13
15
28
72
162
150
206
182
160
284
191
50
176
46
124
13

11
6
11
31
33
59
75
110
143
103
36
81
28
57









TABLE 2 (CONTINUED)

Frequency Mean Epoch Station Listening Hours Activity Hours

27.6 MHz 1959.2 Florida 278 1
1960.2 Florida 417 3
1961.6 Florida 960 25
1962.5 Maipu 500 76
1962.7 Florida 700 26
1963.7 Maipu 1140 32
1963.8 Florida 1520 39
1964.8 Florida 1410 24
1965.9 Florida 1370 21
1965.9 Maipu 880 15
1967.0 Florida 1180 14


DATA COLLECTED ON HIGH-SPEED RECORDERS

18 MHz 1961.5 Florida
1963.7 Maipu
1963.8 Florida
1964.8 Florida
1964.9 Maipu

22.2 MHz 1953.8 Florida
1964.9 Maipu















CHAPTER III

LONGITUDE DISTRIBUTION OF THE MILLISECOND PULSES

Types of Pulses


The sporadic decametric radio storms from Jupiter are received

as bursts of pulses. Three distinct types of pulses are discernible

when one studies the records from a high-speed recorder providing

moderately good time resolution (speed of 5mm/sec). On very rare

occasions there are pulses lasting tens of seconds. Most of the

emission consists of pulses which last from tenths of a second to several

seconds; these are generally referred to as "normal" or "swishing"

pulses. The third type have durations measured in milliseconds; because

of the distinctive sound produced by a train of these pulses, they

have been described as "popping" or "spitting". For this study a pulse

character number was assigned to the three types of pulses. Character 1

is the very long pulses, 2 the normal pulses and 3 the millisecond

pulses.

With the help of the high-speed pen recordings, a character number

was assigned to each 30-second period of Jovian emission. Periods

containing pulses of mixed types were given half-integral character

numbers; if, for example, a period contained pulses of types 2 and 3,

it was assigned a character number of 2.5.

Figure 3 shows the average character number as a function of

System III Jovian longitude. For convenience in analysis the planet












L I











0
I 0~










I
I 0





IW
N 0









z
co

I
4-




I
0I
SU
--










I 0
n 2
.i C nA
~ fl
-N
C ^l B
'CV4-JVH
J r,^ *



*-I~B TUJ
r-3 ^
------ J --I h
s ^
H ^
i -J ii


_| 0

{-J \
-- | 1 _o S
--- i~ i "
L c
n3.'llH iSn <"~""~
I --- *--i b0


(0




IL
t0


-J








was divided, as usual, into 5 zones of longitude. In determining the

average character for a given 50 longitude zone, a computer program

summed the number of 30-second intervals of emission that fell within

that zone, weighting each by its own character number, and this sum

was then divided by the total number of intervals. It should be remem-

bered that a character above 2.0 means that at least some millisecond

radiation occurred for that zone. Thus the figure suggests that the

millisecond bursts tended to cluster in the longitude region from 1100

to 2000, or in the general vicinity of decametric source B, the same

conclusion reached in the preliminary study of the 1962 data (Smith

et al. 1966). There is an indication of further clustering near source C

(3100 to 3600).

It is perhaps even more instructive to analyze the pulses in

the manner shown in Figures 4 through 7. Here the computer has been

used to total the number of 30-second intervals of each character

(1, 1.5, 2, 2.5, 3.0) that occurred while each 5 longitude zone was

on the central meridian. In the figures, types 2.5 and 3 have been

combined to produce the shaded regions of millisecond radiation, while

the unshaded histograms represent emission of type 2 only. No corre-

lation was attempted for types 1 and 1.5, since these classes formed

less than one per cent of the 18 MHz data analyzed.

The striking feature of these diagrams is that while the milli-

second radiation is concentrated in the familiar sources A, B, and C,

it seems to occur with roughly equal probability in all three sources.

On the other hand, type 2 emission is heavily concentrated in source A;

in 1963, for example, 74% of the character 2 activity arose from this

source. This result of course explains the form of Figure 3. Millisecond



























































r4







-J
IL NI >


E 5


E1m


18



















CO
0






o
I-4




co

0rl






-4


.-
ca
0r



1-4
0

u




00
C







0,




ao

o 4J






Sc-





3C
i-4
Z


0

SIVAU3IJNI zIO UJ3aflN





19




















c
o
























-r-I
0
u







m











0
0





r4-
0





























ot
rl
a)


































CO
-0
to




C





0,
*H
1-

















3 C

a r

in(


o 0 o 0
0 0 0 0


S7VA83INI 0O J:9VInN


















C)


















4
C
0




o




a
J u
C) rrC

.4l























0
-H
o


0 -- a)

41-
c















U 0
n40



















ID
0 2
0
to

















-,4
o
- oJ


C


(CJ -
SIVAuMINI --!o U1imanfN


co



wL

-j
co



5-
_.
tu




































































< O
\ Cg

x x
o 0


0 6 0
Uo- 0


S71VAn.LN -O N jI f3lN


ow




U)
I


co


0)

N



I N




-L
ItL

! j

.._


(0
u2
Qa

o











14
C.
0










0
-A
02

















0 1-
0r)




























o
4-'
0
e





4
0O
-z;



-'-l






'o
0











01









pulses represent about one-third of the total activity for sources B

and C, but less than one-sixth of the activity for source A, and thus

the average character is depressed in the A region. It is encouraging

that Baart, Lee, and Barrow (1966) have reported results highly similar

to Figures 4 through 7, although their study was based on a much less

extensive body of data.

A cross-comparison of Figures 4 through 7 suggests that there

may be a temporal variation in the relative amount of millisecond pulse

activity. For example, 1963, 18% of all of the 18-MHz activity

intervals contained type 3 pulses, but in 1964 the figure dropped to

8.5%. The planet-wide average probability of emission for 18-MHz

pulses of all types declined from 0.173 in 1963 to 0.125 in 1964.

It is also interesting to note the differences in Figures 5

and 7, which are plots of the same apparition, 1963, but at different

frequencies, 18 and 22 MHz respectively. The average probability of

receiving Jovian emission at 22 MHz is about 46% that of receiving it

at 18 MHz for this apparition (Register 1968). It is noted in comparing

the histograms that at 18 MHz a far greater percentage of the total

emission contains millisecond pulses than at 22 MHz; in fact, five and

one-half times more millisecond pulses were received on 18 MHz than at

22 MHz for this apparition.


Io Influence


Since it is now well established that the position of the inner-

most Galilean satellite, lo, strongly influences the probability of

Jovian decametric emission (Bigg 1964), an effort was made to see if

this effect extends to pulse character. Figures 8 and 9 show profiles




































Figure 8. Relative number of intervals of 18 MHz type 2
pulses as a function of longitude for various
positions of Io.







FLA--CH-ILE 19 6!- 64 13 MHZ POSiTION OF: 10
120

80- -.

40-

-"TO EART











S40 10
2





S40- f
20- I

S80 -
S40- C A2



80
Go




c 80 -- ,o






0 90 iO00 2170 3600
LONGIUDE, SYS c EM 0
0 -




LOcTUa YSE




































Figure 9. Relative number of intervals of 18 MHz millisecond
pulses as a function of longitude for various
positions of Io.








POSITION OF 10
FLA-CHILE 1961-64 18 -MHZ
FL- C,_ I-_ 45


80 T TO EARTH


i3 60 I
"2 45"
5J A El1
40 Bi
r,0
Ao
20
z 20-





< 0 o "

0 60-
0- c
1 270
lo



H C


IL 20 Q2Y- o I 315
0, T--




0


00 900 1800 2700 3600
LONGITUDE, SYSTEM HI








of 18-MHz activity versus the System III longitude of the central

meridian for the normal type 2 and for millisecond pulses, respectively;

in each case the profile is drawn for eight different positions of Io

relative to the earth-Jupiter line. The points for these smoothed

histograms were obtained from a merge of the Florida and Chile data

from 1961 to 1964. Figure 8 suggests that the type-2 activity of

source A was relatively insensitive to the position of Io; while it

shows the well-known enhancement when Io was roughly 900 or 2400 from

superior geocentric conjunction, a significant amount of activity was

always present that was apparently unrelated to Io. On the other hand,

Figure 9 implies that the millisecond activity from source A was

somewhat more sensitive to the location of the satellite, peaking when

Io was near western elongation; most striking of all is the sharp

peaking of millisecond activity in source B when Io was near 900, and

in source C when the satellite was near 2400. Another method of dis-

playing the marked effect Io has on the millisecond pulses versus the

normal ones is shown in Figures 10 and 11. These are contour plots

similar to those used by Bigg (1964) and Lebo et al. (1965) in their

investigations of the Io effect. To generate these plots, a matrix

was set up with the seventy-two 50 intervals of Io from superior

conjunction as the ordinate, and the seventy-two 50 intervals of the

longitude of the central meridian as the abscissa, and the computer

then summed the number of intervals of normal and millisecond activity

for each pair of intervals. The relative numbers are indicated by the

different shadings in Figures 10 and 11. By comparing the two figures

it is once again evident that the Io influence is far greater for the

millisecond pulses than for the normal ones.

































3-W
ca
0










14-1
O mH




co
4J 4J
















'H 0
U F








0 '





U1

co 0













'-1 *O4
0














.00
on















'H

4- 1
C)
4-1







00
r0c















i-H
0. 0


1oJ

3a,

0 4
4Jl
C '




















In -
N -

*


!2 0 0P
N n
NOID3NflINO t~OWt~dfS I(0dOJ 01






































*H
0










41
C )


um
4-1

Ca



r04








0
l4-






C






*H
0

UC





'4
r-4i

0











mo
0 4J0










Pl
O C'



4)0
-4-




3' 0
(-4










0,z















-r.































































uO U 3
2 C~j C1

Nouo3NnPN03 0183dns wod~J oi








This sensitivity of sources B and C to the position of Io led

to the study shown in Figure 12, an unexpected result of which was a

rather clear-cut confirmation of the division of source B into two

subsources, B1 and B2, an effect only suggested by earlier work (Smith

et al. 1962). This analysis utilized only those intervals containing

millisecond pulses at 18 MHz for the 1963 apparition. The upper curve

is based on radiation received while source B1 was on the central

meridian, the second curve on intervals when B2 was on the central

meridian, and so on. The abscissa is the System III longitude of

the sub-Io point on Jupiter. It is evident that BI, B2, and C, which

were sensitive to Io's position with respect to the earth-Jupiter line,

are equally sensitive to the satellite's System III longitude. On

the other hand, A, which was insensitive to lo's position with respect

to the earth-Jupiter line, is also relatively insensitive to the

satellite's movement in System III. Notice that distinctly different

System III longitudes of Io stimulate B1 as compared with B2. It is

interesting that Figure 12, based entirely on millisecond pulses,

agrees with Dulk's (1965) observation that normal Jovian emission is

stimulated only when Io is over the hemisphere 1100 < X < 2900.

The angular relationships are more striking in the polar diagrams

of Figures 13 and 14, derived from the same observations as Figure 12.

Here the azimuthal co-ordinate is System III longitude, while the

radial co-ordinate is the number of intervals of millisecond pulse

activity received when each longitude was on the central meridian.

Figure 13 includes only those intervals when the System III longitude

of Io lay between the left-hand pair of dashed lines in Figure 12,

that is, along the arc shown by the arrow in Figure 13. Similarly,































Q0

0
Co




c o






CO


-"4
0





o H
w 0o


-H H
w a)
c u














>0
001)
Co
Z0 M


4- 0 a
1-
o 4-J -14

a)
C o o




(1 C otl

U- 0













t-H




-H











u o 0 0
L t) Lr)
LO Co aC))


L<

0



/ 0o
LO




00
r





^ .-- -- .--- -------- -- -------- -.Y------------^ ^ L,.1
~N ......--2-- -- -- -'s ------- -- /-- ------------\ r



S-_------------ -------- --- -" to


LL O







O
oI oo- U)




0,








(K ,NV ''"IVH .))S3S7Fld 311"is1i Ti IO,'llONI S VA! it 'J1J I :1 0'tN
.\ -2 S-


O OO O0 00 0
\I I --'- \I0
LI 1 nI L


f _____ ,
6 Q ~ O "0 o ~ ~0 0 0 0 0
^ ~ ~ ~ ~ ~ I (M~ci\ ; N ^

( NV -?U HOS3 -id33 I'"1''JON~ -IN!S!V L!JI.-0 O













FLA-CHILE 1963
18 M/SEC


180
180


NORTH POLE



r--. 10


2700


Figure 13. Polar plot of 18 MHz millisecond pulse activity for the
indicated range of Io positions centered on 210.












1800


FLA-CHILE 1963
18 M/SEC


900


NORTH POLE


2700


Figure 14. Polar plot of 18 MHz millisecond pulse activity for the
indicated range of o1 positions centered on 2700.








Figure 14 includes only data recorded when Io was between the right-

hand pair of dashed lines in Figure 12. Figure 13 shows that when B1

is active on the central meridian (that is, directed toward the earth),

Io is near System III longitude 2100, and also roughly 90 from superior

conjunction. Figure 14 shows that millisecond emission from B2 also is

received when Io is about 900 from superior conjunction, but in this

case the System III longitude of the satellite must be near 270.

Because the System III longitude of Io advances 27.50/hour,

2.2 hours are required for Io to move from 2100 to 2700, but during

this interval the longitude of the central meridian increases 800.

Suppose B1 is on the central meridian, with lo at System III longitude

2100. By the time Io has moved into position to stimulate B2, that

source is already well past the central meridian. Although the peaks

have finite widths, Figures 13 and 14 indicate that the overlap is

so limited that it is improbable that both B1 and B2 will be detected

during the same Jovian rotation. This hypothesis was checked by an

inspection of the data outputs for the entire apparitions of 1963

and 1964, which showed that on no occasion were millisecond pulses

received from both BI and B2 during the same Jovian rotation. The

same was generally true of character 2 pulses, although on several

occasions during the two years this type of pulse was received from

B1 and B2 during the same rotation; it is possible that these exceptions

represent what has been termed "non-lo-related" emission. The fact

that B1 and B2 are stimulated by distinctly different lo positions

is perhaps the strongest evidence to date of the reality of the division

of B into subsources.

The correlations demonstrated in these studies can be used to





38



support the hypothesis that the immediate environment of the planet

itself must contribute in an important way to the burst structure.

This point of view received additional encouragement from Riihimaa's

(1966) report that the bursts from source B display a distinctive

spectral fine structure. However, the foregoing conclusion should not

be regarded as precluding interplanetary or ionospheric influences.

For example, the mechanism responsible for the generation and escape

of the radiation may produce among the several sources disparate

angular sizes or bandwidths. These factors, then, might well result

in the radiation from each source interacting differently with the

intervening medium.















CHAPTER IV

VARIATIONS OF ROTATION PERIOD BY SOURCE AND BY FREQUENCY

Introduction


A great deal of interest was aroused in 1962 when a change in

the positions of the Jovian radio sources was discovered (Douglas

and Smith 1963b; Smith et al. 1965a). Since most of the theories

of the decametric radiation linked the sources to the Jovian magnetic

field, which in turn was presumed to be associated with the planet's

core, this drift of the sources implied a possible change in the rotation

period of the core.

The data from the first few years after 1960 suggested that

the rotation period of the sources had suddenly changed from the

System III (1957.0) period to one 1.17 seconds longer (Smith et al.

1965a). This change in period is equivalent to a drift rate of 10.450

per year. Lebo (1964) first studied this drift phenomenon in detail

through the 1962 apparition. His work was extended to include the 1963

apparition and greatly expanded in scope by Hayward (1965). Upon

Hayward's departure from the Florida group, the study of the drift was

continued by the author to include the next three Jovian apparitions

(Olsson and Smith 1968). The subject of this chapter is the study of the

source drift through the 1967.0 apparition. More recent studies that

have been made by Register (1968), Donivan and Carr (1969), and Carr

et al. (1970) include one more year's data.









Method of Analysis


Histograms of the reception probability versus the System III

longitude of the central meridian were generated from all of the Florida

and Chile data at the frequencies of 15, 18, 22.2, and 27.6 MHz.

The Florida and Maipu, Chile, data were merged into a single histogram

for all years that data were available from both observatories. Only

these merged histograms appear in the appendix, as the single-station

histograms have been reproduced in several publicationsand dissertations

by the Florida group.

The method used to determine the source position has become

somewhat standardized by our group. A tracing paper overlay is held

in place on the histogram and one approximates the source as well as

possible with a roughly Gaussian curve. The longitude of the peak of

this sketched curve is taken as the location of the source. One can

easily estimate the error involved by slowly shifting the curve on the

tracing paper back and forth in longitude until the fit to the histogram

is obviously poor. At frequencies above 15 MHz one can usually locate

the sources to within 50 using this method. Whenever the source was

too poorly defined or too asymmetric to use the above method accurately,

the source center was defined as the centroidd" of probability; that

is, the reception probability for each 50 zone was summed over the entire

source and the longitude of the midpoint of this sum was determined.

The source positions are listed in Table 3.


Most of the early studies concerning the drift used only the


Frequency and Source Dependence of the Drift













TABLE 3

LOCATION OF THE SOURCES


Frequency Mean Epoch Station* Source B Source A Source C


15 MHz






18 MHz











22.2 MHz









27.6 MHz


1961.5
1962.7
1963.8
1964.9
1965.9
1967.0

1957.1
1958.1
1959.1
1960.2
1961.6
1962.7
1963.8
1964.8
1965.9
1967.0

1958.1
1960.3
1961.5
1962.8
1963.8
1964.8
1965.9
1967.0

1961.6
1962.7
1963.7
1964.7
1965.9
1967.0


142 50
120100
150100
140100
165100
166100

110 50


125 50
130 50
130 50
141 50
137 5
156 50
160 50

105100
114100
132 50
130 5
140 50
138 50
143 50
147 50


242 50
260100
25510
25010
26010
26010


217
235
215
230
235
249
260
268
265
260

220
220
227
243
255
257
247
240

215
240
233
245
245
232


317 50
325 50
325 50
315 50
328 50
330 50

306 50
309 50

323 50
312 50
320 50
327 5
334 50
325 5
332 50


315 50
299 50
312 50
320 50
320 50
315 50
319 50

298 50
335 50
327 50
297 50
318 50


* In the station column, F = Florida and M = Maipu








position of source A as determined from the data collected at frequencies

from 18 to 20 MHz, the reason being that A is the most prominent source

and the data at these frequencies are the most abundant. The question

of frequency and source dependence of the drift was investigated by

Hayward (1965). He found that the drift rate averaged over a three-

year period did vary with frequency in a fairly smooth fashion --

the higher the frequency, the lower the drift rate. This same trend

was still apparent with the inclusion of another year of data (Smith

et al. 1965b). Analysis of the next two years' data showed that some

of the sources had reversed their sense of drift and were then migrating

toward lower longitudes. A study of all of the data by Gulkis and

Carr (1966) revealed that source A may be undergoing a sinusoidal

drift in longitude with a period of about 12 years.

To investigate the year-by-year drift of the three sources A,

B, and C at the frequencies of 15, 18, 22.2, and 27.6 MHz, the longitude

of each source was plotted as a function of date for each frequency.

Figures 15 through 17 resulted. In each of the figures the 18 MHz

data are included as a comparison. Error bars were not shown for those

pairs of data points where the separation was less than the error. Lines

were drawn joining adjacent points of a given frequency, and the slopes,

which are the average yearly drift rates, were calculated and recorded

in Table 4.

Figure 15 shows that all three of the sources at 15 and 18 MHz

drifted together within the limits of the accuracy of the measurement

for all years, except for sources A and B in 1964.8. Figure 16 shows

that the three sources at 18 and 22.2 MHz drifted together until 1964.8,

when sources A and C began to separate, followed by source B a year later.






43
o
0












No
o-O

O







0
o o
0
N 0












\1
LU(
c o







N 4
C, 0

COx




0- 0
10
IV 0







ao'

^ _ _^ _ 4


























>1


N N


NO



0O


w

0

U()


\
1


N- tC)
0) to LOo
0) 0) (Tb)3
ZH1VJA A
1J i, ~^r" i~ '"" o
yv:9A


N


o0
0 CM

'V
ii
00



0
NO N
N a




00
a -


0









O
:1
44
*H
o a)

0 0











0<
<;












-4A


LJ / \0





000

0 (1)



C) 00
a 0
n0-




o 4A


o o 0






\\ a
o 3

0 0



O a







00 ( o
0) w

U) %
c('
CD N





o, t~ IV A
















TABLE 4

YEARLY AVERAGE DRIFT RATE OF SOURCES


Frequency Mean Station Source B Source A Source C
Epoch Drift Rate Drift Rate Drift Rate

15 MHz 1962.7 F -18.3 15.0 6.7
1963.8 F 27.3 -4.6 .0
1964.9 F -9.1 -4.6 -9.1
1965.9 F 25.0 +10.0 +13.0
1967.0 F 0.9 .0 1.8

18 MHz 1958.1 F -- 18.0 3.0
1959.1 F -- -20.0 --
1960.2 FM 4.8 13.6 6.7
1961.6 FM 4.6 3.6 -7.9
1962.7 FM .0 12.7 7.3
1963.8 FM 10.0 10.0 6.4
1964.8 FM -4.0 8.0 7.0
1965.9 FM 17.3 -2.7 -8.2
1967.0 FM 3.6 -4.6 6.4

22.2 MHz 1960.3 FM 4.1 .0 --
1961.5 FM -16.7 5.8 -13.3
1962.8 F -1.5 12.3 10.0
1963.8 FM 10.0 12.0 8.0
1964.8 FM -2.0 2.0 .0
1965.9 FM 4.5 -9.1 -4.5
1967.0 FM 3.6 -6.4 3.6

27.6 MHz 1962.7 F 9.1 22.7 33.6
1963.7 FM 7.3 -6.4 -7.3
1964.8 F 6.4 10.9 -27.3
1965.9 F .0 .0 19.1
1967.0 F 7.3 -11.8 --








The most striking feature in Figure 17 is the wide separation between

source A at 18 and 27.6 MHz for all apparitions except 1962.7. In

these figures there are 57 pairs of points. For 33 of them the sepa-

ration is less than the error bars, and for the remaining 24 the

separation is greater. It should be noted that the separations were

a minimum in all of the figures for the three-year period from 1962.7

to 1964.8, when the amount of data was far greater than for apparitions

before or after. This suggests that the deviations observed for other

years may be the result of poorer statistics.

Figures 18 through 22 were drawn to illustrate the differences

in the drift rates from source to source at a fixed frequency and also

the differences in the drift rates for a given source at different

frequencies. Figures 18 and 19 are plots of the drift rate of the

three sources as a function of frequency for the apparitions of 1964.8

and 1965.9, respectively. Note in these two figures that a point

below the zero-axis represents a negative drift (that is, the source

moved to a lower longitude during the year), and a point above the

zero-axis represents a positive drift. These two figures show that

the drift rate is in the same sense for all three sources only at

15 MHz. For all of the other frequencies the sense of drift is different

for at least two of the sources.

Figures 20 through 22 are plots of the drift rate of the sources

as a function of time for the four different frequencies. These figures

show that there is a wide variation in the magnitude and direction of

the drift rate of the sources for the different frequencies over the

period of time indicated. There seems to be little, if any, correlation

between the frequency and the drift rate of a particular source for a








SOURCE B


5






- 10~


-10



-20


I1 20 25
FREQUENCY (MHz)
Figure 18. Source drift rate versus frequency for 1964.9.


- --------











20-





10-





0


10-





0-


IO-


F--10
LL
K |


15 20
FREQUENCY
Figure 19. Source drift rate versus


SOURCE B


SOURCE A


(MHz)
frequency for 1963.9.


~mrda~pna~i~~~ ---- -~~~--r~~-N-L--~








27.6 MHz


10-

0 --
-5 -


18 MHz

o---o., _/\


15 MHz


59 61 63 65 67
YEAR
Figure 20. Drift rate of source B as a function of time.


_ ~ltlB~I_


~_P


~







20-


27.6 MHz


22.2 MHz


18 MHz


S01-


15 MHz


'II ---------\,--- "-
___ __- 1 9 YEA
1959 1961 1963 1965 1967
YEAR

Figure 21. Drift rate of source A as a function of time.


V pl------------- ----------1----~ ------s~ ------------


r\


~













27.6 MHz


100 2_2- 2.2 Hz







L..

















60 62 64 66
10

o K

[-
10- 18 H z


















Figure 22. Drift rate of source C as a function of time.
-10- 15 z




-=-^ /\





Figure 22. Drift rate of source C as a function of time.









given year. The data in these figures suggest that the distance between

the sources would also vary year by year. They also suggest a yearly

variation in the distance between the positions of any given source

measured at two different frequencies. To check for this variation,

the difference in position of source A as determined at 18 and 15 MHz

was calculated for each year. These differences were also calculated

for source A using the frequencies 18 and 22.2 MHz and 18 and 27.6 MHz.

Similar differences were calculated for sources B and C. These results

are given Table 5. All of the differences are absolute values.

An average difference was calculated for each column in the table and the

error was computed. The error was assumed to be equal to the square

root of the sum of the squares of the error of each source position.

Differences between two adjacent sources at a fixed frequency were

also calculated for each year and are listed in Table 6.

Of the 57 longitude differences recorded in Table 5, only 5

(or about 9%) fall outside of the computed error. This suggests

that the year-by-year source location is independent of frequency.

It should be understood that this result is not in conflict with an

earlier study by the Florida group of a frequency-dependent drift of

the sources (Smith et al. 1965a). In the earlier study, the drift

was the change in the locations of the sources with a change in frequency

for a given apparition, as shown in Figure 23, which is copied from

the paper by Smith et al. The present work investigated the differences

in source position, such as the interval indicated in Figure 23 by

the line element labeled d, and noted how these differences change

from one apparition to the next.

In Table 6, 21 of the 54 differences (or about 39%) fall outside














TABLE 5

LONGITUDE DIFFERENCES IN SOURCE LOCATION
AT DIFFERENT FREQUENCIES


Mean Epoch

1958.1
1960.3
1961.5
1962.8
1963.8
1964.8
1965.9
1967.0






1960.3
1961.5
1962.8
1963.8
1964.8
1965.9
1967.0






1960.3
1961.5
1962.8
1963.8
1964.8
1965.9
1967.0


A18-A22*

15
100
80
6
5
110
180
200
Av.= 1270


B18-B22

110
2
0
10
10
130
130
Av.= 670


C18-C22

80
130
8
7
140
100
13.
Av.= 1070


A15-A18



7
110
5
18
5
00
Av.= 8140


B15-B18


120
100
90
30
9
6
Av.= 8140


C15-C18


50
5
2
190**
3
20
Av.= 670


A18-A27



200
90**
270
230
200
280
Av.= 2170


B27-B18


150
5
80
3
160
120
Av.= 1070


C27-C18


140
150
00
37**
7 **

Av.= 1570


* The notation A18 represents the location of source A measured
at 18 MHz, etc.
** These values exceed the average by an amount greater than
the error.












TABLE 6

DIFFERENCES IN LONGITUDES OF THE SOURCES
AT FIXED FREQUENCY


Frequency

15 MHz








18 MHz











22.2 MHz










27.6 MHz


*These values exceed the average
the error.


by an amount greater than


Mean Epoch

1961.5
1962.7
1963.8
1964.9
1965.9
1967.0


1957.1
1958.1
1960.2
1961.6
1962.7
1963.8
1964.8
1965.9
1967.0


1958.1
1960.3
1961.5
1962.8
1963.8
1964.8
1965.9
1967.0


1961.6
1962.7
1963.7
1964.7
1965.9
1967.0


IC-Al

750
650
700
650
680
700
Av.= 6911

89*
740
93*
770
710
670
66*
600*
720
Av.= 7470


95*
720
690
65*
63*
680
790
Av.= 7370

83
95*
94*
52*
730

Av.= 7970


IA-BI

1000
1400*
1050
1100
95
94
Av.= 107140



1050*
105*
1190
1190
1310*
1090
100*
Av.= 11370

115*
1060
950*
1130
115*
1190*
1040
93*
Av.= 10770

1000
1150
1000
1050
1050
84*
Av.= 10170




































Figure 23. Positions of sources A, B, and C at various
frequencies in 1962, with lines tracing a
possible mode of "drift" of the sources.













































00 1200 2400
LONGITUDE, SYSTEM Il


1200








of the calculated error, indicating a possible variation in the distances

between the sources. This variation could, however, be due to the

difficulty one has in accurately locating sources B and C for years

of low activity.














CHAPTER V

A STUDY OF THE ROTATION PERIOD BASED
ON STORM COMMENCEMENT TIMES

Introduction


Several hypotheses have been suggested to explain the apparent

sinusoidal drift of source A. Gulkis and Carr (1966) assume that the

radiation from the source is beamed into a thin, curved sheet which

is fixed with respect to the Jovian magnetic field. Since Jupiter's

magnetic axis is known to be tilted some 100 with respect to its

rotational axis, the sheet of radiation will wobble as viewed from the

earth as the planet rotates. Thus one can receive radiation only for

certain orientations of the earth and Jupiter's magnetic field. This

hypothesis therefore explains the variation in source position in

terms of the rate of change of the Jovicentric declination of the

earth. The hypothesis has been strengthened by the recent work of

Donivan and Carr (1969) and Carr et al. (1970), who have included

several more years of data in the study, which continues to show a

strong correlation between source position and the Jovicentric decli-

nation of the earth.

Another possible explanation of the drift of source A was

recently proposed by Register and Smith (1969). For several years

it has been known that source A seems to be composed of an lo-related

component and a non-Io-related component (Olsson and Smith 1966;

Shever 1967). If one separates the radiation from the source into









these two components and plots yearly histograms from these data,

several differences are noted between the curves. The centers of

source A of the lo-related curves consistently occur at a lower

longitude than do the non-lo-related centers. Secondly, the source

centers display completely different drift characteristics. The drift

of the center of the source in the lo-related curves is almost linear,

whereas the non-Io-related curves strongly suggest a sinusoidal drift.

These histograms also show that while the amplitude of the lo-related

source is fairly constant year after year, the non-Io-related source

varies considerably in amplitude. Therefore, it is possible that

the drift of the blended source is the result of superposing a non-Io-

related component of variable amplitude and variable period on a more

stable lo-related component.


Duncan's Hypothesis


A third explanation has been offered by Duncan (1967). Approaching

the question from a different point of view, Duncan defined the Jovian

radio rotation period in terms of the commencement times of the storms

from source A. Duncan's analysis utilized the observations of other

workers, which contained just over 1000 storms from all Jovian sources

at frequencies ranging from 18 MHz to 30 MHz. He plotted histograms

of storm commencement times and showed that the leading edge or front

of source A seemed to remain stationary in a coordinate system rotating

with a period of 9h55m29s70. When he studied the center of the source

in these histograms, Duncan found that some source drift remained,

although he regarded it as considerably smaller than that reported by

others. He concluded from these observations that the source drift








was in fact due entirely to a long-term secular variation in the

average durations of the Jovian noise storms,an effect that would

shift the center of the apparent source but not its leading edge.


Analysis of Florida and Chile Observations


Since the extensive data of the University of Florida and

University of Chile radio observatories were not available to Duncan,

it seemed worth while to examine them as a test of his very interesting

hypothesis (Olsson et al. 1969a, 1969b). The study was confined to

the 18 and 22.2 MHz observations, which are consistently the most

numerous and most reliable. The University of Florida observations

utilized in this investigation extend over ten apparitions, while those

from Chile cover seven apparitions. The daily records from each

station were examined and the System III (1957.0) central meridian

longitude was recorded for the time at which each noise storm began.

Groups of bursts separated by intervals of quiet of less than 15

minutes (about 90 of Jovian rotation) were regarded as belonging

to the same storm. Altogether, over 1800 noise storms from source A

were included in the analysis. A histogram was then prepared for

each apparition, showing the number of storm commencements that had

been observed when each 5 zone of System III longitude was crossing

the central meridian. Since it is known that the source structure in

the ordinary probability histograms changes with frequency (Smith

et al. 1965a),separate histograms were plotted for 18 MHz and for

22.2 MHz. To provide a cross-check, individual histograms were also

prepared for each of the two observatories.

The 33 histograms that resulted from this analysis are shown in








Figures 24 through 28. It should be noted in several of the histograms

that the source A peak is bifurcated; that is, there is a second

grouping of commencements at the higher longitudes. Those histograms

which display this bifurcation most prominently are the 18 MHz Florida

ones for 1963.8 to 1965.9, the 22 MHz Florida ones for 1964.8 to 1967.1,

and the 22 MHz Chile histogram for 1965.9. This same bifurcation is

evident in Duncan's second histogram (Duncan 1967, Figure 6), which

is a merge of data spanning the years 1962-1966, although it is absent

in his earlier histogram for the period 1951-1961. Figure 29 is a

merge of the 18 MHz Florida data for the periods 1958-1961 and 1962-

1966. These data produce a distribution strikingly similar to Duncan's

histograms. In both cases the steepest part of the commencement

front is preceded by a more gradual rise covering about 250 in longitude.

The bifurcated peaks span approximately 600 and are followed by

tails extending to longitudes near 3000.

Figures 30 through 33 are the histograms produced when the

Florida and Chile data were merged. For each curve the "front" was

selected as the first significant rise following the well-defined

minimum which characteristically separates sources A and B. It is

evident from the method of analysis that any movement of the front

must occur in multiples of 5. Furthermore, in several instances

there was more than one feature that might have been selected as the

front; where these ambiguities exist two choices were made. The

fronts so chosen are indicated on the histograms of the merged data

with vertical dashed lines.

Figure 34 traces the position of the commencement front of source

A as it was determined from the merged Florida-Chile histograms.














FLORIDA 18.0 MHz SOURCE A


5 1968.1 37 STORMS

0

o 10
1967.0 93 STORMS
U,
'L 5-
o _


m 142 STORMS
S 1965.9
Z 10-




1500 200 250 300 350
LONGITUDE OF CENTRAL MERIDIAN


Figure 24. Commencement time of storms histograms of
source A for 1965.9 to 1968.1 at 18 MHz.























































1500 2000 2500 3000 3500
LONGITUDE OF CENTRAL MERIDIAN
Figure 25. Commencement time of storms histograms of
source A for 1958.1 to 1964.8 at 18 MHz.










MAIPU 18.0 MHz SOURCE A

5 1967.0 10 STORMS-

0 -_o__uJl_____-
1965.9 35 STORMS



5 1964,8 32 STORMS!

0 C--l --0
10
0 1963.8 88 STORMS
Cc 5-


o
v-I


w1O

S 1962.7 121 STORMS
S5

0 --

1961.2 79 STORMS


0 JED I------- -----

1960.2 42 STORMS
5-

0 _.^=L ZLZ.
1500 2000 2500 3000 3500
LONGITUDE OF CENTRAL MERIDIAN

Figure 26. Commencement time of storms histograms of
source A for 1960.2 to 1967.0 at 18 MHz.







FLORIDA 22.2 MHz SOURCE A
1968.1 r J 19 STORMS

1967.0 41 STORMS

0

10 -
1965.9 r 65 STORMS
5-



1964.8 78 STORMS
u) 5 _[

0- 0- -[f

_ 1963.8 I01 STORMS
05-

ca 0

z 5 1962.7 51 STORMS -



51961.6 34 STORMS


5
S1960.2 I I STORMS
o --. --C
3 1959.1 7 STORMS

3 1958.1 8 STORMS
0 r- -J .. ,L ._ __.n n

1500 2000 2500 3000 3500
LONGITUDE OF CENTRAL MERIDIAN

Figure 27. Commencement time of storms histograms of
source A for 1958.1 to 1968.1 at 22.2 MHz.










MAIPU 22.2 MHz SOURCE A
y f -


1967.0
r,Q-Lh r, L,


19 STORMS


Id-


w


z


1500


3500


Figure 28. Commencement time of storms histograms of
source A for 1960.2 to 1967.0 at 22.2 MHz.


0-
1965.9 31 STORMS
5



5 -
1964.8 26 STORMS

0
O -
1963.0 75 STORMS
5


0
0-

1961.6 49 STORMS
5 -


0- -3 El -

1960.2 27 STORMS
5-


2000 2500 300
LONGITUDE OF CENTRAL MERIDIAN


(




























VI)
0
Dr
S40
u)

0
ir

m 30


150, 2000 2500 3000 350
LONGITUDE OF CENTRAL MERIDIAN



Figure 29. Merged histogransof commencement time of storms.











FLA-CHILE MERGE 18.0 MHz SOURCE A


U) ",

0 0
OO
15
0

S10
m

2 5


350E


Figure 30. Histograms showing commencement fronts with
dashed lines for 1964.8 to 1968.1 at 18 MHz.


LONGITUDE OF CENTRAL MERIDIAN







FLA.-CHILE MERGE 16.0 MHz SOURCE A


1- 1963.8 237 STORM

IO-













0
S15-




I
r-



co 1961.6 I 178 STORM

z
710-


3 (FLA. ONLY)
1958.1 II STORMS"
0 L-.-IJ-.,- JIL_.Lt. -_ _. _
1500 200 2500 3000 3500
LONGITUDE OF CENTRAL MERIDIAN

Figure 31. Histograms showing commencement fronts with
dashed lines for 1958.1 to 1963.8 at 18 MHz.















































LONGITUDE OF CENTRAL MERIDIAN


3500


Figure 32. Histograms showing commencement fronts with
dashed lines for 1964.8 to 1968.1 at 22.2 MHz.











FLA.-CHILE MERGE 22.2 IMHz SOURCE A


5-

1963.8 176 STORMS
0- 1


5-


0 6
O -- -1----
1962.7
1 51 STORMS
5- I



1961.6 88 STORMS
0-


5- ... 1

o i10IfL __


1960.2
I I -
-%


38 STORMS


1959 1 (FLA. ONLY)

1958.1 (FLA. ONLY)
Fr FL 8 STORMS
I--) n it IJ__J


150


2000 2500 3000
LONGITUDE OF CENTRAL MERIDIAN


3


Figure 33. Histograms showing commencement fronts with
dashed lines for 1958.1 to 1963.8 at 22.2 MHz.


en
or.
I--
t)

o 1

03


Z
Ct
wJ
cc


500


r - --- - - -


---


r.


I#


,.~,~ I*~~~^--cl-F-~-------------













1970-- 18 MHz 22.2 MHz





1968 0


0- -0


1966 3


O- 0

1964-
S0-- 0
LJ
I- -


1962
0




1960




S1958

1600 2000 2400 160 2000 240
LONGITUDE OF CENTRAL MERIDIAN


Figure 34. Location of the commencement front of source A
during ten apparitions of Jupiter.









Where two fronts were chosen for one year, both are recorded and joined

by a horizontal tie. The straightforward interpretation of Figure 34

is that it supports Duncan's contention; over an interval of a decade

the commencement front has remained substantially fixed in longitude,

which is to say that its rotation period has been constant. It is

also interesting to note that the longitude of the front is the same

at 18 and 22.2 MHz, whereas there is generally a significant displacement

of the center of the source between these frequencies in the conventional

probability histograms.

The fact that each set of points in Figure 34 is well fitted

by a vertical straight line implies that the period defined by the

commencement fronts in this study is that of System III (1957.0),

i.e., 9h55m29s37 with a standard error of 0.08. This is about

O03 less than the period favored by recent decimetric observations,

as well as the mean period adopted by Gulkis and Carr (1966) in their

oscillatory theory of the behavior of the standard probability histograms.

In the present state of the art it is well to exercise caution in

assuming that each of these highly disparate approaches is measuring

precisely the same phenomenon.

Figure 35 shows the location of the center of source A in the

storm commencement histograms for the ten apparitions. The points

were obtained by determining the central meridian longitude corresponding

to the average storm commencement time for each apparition. It is

evident that, unlike the commencement front, the source center displays

an oscillatory drift of large amplitude; the solid curves are sinusoids

with 11.9-year periods which were least-squares fitted to the empirical

points. The dashed curve, adapted from Gulkis and Carr (1966), depicts






















0 4J



ca
mou





in o
o








CO


o a
S0





o o
UO




0 .0
-z












0 o
O eL
g o

00

0









o a
n 4j e








0 H41







0 C


0 o

00.0

0.0
o -i







0J CO


C (


0 3


*r C U













< -
<1 "/? 1 / \ -

)J /





o 7-





'i /0 U
CiI






OJ 0n



0
-0


C)
0J -



N ao
0
00
0 ro







0 0 o

3. LVG








the oscillation of the center of source A in the usual probability

histograms. It is evident that the drift amplitude of the source

center is substantially the same in both types of histograms.


Storm Durations


A final question is whether the oscillatory behavior in Figure 35

can be attributed to a long-term periodicity in the mean durations

of the noise storms. Experienced observers are aware that at any

frequency it is unusual for a storm simply to begin at the leading

edge of the source (point a in Figure 36) and persist right across

the source to b. Far more commonly, storms begin "within" the source

at some point such as c and end at another point d well before the

trailing edge of the source reaches the central meridian. In other

words, the "source" is largely a superposition of numerous fairly

short storms with staggered commencements and endings (such staggering

of at least the commencements is of course implicit in the distribution

shown in Figures 24 through 28).

To investigate the distribution of storms within source A, plots

in which each storm is represented by a horizontal line were prepared

for each apparition from 1959.1 to 1967.0 from the 18 MHz and 22.2 MHz

Florida data. These lines are plotted against the longitude of the

central meridian, and thus indicate where the storm began and how long

it lasted. The results are shown in Figures 37 through 46. In each

case there was a wide distribution of storm durations and of starting

and stopping times. To study this result further, a distribution curve

was plotted as a function of the duration of the storms and this is

shown in Figure 47. In this figure the duration is measured in 5 zones




















































N
(D
0)

X X
w



-Jc
I

LL


0


A i I TOV-Uojc









I I


' I


I
- I


I I


I I


-18 MHz FLA
---1961.6
99 STORMS










-- I












. f I I I


1500 2000 2500 3000 3500
LONGITUDE OF THE CENTRAL MERIDIAN


Figure 37. Distribution of storms within source A for
1959.1 to 1961.6 at 18 MHz.


18 MHz FLA
1959.1
18 STORMS





18 MHz FLA
1960.2
23 STORMS























=1


I -


I



____________________I





______ I


_______ I



I

__________________________________ I


S I


2000


250


I I I


300


350


LONGITUDE OF THE CENTRAL MERIDIAN




Figure 38. Distribution of storms within source A for
1962.7 at 18 MHz.


18 MHz FLA
1962.7

138 STORMS


150


I~ar~urrpl~.*r~-~-ns rr-j-;l~nu~ur-~vl.-raP~Y3YC~D~-~-~-PI


------nl il-r-a-l---P---~----~--r~C-










r ~ I __ar~urnw-- 'Innnnnnnnn~~~~~~~~~


- _


= I





____ I


I --------------------------
~"--~-~--I-~- I`--~~-8-s~
-I---~--`---~--~
-~-~""1---~-~-~-~-31'--^"~--

1 "I~-"I-- --------------
--I~'-------~


I I
Si !


II I


2000 2500 3000 3500

LONGITUDE OF THE CENTRAL MERIDIAN


Figure 39. Distribution of storms within source A for
1963.8 at 18 MHz.


r- _


1500


ru~a~-jpiF~n~aarw~yrran~-rsrrp-n'uu~a ~ -b~ 'm


18 MHz FLA

1963.8

150 STORMS


___________ I


-I






_I






"-- ----mi I


I- --

-X- _Pl- I






I----




________________________________________
_ I~--~-~










I -


I ---


----p---- ~------r -----;


=_I
- I
-- I


-- I


__I


--- ----- ---
I


I
^"' .'-'""I
"^^^-^^^ ^ ^J,_


r_ i.IIJ 1_


2000 250 3000 350

LONGITUDE OF THE CENTRAL MERIDIAN


Figure 40. Distribution of storms within source A for
1964.8 at 18 MHz.


1500


~___~__1~ 11~1---- :M-Iy~-


' I












-- I

= ..I
-




-


1 i


















18 MHz FLA

1964.0

124 STORMS


7!


-i-


!




83


















-- 18 MHz FLA
1965.9

142 STORMS










S------


II











1500 2000 250 3000 350'
LONGITUDE OF THE CENTRAL MERIDIAN



Figure 41. Distribution of storms within source A for
1965.9 at 18 MHz.

















--_ I
= I





S 18 MHz FLA
1967.0
I 88 STORMS








I -I
II









1500 2000 250 3000 3500
LONGITUDE OF THE CENTRAL MERIDIAN







Figure 42. Distribution of storms within source A for
1967.0 at 18 MHz.










22 MHz FLA


i I
























I~


1959.1
6 STORMS


1960.2
10 STORMS




1961.5
31 STORMS


SI 1962.7
S51 STORMS







,. . ... -.I


200 250 300 350
LONGITUDE OF THE CENTRAL MERIDIAN


Figure 43. Distribution of storms within source A for
1959.1 to 1962.7 at 22.2 MHz.


1500


L


II

















I =-
I -=I






S- 22 MIIz FLA
"-- 1963.8
S-- 102 STORMS


















II



1500 2000 2500 3000 350
LONGITUDE OF THE CENTRAL MERIDIAN





Figure 44. Distribution of storms within source A for
1963.8 at 22.2 MHz.


















-- I I I

I





-- I
-- I
I











-. I


_______________________________________________________ I
-.- I
-- I

___________________________________________________________ I


I I -I









22 MHz FLA

1964.8

84 STORMS


I I


2000

LONGITUDE


2500

OF THE


,,,----J-L

3000

CENTRAL


3500

MERIDIAN


Figure 45. Distribution of storms within source A for
1964.8 at 22.2 MHz.


1500


~~~nc~-~o~xrxu p~vorirrui~m~ura-an


I ---~




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

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