Jupiter's decameter and kilometer emissions

MISSING IMAGE

Material Information

Title:
Jupiter's decameter and kilometer emissions satellite effects and long term periodicities
Physical Description:
xii, 176 leaves : ill. ; 28 cm.
Language:
English
Creator:
St. Cyr, Orville Chris, 1950-
Publication Date:

Subjects

Subjects / Keywords:
Radiation -- Jupiter (Planet)   ( lcsh )
Satellites -- Jupiter (Planet)   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1985.
Bibliography:
Includes bibliographical references (leaves 163-175).
Statement of Responsibility:
by Orville Chris St. Cyr.
General Note:
Typescript.
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 - 000869519
notis - AEG6556
oclc - 14399203
System ID:
AA00003819:00001

Full Text
















JUPITER'S DECAMETER AND KILOMETER EMISSIONS:
SATELLITE EFFECTS AND LONG TERM PERIODICITIES










BY

ORVILLE CHRISLST. CYR


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


UNIVERSITY OF FLORIDA


1985

























Unfortunately, my father passed away before he could delight in

the completion of my dissertation. It is to him and to my mother that

this work is dedicated.














ACKNOWLEDGMENTS

The work described in this manuscript has taken much longer to

prepare and present than originally intended. To a large extent, this

delay may be attributed to the author's premature departure from

graduate school--premature in that he left the academic world before

completing the work. One consequence has been that the work remained

incomplete for five years; one benefit has been that the author's life

has been touched by many more people and experiences than it would have

been otherwise. For this reason, the acknowledgments for support, both

moral and intellectual, will be quite lengthy.

Thomas D. Carr, the advisor for this work, has displayed patience

and support far beyond that needed by most graduate students. He has

been called upon to read and critique innumerable drafts submitted by

this author which, in many cases, were not as clear or well-organized

as they should have been. In all these instances he was able to

ascertain those problems which the student could not see, and he was

able to provide a framework so that the author could better organize

his thoughts. His questions were simple--the answers he sought were

fundamental but were never beyond the grasp of the student. His

tenacity for attempting to understand the sporadic low frequency

emissions from Jupiter is exemplary, and many times over he has earned

this author's respect and admiration.

While at the University of Florida, the author benefited from

classroom teaching and out-of-class interactions with the entire

faculty of the Department of Astronomy. In particular, several of the

faculty members have served on the doctoral committee, and their

contributions are appreciated. These members are Alex G. Smith,
iii







Stephen T. Gottesman, and George R. Lebo. For the preliminary oral exam-

ination, Charles F. Hooper of the Department of Physics served on the

committee, and the author is grateful for his participation.

The author has also benefited from outstanding interaction with

other mentors during the years in Florida. Most notably, Stanley S.

Ballard, Distinguished Service Professor of Physics, has taught this

author much about optics, teaching, and professionalism in science. He

also served on the doctoral committee for the final examination, and this

and his other contributions to the author's maturation are appreciated.

Other faculty mentors also deserve special recognition. F. B. Wood, J.

E. Merrill, and H. Eichhorn have all left positive impressions on this

author's life.

Much of the experience one gains during graduate school comes from

other graduate students. This author's life has been enriched by many

other students who endured these experiences, but the names of Andrew

Pica, Greg Fitzgibbons, Joe Pollock, Dan Caton, Francisco Reyes, Roger

Ball, Douglas Johnson, and Glenn Schneider are immediately thought of.

Upon leaving Florida, the author was employed as a Systems Analyst

for Prime Computer, Inc., in Manhattan, NY. At Prime, Kent Fielden (and

others) provided the author with a "virtual B.S. in Computer Science" by

his many excellent tutorials. Moreover, the author learned something of

the business world through positive interaction with Roy Berardelli,

Steven Pittas, Elven Riley, James McManus, Dan Brodsky, Bruce Yellin, and

Gordon Monsen.

In early 1984, the author accepted employment as a Support Scien-

tist with the High Altitude Observatory (HAO) in Boulder, Colorado. The

job was to act as Chief Observer for HAO's coronagraph onboard the

satellite Solar Maximum Mission. Training for this job was conducted
iv








by Bob Lee. His influence has been a positive force in this author's

life, as he offered guidance in solving engineering problems and in solv-

ing life's problems. Also at HAO, the tutelage of Bob MacQueen has con-

tributed positively to the author's character.

The Solar Max satellite operations have been conducted at Goddard

Space Flight Center in Maryland. Here again, many have contributed to

the author's ideas and actions. The support of the C/P operations team

has been necessary to complete this manuscript, and thanks go to each one

of them--Bob Lee, Kathleen Walsh, Stacey Rosenberg, David Kobe, and Al

Gross. Others at Goddard who have made the satellite work an enjoyable

and educational experience are Joe Gurman, Bruce Woodgate, Brian Dennis,

Keith Strong, Dave Speich, Dave Douds, Jim Harrison, and Walsh Barcus.

Also at Goddard, the author's understanding of planetary radio emissions

has been enhanced by many discussions in Building 2 with Michael Desch

and Michael Kaiser. Their contributions are cited throughout this work,

but they deserve special recognition for supporting the author's efforts

to complete this manuscript. Ron Parise at Goddard has also given

encouragement and provided enlightened discussion.

Many others have not been mentioned but they have nevertheless con-

tributed to the work presented here. Suzanne St. Cyr has produced much

of the professional artwork included here. Irma Smith of Gainesville has

provided expert typing, in spite of physical difficulties, to yet another

dissertation from the Department of Astronomy. Greg Fitzgibbons has

assisted in the author's submission of the dissertation to the Graduate

School in Florida, and the time he has spent is appreciated. Colin Bar-

row supplied the data from the University of Colorado radio spectrograph;

Jorge May has supplied data from Maipu, Chile.

v











W. Richardson assisted the author through the process of data reduction

from U.F.R.O., and his efforts are greatly appreciated. In 1977, plan-

etary scientist Alex Dessler suggested that the author investigate the

roles of the Jovian satellites in the production of DAM. At the Uni-

versity of Oklahoma, Tibor Herczeg and David Branch first taught the

author the fundamentals of astronomy and astrophysics, and it is pleas-

ing to be able to extend that foundation.

Finally, the only people left to acknowledge are the author's

family. His wife, Suzanne Beairsto St. Cyr, has never lost faith that

this work could be completed. Her support extends far beyond the

others mentioned here, and she very much represents the watershed for

his motivation. The imperturbable support of his stepson, Odin Austin

Shafer, and of J. Johns are also acknowledged.












TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS . . ... iii

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

LIST OF FIGURES . ... x

ABSTRACT. . . ... . xi

CHAPTERS

I. INTRODUCTION . . ... 1

Background . .... 1
Dissertation Overview. . ... 10

II. ELECTROMAGNETIC INDUCTION IN PLANET-SIZED BODIES 14

Introduction . ... 14
Induction by Time-varying Magnetic Fields. ... 15
Induction by Motion in a Magnetic Field. 17
Conductivity Considerations. . ... 28
Completion of the Electrical Circuit ...... 38

III. THE JOVIAN MAGNETOSPHERE AND SATELLITES. .. 43

The Inner Magnetosphere. . ... 46
The Middle Magnetosphere . ... 50
The Outer Magnetosphere. . ... 52
Io . . 55
The Icy Satellites . ... 64
Amalthea . . ... 71
Discussion . .... 74

IV. THE 11.9 YEAR PERIODICITIES. . ... 79

Observations . . ... .79
The Solar Influence. . ... 83
Probability versus D . 93
Polarization versus E . 102
Spacecraft Results versus Jovicentric Latitude 107
Discussion ............... .108










Page

V. SEARCH FOR SATELLITE EFFECTS--DAM (GROUNDBASED). .. 111

Introduction ..... ...... 111
The Satellites' Orbital Resonance. ... 112
Previous Studies . ... 121
Analysis and Results . .... 124
Discussion . .... 139

VI. SEARCH FOR SATELLITE EFFECTS--KOM (SPACECRAFT) .. 144

Introduction . ... .144
Analysis and Discussion. . ... 149

VII. SUMMARY AND CONCLUSIONS. . ... 158

BIBLIOGRAPHY. . . ... ..... 163

BIOGRAPHICAL SKETCH . .... ..... 176


viii












LIST OF TABLES


TABLES Page

I-1 DAM SOURCE GEOMETRIES. . . 6

II-1 UNIPOLAR CLASSIFICATION. . ... 30

III-1 THE GALILEAN SATELLITES. . ... 57

III-2 CHARACTERISTICS OF THE GALILEAN SATELLITES .. 58

III-3 GALILEAN SATELLITE ENVIRONMENT . ... 59

III-4 AMALTHEA . . ... .. .73

III-5 INDUCED ELECTRIC FIELD INTENSITIES FOR THE MAJOR
SATELLITES . .... 75

IV-1 OBSERVATIONS USED IN THIS STUDY. . 81

IV-2 DATA FROM FLORIDA 18.0 MHz POLARIMETER ... 106

V-1 Eur OCCURRENCES FOR FIXED Io . ... .120

V-2 SATELLITE BEAT PERIODS . .... 141

VI-1 FFT SPECTRAL PEAKS IN nKOM . ... 155

VI-2 BEAT PERIODS OF THE JOVIAN SATELLITES WITH nKOM 156


I













LIST OF FIGURES


Figures

1.1

1.2

2.1

2.2

2.3

2.4

2.5

4.1

4.2


4.3

4.4

4.5

5.1

5.2

5.3


5.4

5.5

5.6

6.1

6.2

6.3


Probability versus XIII for DAM .

DAM for ( III, 'Io) . .

Induction by motion . .

Characteristics of both TE and TM induction. .

Summary of the stagnation field characteristics.

Superposition of magnetic fields .

Alfven wave generation . .

Solar activity and DE . .

Observing time for Florida/Chile 18.0 MHz and
Colorado . . .

Probability versus opposition date .

Probability versus DE for each source .

18.0 MHz polarimeter axial ratio .

II Io' 'Eur' and 'Gan geometry .

(XIII, rEur) and (XIII, 'Gan) for 'Io fixed. ..

18.0 MHz Io-B-probability versus "Eur for 1962
and 1974 . . .

18.0 MHz non-lo-A probability versus 'Sat. *

22.2 MHz Io-A probability versus 'Sat .

15.0 MHz Io-B probability versus Sat .

Dynamic spectrum showing KOM . .

Power spectrum for bKOM . .

Power spectrum for nKOM . .


Page

4
. 420

20

. 23

. 27

. 33

. 42

. 85


88

91

95

105

115

118


128

131

133

135

147

151

153


. .












Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


JUPITER'S DECAMETER AND KILOMETER EMISSIONS:
SATELLITE EFFECTS AND LONG TERM PERIODICITIES

By

Orville Chris St. Cyr

December, 1985

Chairman: Thomas D. Carr
Major Department: Astronomy

We present results of observational studies of Jupiter's deca-

meter (DAM) and kilometer (KOM) radio emissions. Our initial goal was

to investigate the role of Jupiter's major satellites in modulating

these sporadic emissions. We review the concepts of electromagnetic

induction in planet-sized bodies and describe the Jovian magnetosphere

environment and the physical characteristics of the Galilean satellites

and Amalthea. Since the phase of Io is known to modulate the DAM emis-

sions, we address the problem of orbital resonance among the Galilean

satellites. When lo's influence in the emissions is eliminated from

more than two decades of groundbased observation, no significant

enhancements or deficiencies in the detection probability were found

for the phases of Europa, Ganymede, or Callisto. An upper limit

estimate of the fraction of radiation possibly controlled by these

satellites has been determined. An analysis of the Voyager 2 KOM data

to search for possible satellite modulation has also been completed.

Although periodicities other than that due to Jupiter's rotation were

present in the nKOM component, no evidence of satellite control

xi










present in the nKOM component, no evidence of satellite control

(including lo's) was found. A very low upper limit has been placed on

satellite control of the emissions in this portion of the Jovian

spectrum.

A secondary goal has been to search for long term periodicities

and other effects in the groundbased DAM data. As expected, we found a

strong tendency toward an increase or decrease in detection probability

with an increase or decrease, respectively, in DE. However, we found

that for DE values between +3?1 and +3?3 (the latter being the

upper limit of the DE range) there is a relatively small minimum where

the maximum of the probability versus DE curve was previously thought

to be. We found that the non-Io-related, left-hand circular component

of these largely polarized DAM bursts disappears near the maximum

positive DE. We cite spacecraft results in the published literature

which appear to confirm this discovery, and we discuss the relevance of

this new finding to models of the emission source location.














CHAPTER I
INTRODUCTION

...Our own eyes show us four stars which wander around
Jupiter as does the moon around the Earth, while all
together trace out a grand revolution about the sun in
the space of twelve years. (Galileo, 1610)

Background

With these words Galileo announced the discovery of the four

Jovian satellites which now collectively bear his name. The existence

of this miniature solar system was taken as damning evidence against

the geocentric universe, and it transplanted the center of the cosmos

to the Sun. In contemporary endeavors the Galilean satellites have

again provided startling discoveries of natural phenomena. In two fly-

by's of Jupiter and the retinue of satellites, the Voyager spacecraft

revealed active vulcanism on lo, a curious grooved terrain on Ganymede,

and extensive linear markings on the nearly craterless, icy surface of

Europa.

Much of the work presented in this dissertation is about Jupiter

and its major satellites. The giant planet and its first Galilean

satellite, Io, act together to produce sporadic decameter wavelength

noise storms. Discovered by Burke and Franklin in 1955, Jupiter's

decametric activity has been monitored by planetary radio astronomers

at the University of Florida every apparition since 1957. The result-

ing accumulation of thousands of hours' observation covering more than

two decades provides a unique database of this phenomenon. Soon after







2

the 1955 discovery of the decameter noise storms, several investigators

(as will be cited in this work) realized that the probability of

detecting Jovian activity was dependent upon the Central Meridian Long-

itude (designated CML or X) of Jupiter. Histograms of Jovian activity

as a function of CML appear similar from apparition to apparition when

a planetary rotation rate of 5h 55m 29.71s is invoked. This period is

approximately five minutes longer than that of visible features at the

giant planet's equator, but it is believed that this longer period

represents the rotation rate of the planetary core which generates Jup-

iter's magnetic field. The fundamental longitude system adopted by the

International Astronomical Union has an initial epoch of January 1.0,

1965, and is designated XIII (1965.0).

A representative histogram of the probability of detecting Jovian

decameter (or DAM) activity as a function of XIII is shown in Figure

1.1. Early investigators named the three maxima according to their

relative strengths at an intermediate decametric frequency--Sources A,

B, and C. While it is now recognized that A, B, and C may be due to

multiple beams originating from a common location, the convenience of

referring to them as separate "sources" has persisted through the

years. In 1964 the statistician Bigg reported that the orbital phase

of lo, the innermost Galilean satellite, had a profound effect on the

occurrence of the Jovian decameter storms. Using observations from the

University of Colorado, Bigg showed that two orbital positions of Io

were favored. Satellite orbital positions or phases (denoted YSat) are

measured from Superior Geocentric Conjunction--where Earth, Jupiter,

and the satellite lie along a line, with Jupiter between the Earth and

the satellite. Bigg (1964) found a very large enhancement in the


























S)

4-)
L
a)

a)N


ea-r V-A
4J0 to
L. 4- 0
Cr 01





.0 4-)

4- 'O r- ~
1-) 0 0
4-) E LL-


0. 0




-o S- "a-
o d) U. M c
c O~00

E E












> 0 Ca



-0-0 2 4~)

S- *'=- 0
S. U 4-) CU
S-4- > ,- .
4-. 40-














M 4J 0.-
"-i~n a) 0 a






















C3 S-j 0
UQIOU
< E 0 I -
















>~ 0 0.

1. = 0~. U
V) CL 41
p-C L ~





o.-0) I 4.) U)
*'-.4) a)
1-4 >Q 1-



- ~ E '- ) +
*r U)0~

i--J. a)


aL0C)
1-4) .0 1-



0 C, I-.


-:L) 0..






























CIO


CD






cc f.0 CI


ca~
=" _
























P-4















0CD
= z
=
00o






u=

a1c,







5

detection probability for Io phase angles of 90 and 240. The single

parameter view of favored Jovian longitudes was thus expanded to a two

dimensional picture of dynamic interaction between planet and satel-

lite. In an extension of the earlier nomenclature, the regions of

(XIII, flo) configuration space which showed dependence on both para-

meters were named Source lo-A, Source lo-B, and Source lo-C. Those

regions showing no Io dependence became non-lo-A, non-Io-B, and non-

lo-C (Table I-1). Evident in Figure 1.2 is the fact that a large

fraction of the radiation is not dependent upon lo's phase. This fact

has contributed to the idea that influence by one or more other Jovian

satellites may exist. Early investigations, however, were limited in

the conclusions that could be drawn concerning influence by the other

satellites (e.g. Bigg, 1966) due to their orbital resonance.

More recently, the Jovian spectrum has been extended into the

kilometric (KOM) wavelength ranges by instruments onboard the Voyager

spacecraft (Warwick et al., 1979). Although these emissions are spor-

adic and composed of two distinct components (narrow- and broad-band-

width), there is at present no evidence for Io modulation of them.

With a periodicity of 10t17, the narrow band kilometric component

appears to be the only Jovian electromagnetic phenomenon which does not

rotate with the XIII (1965.0) rate.

With this background, we may now state the primary problem

addressed in this dissertation: What are the roles of Jupiter's major

satellites in modulating the planet's electromagnetic emissions? Does

any satellite other than Io affect the sporadic decameter emissions?

If so, why has it eluded detection for over 30 years? If not, what

quantitative upper limit can be set as determined by the observations?

Further, do any of the Jovian satellites affect the newly discovered











TABLE I-1
DAM SOURCE GEOMETRIES


Source XI

(degrees) (degrees)


Null 5 95 All

(lo-B) 70 110
B 95 195
(non-lo-B) All others


(lo-A) 220 260
A 195 285
(non-lo-A) All others


(lo-C) 220 260
C 285 365
(non-Io-C) All others


NOTE: The Jovian longitude and Io phase definitions
for the probability features seen in groundbased
observations of Jupiter's DAM (cf. Carr and Desch,
1976). All work in this study has been based on the
System III (1965.0) rotation rate for Jupiter.



















O

- CO


0
o

ci-



0

OE



0






s,- "-
L


ei- .
OCI
o <,





0 1
L-
Os-
0 w
C *

*- (


0 U0


C,,


0 1


err












00


* *

o
U



0




4-)
0 0
C
4-) L.







- 0

11
L









t i
LO




4- -




















L
C *'-
0r





03
*r- V
+> L
0 0.
>>
L



OU,

0 '4.


N
E -




(0 r

N" (0 A
s Io '





























00
in




r-
<*t


0


0

<~

It


CI

cc)



N
m
mp
E


VM









kilometric emissions? If not, what upper limit on satellite inter-

action can be determined using these observations? Most importantly,

from our knowledge of the Jovian moons and their environment, do we

expect that satellites other than Io should exhibit the strong electro-

magnetic interaction which could be detected through the low frequency

radio emissions?

A secondary goal of our work has been to investigate the ground-

based decameter observations for long term periodicities and other

effects. An understanding of the long term effects in the data has

been necessary to disentangle the orbital resonance of the Galilean

satellites.

We focus this study on the Galilean satellites and Amalthea. A

more complete statement of the complexities involved in the resonance

of the Galilean satellites is presented in Chapter V. The problems

associated with groundbased observations of Amalthea are described in

Chapter III. Our motivation in undertaking this investigation is based

on interest in the nature of the electromagnetic interaction of satel-

lites with planetary magnetospheres. Goldreich and Lynden-Bell (1969)

modeled the lo-Jupiter interaction, and they predicted a modulation of

the Jovian emissions at low decametric frequencies (1 8 MHz) by Europa.

We now know that several of the assumptions made in this model were

incorrect. More recently, Dessler and Chamberlain (1979) suggested

that observations of Jovian auroral hotspots could be attributed to the

position of the magnetic flux tube linking Europa with Jupiter. Also,

in a discussion of possible energy mechanisms to support life at

Europa, Reynolds et al. (1983) have proposed a novel electromagnetic

interaction for that satellite with the Jovian magnetosphere. These

ideas will be discussed further in Chapter III.









10

At this writing, the only purportedly significant evidence for an

an Europa modulation of the groundbased observations of the decameter

emissions was reported by Tiainen (1967). Until this dissertation,

Tiainen's report has not been confirmed or disproved; however, a very low

upper limit on satellite effects (other than lo) was measured by Kaiser

and Alexander (1973) and by Thieman (1979). The in situ observations

at Jupiter and Saturn by Pioneer and Voyager spacecraft have rekindled

much interest in the possibility of satellites interacting with their

environment. A thorough analysis of this problem is therefore both

useful and timely. Kurth et al. (1981) reported Dione modulation of

the Saturnian kilometer emissions; however Genova et al. (1983) have

discounted this effect as spurious. In view of the excitement

generated by lo's interaction, it is surprising that this evidence from

Saturn has not produced much interest in the published literature.

Dissertation Overview

Chapter II of this dissertation addresses the theoretical consid-

erations in electromagnetic interactions between a ponderous (i.e.,

planet-sized) body and an external magnetic field. We distinguish be-

tween the transverse electric (TE) and transverse magnetic (TM) modes

of induction, and we discuss the physical locations where interaction

could occur. We attempt a general view in this discussion, drawing

examples from measurements made on the terrestrial Moon. This body is

of similar size to the Galilean satellites, although the Moon's envi-

ronment differs significantly from that found in the Jovian magneto-

sphere. While the composition of the Moon may be close to that of Io









11

or Europa (at least from density measurements), the outermost layers of

the three bodies are dissimilar and represent a range of possible elec-

tromagnetic interactions.

Following the discussion of factors to be considered for induc-

tion, Chapter III outlines the physical conditions found in the Jovian

magnetosphere and at each of the major satellites. Jupiter's magneto-

sphere is divided into three regions. The inner magnetosphere is domi-

nated by Jupiter's radiation belts and by the relatively dense plasma

torus located at lo's orbit. The material in this torus is probably

supplied by lo's active vulcanism, and the mechanism producing the DAM

emissions is some induction process (not yet understood) caused by the

sweeping of Jupiter's powerful magnetic field past Io. The dynamic

region of space which is characterized by the inner magnetosphere

presents a marked contrast to the more benign middle and outer portions

of the magnetosphere. In the outer regions the predominantly dipole

figure of the internal Jovian magnetic field is distorted by azimuthal

electric currents, arising from a plasma disk confined to the plane of

the magnetic equator.

From the preceding paragraph we see that the icy Galilean satel-

lites (Europa, Ganymede, and Callisto) exist in a radically different

environment from that found at Io. Moreover, the composition of the

surface layers of these satellites allows only remote possibility for

electromagnetic interaction with the Jovian magnetosphere. With the

luxury of hindsight we note that prior to 1964, very few researchers

would have entertained the idea of significant interaction by any sat-

ellite. The detection of lo's signature in the DAM, and the more re-

cent discovery of Dione's modulation of Saturn's kilometric emissions








12

(Kurth et al, 1981) have shown that the subject of satellite-magneto-

sphere interactions warrants serious consideration. The electrodynamic

properties of each of the major satellites will be addressed in Chapter

III.

A description of the observational data used in this study is

presented in Chapter IV. There are two long term modulations (DE and

solar activity) that have played a prominent role in the collection of

the groundbased data, and these are described in this chapter. The

original purpose of analyzing the DAM observations was to disentangle

long term effects from the orbital resonance effects of the Galilean

satellites. An unanticipated result has been the discovery of subtle

modulation of the data, which repeats after 11.9 years (one Jovian

year). This new finding and its possible implications are discussed in

Chapter IV.

As we will see in Chapter V, the orbital resonance between the

inner two Galilean satellites leads to a repetitive geometry for

lo-Europa configurations; hence mutual effects of these two moons are

the most difficult to separate. From our discussion of both the

theory of satellite interactions and the Jovian environment, we do not

expect Europa's phase to affect the DAM; however, on the same grounds,

we suggest that an Europa signature would be more likely than either

Ganymede's or Callisto's. In Chapter V we describe the analysis and

results of our examination of the groundbased DAM data. For each sat-

ellite we determine an upper limit for the detection of its mutual

interaction with lo and the Jovian magnetic field, and we discuss the

implications of these measurements for the idea of intrinsic magnetic

fields in the Galilean satellites.









13

In Chapter VI we report on a search of the newly discovered kilo-

meter emissions for signatures of Jupiter's satellites. Although the

radiation is sporadic, the Fourier-transformed power spectrum of it

shows unexplained periodicities; however none of these coincides with

satellite periods (or beats with known periods). We conclude the dis-

sertation with Chapter VII, a summary of results. We find that only

lo's interaction with the Jovian magnetosphere results in electromag-

netic emissions which are detectable at this time. Only the DAM radia-

tion is modulated by lo, and the KOM emissions do not appear to be mod-

ulated by any satellite's phase.














CHAPTER II
ELECTROMAGNETIC INDUCTION IN PLANET-SIZED BODIES

Introduction

This chapter is an introduction to the body of knowledge con-

cerning the electromagnetic interactions between planets or satel-

lites and magnetized plasmas. This information serves as background

material for a discussion in Chapter III of the Jovian satellites and

their interaction with Jupiter's magnetic field. A concise account

of the phenomenon of electromagnetic induction by one system of

electrical currents into another was first described by Michael Fara-

day in the 1830's. He established the fact that electrical currents

can be induced in conductors by a changing current in a nearby cir-

cuit. He also found that the steady flow of a nearby current has no

effect unless there is relative motion between the two circuits. We

now describe these phenomena as two separate forms of electromagnetic

induction, each with its own properties and effects. In the first

case an electric field can be induced in a conductor as a result of

temporal variations in the external magnetic field (i.e., the

magnetic field which is associated with current flow in the inducing

circuit). Maxwell formalized this concept by expressing the induced

electric field (E) as a function of the time rate of change of the

inducing magnetic field (B). In differential form this is

7 x E = B
t









In the case of a steady current flow, dB/dt = 0 and any induced

electric field must be curl-free. This does not eliminate the

possibility of electrical induction though, since the relative motion

of a conductor through a static magnetic field leads to an induced

potential difference across the conducting body. Expressed in terms

of the electric field, the magnetic field, and now the relative

velocity (v) of the conductor with respect to the magnetic field, the

induction is seen to be

E = -(v x B)

In this chapter we will discuss the electrical induction which takes

place in large, planet-sized spherical bodies. Both time-varying

(dB/dtIO) induction and relative motion (v x B) induction are

believed to have observable consequences in bodies within our own

solar system. We will outline the general principles involved in the

electrical interaction of these bodies, citing specific solar system

examples where possible.

Induction by Time-Varying Magnetic Fields

One of the more important examples of time-varying magnetic

fields in the solar system is associated with the solar wind. This

more-or-less radial flow of ionized particles away from the Sun

carries a magnetic field whose strength and orientation have both

periodic and transient components. Since the Sun rotates once every

28 days, the radial flow appears to spiral out regularly, much like

the spray of water from a rotating garden sprinkler. During a solar

rotation the orientation of the solar wind's magnetic field changes

by 180 several times, leading to the concept of a "sector structure"

existing through the interplanetary medium. Short term transients









16
also occur in the solar wind which are generally associated with

flares and other types of solar activity.

Several of the planets in the solar system are known to possess

observable intrinsic magnetic fields. These are Mercury, Earth,

Jupiter, and Saturn. At least in the cases of Earth and Jupiter, the

magnetic and rotational axes do not coincide. Thus we have another

example of time-varying magnetic fields. An observer near one of

these bodies measures a periodic variation in the magnetic field

strength such that

6B = iB
6t

where w is the angular rotation rate of the planet and i is the

imaginary root of (-1).

The effect of a time-varying magnetic field on a planet-sized

body (or satellite-sized body) is to induce electrical currents to

flow within the interior of the body. These are called "eddy cur-

rents," and the closure of these currents within the body is com-

plete, as is indicated by the formulation of Faraday's law of induc-

tion. The magnitude of the currents is given, to first order, by a

statement of Ohm's law

J = OE

where J is the electrical current and n is the conductivity inside

the body. The penetration of these currents can be approximated by

the "skin depth" formula

S 1/2
PUIxJI








17

Here p is the magnetic permeability of the material composing the

body, and is again the radian frequency of the inducing field. If

both the conductivity and are large, the current can be confined to

a thin shell within the body. The material within that shell is then

shielded from the magnetic variations. This effect is commonplace in

the highly conducting metals used in everyday electrical circuitry,

where the current flows only in the outermost layer of the con-

ductor.

To distinguish the time-varying type of induction from that due

to relative motion in a static magnetic field, the time-varying mode

is called "transverse electric," or simply the TE mode. When induced

currents flow in the conducting body, a magnetic field results. This

induced magnetic field is of the poloidal variety, and it is oriented

so as to counter the inducing field. If not driven periodically, the

induced currents and poloidal field will decay over a characteristic

time scale r, called the "Cowling time"
2
r = L-o

Here Lo is a "characteristic" length of the conducting body (e.g.

for a sphere Lo is usually taken as the radius R). Parker (1979)

describes this time as a measure of the current-carrying capacity

(area times conductivity) of the body, and he calls it the magnetic

relaxation time.

Induction by Motion in a Magnetic Field

In contrast to the time-varying case, the motion of a body

through a static magnetic field causes an induced electric field

whose effect depends only on the electrical conductivity in the











outermost shell of the body. This type of induction is called

"homopolar" or "unipolar" induction, and it is often described in

terms of the model shown in Figure 2.1. Here a rotating metallic

disk is placed in an external magnetic field. The velocity of any

point on the disk depends on its distance from the axis of rotation,

hence a potential difference exists between the outer edge and the

center. Electrical contact is maintained at the edge of the rotating

disk (through some kind of sliding contact), and the circuit is

closed by connecting the other end of the wire back to the axis of

rotation. Alfven and Falthammar (1963) describe a variation of this

concept which more closely parallels the planetary magnetosphere. In

their model a bar magnet is rotated about its long axis, producing an

electrical current in the pictured circuitry.

In order to scale the classical unipolar generator system up to

planetary-sized bodies, we can consider different aspects of the two

examples used in the TE mode discussion--the solar wind impacting the

planets, and the planets with intrinsic magnetic fields. First, each

of the planets in the solar system represents a conducting sphere

(conductivity considerations of these bodies will be examined later).

As the solar wind plasma and its magnetic field flow past the

planetary bodies, a (v x 5) electric potential difference is set up

across the planet's diameter. In the second example, we note that

several planets with intrinsic magnetic fields also possess one or

more satellites. When the planetary magnetic field rotates past the

more slowly moving satellite, a potential difference exists across

that moon's diameter.




























Figure 2.1. Induction by motion. An example of the produc-
tion of electrical current by induction is shown in (a), where a
rotating metallic disk is placed in an external magnetic field. A
(v x B) potential difference exists across the radius of the disk.
A variation of this concept is shown in (b).





















ROTATING
DISK

















ROTATING
BAR MAGNET


\ Brushes/Sliding Contacts


(a)


Galvanometer


(b)







21

In the (v x B) induction, closure of the electrical currents

must take place through the outermost conducting layers of the body,

since

x E = ,^_B = 0
St


The mechanism by which current closure is attained is crucial to the

unipolar mechanism, and the subject will be treated throughout the

remainder of this chapter. The unipolar mechanism is possible only

if a conducting part of the body (e.g., its surface or ionosphere) is

in electrical contact with the surrounding plasma. The (v x B)

induced electric field produces axial currents, leading to a toroidal

magnetic field. This type of induction is known as "transverse mag-

netic" or TM mode. Characteristics of both TE and TM induction are

shown in Figure 2.2.

In 1966, Gold suggested that one consequence of (v x B) induc-

tion might be observable in the Earth's Moon. He hypothesized that

the diffusion of the solar wind's magnetic field lines through a

solid body such as the Moon could not occur as fast as the unimpeded

flow of the solar wind plasma past that body. The magnetic field

lines would be "hung up" in the outer layers of the Moon, and a

"stagnation field" would result which could masquerade as an intrin-

sic lunar magnetic field. Through manned exploration we now know

that such a stagnation field does not exist in the Moon, but it is a

worthwhile exercise to formulate the concept for possible discussion

of other bodies immersed in magnetic fields. Gold reasoned that the

diffusion of magnetic field lines through a spherical body could be

no faster than the Cowling time, which we described earlier. The


















u 4+- 4-J 00 V4-

cu 0 0
-- I L L- Oc/
- S I- O -CL
A- '- s.U L 0 >
(A cu? V
-- IW ut> 0


4 U 4J CC 4--
U 11 **- u L- 4-
- C: O ( U +O ) C
>) a) u- S- ZL

-Z r S,- U- A00
LU' U =3U-E 3









im
UU












CI













ME I CC
II






































1W






LL


C 0*M
WI IX I
LL >,J II
(3 IC Q I

LU .1 C






( 0











-> >
L L S.J

>- ri I > CT 0
I-- S- 1 L.



C +-) C -
M c










unimpeded flow of solar wind plasma, and hence the solar wind's

magnetic field, past a sphere of radius R requires a time (t) such

that

t = R/v.

Assuming a highly conducting lunar surface, Gold found a time of sev-

eral months would be necessary for diffusion of the magnetic field

lines through the Moon. Tozer and Wilson (1967) came to a different

conclusion because they (correctly) assumed that the lunar surface

was not composed of highly conductive material. They did suggest,

however, that the Moon might possess a highly conducting core, and

that the solar wind's magnetic field lines could stagnate there,

rather than at the surface as Gold had predicted. Recently Russell

et al. (1981) have reviewed the measurements of the Moon by magnetic,

gravitational, and seismic means, and they question whether the

balance of evidence indicates the existence of a molten, highly

conductive lunar core.

We can describe the consequences of (v x B) induction in

another way. In the rest frame of the conducting body, the resulting

electric field induces an electric current within the body in such a

sense as to increase the magnetic field strength on the windward side

of the body (Dessler, 1968). The electric field is not seen by an

observer moving with the plasma flow; instead, in the frame of the

solar wind an increased magnetic field is seen on the impacted sur-

face of the body. As mentioned before, if a stagnation field is

formed, the "piling up" of field lines in one hemisphere could look

like the bow shock of an instrinsic magnetic field. Also, a cavity

in the solar wind flow would give rise to a wake. In his original










paper, Gold (1966) suggested that the accumulation of field lines

would continue until the stagnation field pressure balanced the ram

pressure of the solar wind plasma
1/2 NMv2 = B2
2u
Here N is the number density of ions, m is their mass, and v is their

velocity. The maximum stagnation field is given by B. Sonett and

Colburn (1968) describe how the unipolar system is "self-regulating."

When the stagnation field pressure is at its maximum (assuming large

conductivity in the body), the diversion of plasma flow and field

lines away from the body effectively decreases the inducing magnetic

field strength, thus diminishing the net electric field available to

drive the unipolar currents. Figure 2.3 summarizes the stagnation

field characteristics.

What are the energy considerations in the planetary unipolar

mechanism? A conducting body in the presence of a (v x B) induced

electric field removes an amount of energy from the impinging plasma

which is proportional to the rate of Joule heating in the body.

Sonett and Colburn (1968) give this rate as
E 2
H = OE



Here the new variables are H, which is the rate of Joule heating, and

6, which is the mass density. The parameter a here represents the

bulk electrical conductivity. The preceding discussion has been

based on the assumption that the impinging plasma was spherically

symmetric about the planetary body. This may not be a bad approxima-

tion within some regions of the Jovian magnetosphere, but the Earth's

Moon produces a cavity in the solar wind flow (Ness, 1966). Horning















o I





>- r0 -- n






> 4-- --


















o=
C: 0i) 3 3
1- D 0 N
U C C :**

) 0 *- S- 0"


Z a) 0u Ca
I0 3 z-- 1
- -- -)






4- -C
,- *- Q- C






























m
4- =-- *




















c Ln 0 C-
( 0
., --























U-
ra O'
V)) V)
1-
















0 a>
L -0- (A
C 0) J3

j, i ^4 L


















| E t.
SU


II W 4





_ EI (a
W L (A


















u. r-,
LL3



4-
4- )
0-- 0 S-
O 4-2





0 (A
c( -
O m
L i. -


>- +
- ai va
C: C










and Schubert (1974) analyzed the case of a uniformly conducting

sphere in an asymmetric plasma and found that the total power dissi-

pated is only about 0.76 that of the spherically symmetric case.

Also, when rotation can be neglected, the volumetric heating rate at

the sub-flow point is nearly an order of magnitude larger than at the

anti-flow point.

In a description of possible satellite effects in the Jovian

magnetosphere Goertz (1981) describes the induction phenomenon from a

different perspective. He discusses the distortion of the (v x B)

generated electric fields in the vicinity of a satellite for three

cases of bulk electrical conductivity in the satellite. If the body

is a perfect insulator ( = 0), then the plasma flow impacts the sur-

face directly since there is no distortion of the surrounding elec-

tric field. In the limit of a perfectly conducting satellite

(o = -), the electric field within the body is reduced to zero by the

accumulation of polarization charges on its surface. Since the

charges are free to move in a conductor, the surface charge on the

sphere builds up until it cancels the applied electric field within

the interior. Outside the body, and far away from it, the external

field will be undistorted. However, nearby the sphere, the body will

act as an electric dipole. The motion of charged particles along the

electric equipotentials will not have access to the sphere's

surface.

Conductivity Considerations

An Ionian unipolar mechanism was offered by Piddington and

Drake (1968) as the driving force of the Jovian DAM radiation. This

concept was formalized by Goldreich and Lynden-Bell (1969), and






29

much work has been done to refine this process as in situ observa-

tions became available (e.g., Neubauer, 1980; Gurnett and Goertz,

1981). The material conductivity considerations in TE (time-varying)

induction are relatively straightforward compared to the TM (uni-

polar) case. We will therefore focus on the unipolar induction mech-

anism and discuss the various ways that electric current closure can

be attained. Sonett and Colburn (1968) outlined the three possibil-

ities for circuit completion in the unipolar process. We will dis-

cuss each of these cases in the following sections (Table II-i).

When a planet-sized body possesses a sizeable intrinsic magnet-

ic field (Type I body), the discharge of (v x B) electrical currents

will take place at the magnetopause. Therefore, the surface covering

of the body (i.e., atmosphere present or not) is not relevant to con-

ductivity discussions for the Type I body. The magnetopause is the

boundary that separates the flowing solar wind plasma from the mag-

netospheric plasma (or, within a planetary magnetosphere, the satel-

lite's magnetopause separates the ambient plasma which corotates with

the planetary magnetic field from the space dominated by the satel-

lite's magnetic field). Stated in other words, electric currents

must flow within this boundary region because both the intensity and

direction of the impinging magnetic field must be altered from the

unimpeded solar wind characteristics to magnetospheric values. The

Earth is a Type I body, as are Jupiter, Mercury, and Saturn.

From observation, we know that the main dipole field of the

Earth could be caused by uniform magnetization throughout the sphere.

However, when we examine the crust of the Earth we find that the

upper layers are not magnetized to the extent required to match the













TABLE II-i
UNIPOLAR CLASSIFICATION


INTRINSIC
TYPE MAGNETIC FIELD ATMOSPHERE


I Yes N/A

II No No

III No Yes


NOTE: Shown here is the Sonett and Colburn (1968) class-
ification scheme for unipolar induction in planet-sized
bodies. The classification is based on two basic prop-
erties of a planet--the existence of an atmosphere and the
existence of an intrinsic magnetic field.






31

measurements. Moreover, we know that temperature increases rapidly

with depth in the Earth so that within a few 10's of kilometers the

Curie point of most ferromagnetic materials has been exceeded. This

requires that we find an alternate explanation for the Earth's mag-

netic field. Presently we attribute the generation of intrinsic

planetary fields to dynamo action within a molten core. Throughout

this discussion we will accept as "given" the concept of intrinsic

magnetic fields, since we will only be concerned with their observ-

able consequences.

When a spherical body possessing an intrinsic magnetic field

is placed in an external field, the relative strength and orienta-

tion of the fields must be taken into account in order to describe

the subsequent motions of charged particles. While electric fields

can accelerate particles, magnetic fields can only constrain their

motion and guide them toward or away from the body. In Figure 2.4

we display four possibilities for the superposition of magnetic

fields. In (a) and (b) the intrinsic field is parallel to the

external field, and the effective diameter of the sphere to charged

particles which are following the external field lines is actually

larger than its true diameter. In (c) and (d) the intrinsic field

is antiparallel to that of the external field, and particles may be

carried around the sphere without impacting it. This is the so-

called "closed field" model where a magnetic bubble forms around

the sphere, protecting its surface from plasma. There will be

very little asymmetry between the upstream and downstream plasma flow

in this case. Following the description of Goertz (1981) we note

that the motion of charged particles in the antiparallel case is





























s-



-- =-) .-
4-) 4-> U -- 4- 4

C- 3 ~ r.- a. S 4-)
*L- ct c-C *V-

a)- E -E- -in fr CU
4-0 -0 U *r--
S*r- 0 3 C- --
a- "0 +. -- CD

M 4- (n --- .C"
S TC S-- -)
S04- S- L- V C
0C C 4-C ''
C/a C 0 a) )
CU -: X< (/4->
-C C 4 -) )-
- *'- --- 5 l -


S-0 s- c C
CU- CA 4. CL -
4--L *O 'cU3
E- U 1 3) : -C
( E c; I 3 0 *
- E .- Ca-
4-'- *- n u4-J C CC
0U r- C 3 )
C*- ir- 0C 0 -'
m 4- *r- = Qr-
O -(5 L. n CnCU
E Jn.r--i 1 *r- N1
*E- fO Q to a. .- N

C.. r V- v CO
E- U4-- S- S-
0 W)_S- 4 -0 W0 -

*r- *C CU -- -o- C W O
*r-- E a *-- c: -0
4- ) 4-) aJ O4
.1 4-1 E -'"- I v--. S
SC 4C a) 4-
0 (C E- E O (D
+ -0 4- ) -4 ) C 4)
* 3 C -- to t En *- o
CL 0 Wr C E4- C
3 0 *i- *- ra o 0
OV 5- 4- a) S- S-C- -C 4-)

CU 4- C 0 RC (A
*r- *- *- O L-
4 4- +- ) -0 0- 0a-
* S- in a) cm-
Cs c a)- OiCc E
=3 cD cu 3 Lo)r
)< CU CU U A CU Q. --

Cn O r- : C ) *r-
*r- ) *m a) o c vi
LL ll 4- + *L- *r- Q)
Q C a ) 4-- -
*- *- 0 t t 3
(n L. r-- I r- -e

O C C 0 *r- LL.
0 r- 4- 4- 4-) S- 4- ---















0
a


E












0
















0
ac


V














0
A V






34

equivalent to a perfectly conducting sphere in an electric field--the

(analogous) electric field lines outside the body are distorted and

charged particles are transported around the sphere. The parallel

case is equivalent to an insulating body in an electric field, where

particles may strike the sphere directly. In this case, the voltage

drop available for the (v x B) mechanism is actually larger than the

physical diameter of the body (Ip, 1981).

With the recent expansion of knowledge concerning the satel-

lites of the outer planets, it appears that most bodies in our solar

system are of Type II--they possess neither atmosphere nor intrinsic

magnetic field. The Moon is the best studied of these bodies. If

any unipolar interaction takes place, it must occur in the surface

layers of the body. On the Earth, most rocks and minerals are poor

conductors of electric current, with passage of current almost en-

tirely due to electrolytic conduction through interstitial water

rather than by electron transport. Feldspar is the most common rock-

forming mineral group, accounting for -60% of the Earth's crust

(e.g., granite, basalts). For most rocks, however, chemical composi-

tion is of secondary importance in determining their bulk electrical

properties. Porosity and fracturing are much more important (Grant

and West, 1965). Pressure has little effect on the electrical prop-

erties of rocks, except that they become more insulating as their

pores and cracks are closed with increasing pressure (Parkhomenko,

1967; Schmucker, 1969). For metals, conductivity decreases with

increasing temperature. However, for the semiconductors which are

believed to comprise a majority of the constituents of the solid

bodies in the solar system, conductivity increases exponentially with









increasing temperature (Anderson and Leaver, 1969). This occurs

because more charge carriers (electrons and holes) are available with

a rise in temperature.
Although the Earth is a Type I body and it retains an atmo-

sphere, we note that over 70% of its surface is covered with water.

The water molecule has a relatively large dipole moment, arising from

the net separation of positive and negative charge centers between

the two hydrogens and the oxygen. Ground water, usually carrying

large amounts of salt in solution, is the rock component which is

most important in determining the conductivity in the first few kilo-

meters of the Earth's surface (Parkhomenko, 1971). However, since

the conductivity of seawater is so much larger than that of typical

crustal rocks (by more than an order of magnitude), induced electri-

cal currents in the oceans due to time-varying magnetic fields should

be much larger than in the land masses (Price, 1967). The conductiv-

ity of seawater is ~1 -1n-1.

Europa appears to be completely covered by a thick layer of

ice, while Ganymede and Callisto show at least a partial ice cover.

At the present time, about 10% of the remaining land masses on Earth

are covered with ice or snow. Glen and Paren (1975) describe the

difficulties in measuring the electrical characteristics of ice and

snow. Theoretically, all of the electrical properties of snow (or

frost) may be derived from a knowledge of the characteristics of ice.

However, the compositional purity, the fraction of unfrozen liquid

water, and the uncertainty in the density of crystals makes a real-

istic treatment of the problem quite complex. The low frequency

(bulk dc) conductivity of pure ice at -100C is in the range of 10-7 to

10-8 -1m-1. The conductive response to higher frequencies is








36

about two orders of magnitude larger (Mellor, 1977). As will be

addressed in a later section, it is believed that the existence of a

mixture of rock and liquid water near the surfaces of the icy Gali-

lean satellites may have important consequences in the closure of

unipolar induced electric currents.

The Moon is certainly a Type II body, and it is the second most

studied body in the solar system. Although no large scale magnetic

field is currently present, significant levels of remnant magneti-

zation were found in the Apollo lunar samples. Further, magnetic

measurements of the surface revealed spatial variation over kilo-

meter distances, implying sources within the crust. Runcorn et al.

(1983) review the disagreement which still exists at the time of this

writing concerning the origin of the Moon's remnant magnetization.

Leading theories include meteoritic genesis by impact magnetization

(e.g., Cisowski and Fuller, 1983) and internal dynamo generation of an

early lunar magnetic field. For the Moon, analysis of poloidal field

variations (i.e. those due to time-varying magnetic fields) gives the

most accurate conductivity information at depths between 200 km and

800 km. Dyal et al. (1977) model this region with a relatively high

conductivity of 10-2 to 10-3 Q-1m-1. The outermost 80 km of lunar

crust is believed to have a much lower conductivity of -10- n-1m-1.

We may compare this to the conductive properties of the Earth. In

the first 250 km depth o is -0.01 B-m-1; at 600 krm depth effective

shielding sets in and a is -0.2 n-1m-1 (Schmucker, 1969).

For a Type III body which possesses no intrinsic magnetic field

yet retains an atmosphere as its outermost layer, the closure of

unipolar electric currents should occur in the ionosphere. The










interaction will take place at the ionopause where the plasma pres-

sure must be balanced by the static and induced electromagnetic pres-

sures in the ionosphere. Only a few substantial atmospheres exist in

the solar system for the rocky bodies which have been studied.

Venus, Earth, and Titan have substantial atmospheres; Mars has a more

tenuous gaseous shell, with a surface pressure -0.01 that of Earth.

No atmosphere has been detected for Europa, and a Voyager-based

occultation experiment placed a very low upper limit on the existence

of an atmosphere at Ganymede. An ionosphere is known to surround lo,

even though by Earth standards its atmosphere is rather tenuous.

Kumar and Hunten (1982) provide a comparison of the peak electron

densities found in several ionospheres:
Earth 1 x 10 6cm -3

Venus 5 x 10 5cm -3

Mars 1 x 10 5cm -3

Io 6 x 10 4cm -3

Surface effects will be radically different between Type II and

Type III bodies. Without an atmosphere, sublimation of ices into

space may occur, while contamination of the surface by the surround-

ing medium may also be important. In such exposed bodies, charged

and neutral particles may also be lost from the ambient plasma to the

sphere's surface. Some neutralization of the plasma will take place

as it strikes the surface of a body without an atmosphere. Chemical

changes caused by solar ultraviolet radiation may take place on the

surface. Sputtering of the surface layers by high energy particles

may also occur, causing fresh ions and neutrals. If an atmosphere is

not initially present, the sputtering may create a weak corona.








38

Vulcanism and other less energetic forms of outgassing may contribute

to the maintenance of the outer shell. If the body can retain an

atmosphere, there exists the possibility of an ionosphere created by

solar photoionization, by electric field acceleration, or ionization

by the plasma.

Completion of the Electrical Circuit

The unipolar mechanism can be viewed as being composed of three

sub-systems, each of which is crucial to the existence of the

process. Given a sufficient potential difference (v x B), these

parts can be summarized as:

(1) The conductivity of the surface layers of the conductor.
This has been reviewed in previous sections and, for the
remainder of the discussion, we will assume that the
requirements for sufficient conductivity have been
fulfilled.

(2) The closure of the electrical pathway between magnetic
field lines. In the case of Jupiter-lo interaction the
Jovian ionosphere is believed to provide the conducting
path between field lines which pass through Io. In the
example of the solar wind, Alfven (1981) points out that it
is not clear how the field lines are closed at the Sun's
surface.

(3) The electrical conductivity of plasma along the magnetic
field lines which link the driving force described by (1)
with the completed circuit in (2). Once again, in the
planetary example the continuity of the field lines is more
readily apparent. For the solar wind's magnetic field this
description may not be adequate.

The problem of plasma conductivity is therefore intimately connected

with the closure of the unipolar current system. (In most cases,

however, it appears that the major impediment to the process is the

conductivity of the solid body, Sonett and Colburn, 1968.) Alfven

and Falthammar (1963) describe the steps leading to a "standard"











approximation for the conductivity of a plasma as a function of

temperature (T)

o =(constant) T3/2

where the constant is defined for a given plasma and depends on such

quantities as the fraction of ionization and the atomic number of the

constituents. Once charge carriers are freed from the solid body by

the unipolar mechanism, it has been generally assumed that they would

be swept away with the ambient plasma along the magnetic lines of

force, thus completing the electrical circuit. Goldreich and Lynden-

Bell (1969) considered the conductivity of the plasma to be infinite

along the magnetic field lines connecting Io to the Jovian ionosphere

(i.e., the Io flux tube). In both Sonett and Colburn's general dis-

cussion and in Goldreich and Lynden-Bell's description of the uni-

polar mechanism at lo, the generation of Alfven waves as an alterna-

tive method of current conduction was mentioned. In view of the

current theories of Io's interaction with the Jovian magnetosphere, a

discussion of this alternate mechanism is warranted.

Drell, Foley, and Ruderman (1965) were the first to present a

formal description of the completion of (v x B)-induced electrical

circuits by Alfven waves. They applied the theory to the orbit of

the Echo satellite through the Earth's ionospheric plasma, and they

found that a significant decay of the satellite's orbit could be

explained by the conversion of mechanical energy into Alfven disturb-

ances. They determined a lower limit to the size of a conducting

body necessary to generate an Alfven disturbance, by relating the








40

dominant ionic gyrofrequency (-i) to the characteristic length (Lo)

of the conductor and its velocity
Wi v/Lo

Solving for Lo in the near-Earth orbit yields a value of -10 m.

As seen from the rest frame of the plasma, electric currents

arising from the spherical obstacle are carried off along the magnet-

ic field lines by the transverse Alfven waves. In the sphere's frame

of rest however, the electric currents are carried back into the

plasma flow at the flow velocity while moving along the magnetic

field lines (Southwood et al., 1980). The deceleration of plasma

from its unimpeded flow is accomplished by the (v x B) force in the

conducting layer of the sphere (i.e., crustal shell or ionosphere).

A transfer of momentum between the plasma and the sphere occurs as

the plasma flow energy is dissipated by Joule heating within the con-

ductive layer. The magnetic field lines are distorted by the sphere,

and the flow momentum is directed perpendicular to the lines as

Alfven waves. One can view this phenomenon as a transfer of stress

from the impeded plasma flow, through the distortion of magnetic

field lines in the sphere, which is then communicated back to the

plasma along the field lines as an Alfven wave. The accompanying

Figure 2.5 shows the distortion of magnetic field lines from two

views. Drell et al. (1965) drew an analogy to a series of transmis-

sion lines which, when contacted by the moving conductor, carry the

resultant charge separation produced by the (v x B) electric

induction. The electric current is carried through the circuit as

Alfven waves, driven by a d.c. battery (the moving conductor).
















C
> C
r -- C 5.O


C O -- LWC
C-4- = i= > (
4- 4--
O S0 C
- 0> > 0,


C 0 *- *r-
o 10 / 0 )

4-3 (0 -0 4- .



W C (v

L0- u- --




T W 0 4-/
Sr- 0 >
0r 4- 4- 0W



40 W.0 -0
-E U O -
* 4-' 3 0 W4 >


L 4- (M
W 0- 0
0 3: Un
U 4- >, *-
01 T .4-'
SL ) 3 *I- o

71 *r- .C C 0 (U
S- S L /) *r- 0
4-4- i- O
C 0 > Vn

*i- r-- C J

I- 3 r- >
0-4- 4- L.
-, -0 O
4- C U C .1 *r-


4-' .CI-S
mo >1 Co cm

0 .C E L
C .- -
0) C7) .3
Cm i- C 0) E
( 0) L *r- 3
U 0 S. (U 4- C
> *r- 4-' 0
M0 to U r)- -
3: o u
-E 4- w0 C
C: CL- 0 S-
C 0.O ..C L.
o) n 4-' )
> 0 4- C
4- 4r- 4- X W
- *r- W 0 >
Sr- *r- 4-
04- C 0 -
U) o -C C
0 *- 4 -'
Lfl 10 *W- 4 0
4 S- 0
c\J Un 0 o4 4 I-
W0 C 4-)

- W U *r- 0 *

3 ( E ( ) C>
L0) W CC )0-
**- (U i- Uin & >
L=3 0- W 4-


L C W L.-

*- X Q. *r- in
t0 0n S *r-


-C
4-3



0
r(



4r-
-'

E
v,
C



c-




0
> *


0)
C>
O *r-






4- )
a
o th



L(-
S0


W in







ea c
-- o

-)
4 C

O 0r-
Er
>O


u- 0

-0
0)


LCL
UQ)0

W 0.
0
r-O

< r-
-I-'
4- D)


a -
l < EE



14->






< E


0-
L

0

U)
I>


































N. 4
Zoo%
PC.
N. 'N.







0r


















-J4













CHAPTER III
THE JOVIAN MAGNETOSPHERE AND SATELLITES

The magnetosphere of Jupiter is certainly the most voluminous

object in the solar system, with the exception of the heliosphere

itself. Since the 1955 discovery of the DAM emissions, we have known

that the space surrounding Jupiter and influenced by its magnetic

field is a dynamic environment. Measurements of the internal heat

flux of Jupiter have shown that the ratio of emitted thermal energy

from the disk of the planet to the absorbed solar energy is about

1.67 (Hanel et al., 1981). It is widely believed that the immense

energy source necessary to drive the diverse magnetospheric processes

is provided by the rotation of the giant planet (e.g., Goertz, 1983;

Hill et al., 1983). Prior to the Pioneer and Voyager in situ

measurements, most of our knowledge of the Jovian magnetosphere was

through observation of the high energy electrons which produced the

nonthermal DIM (decimeter) and DAM emissions.

For the Earth's magnetosphere, the solar wind is the primary

source of plasma and energy. In the case of Jupiter, the total power

from the solar wind over the entirety of its magnetosphere is about

1015 watts. While this is a substantial load, an energy transfer of

-10% would be necessary to produce the observed Jovian aurora, and an

efficiency this high is unlikely (Thorne, 1983). The true energy

source driving Jupiter's magnetosphere is most likely the planet's











rotation. The transfer of energy occurs through the torque exerted

by Jupiter on the heavy ions scattered throughout the magnetosphere.

The process is accomplished through the Birkeland current system--ex-

tracting rotational energy from the upper layers of Jupiter's atmo-

sphere in an attempt to force corotation on the magnetospheric

plasma.

Since the spacecraft encounters, it is most common to discuss

the Jovian magnetosphere as divided into three major spatial regions--

the inner, the middle, and the outer magnetosphere.

The inner magnetosphere is dominated by the magnetic field of

Jupiter itself. This region extends outward from the planet's cloud-

tops to about 6 Rj and includes the satellite Io and the associated

plasma torus. Also the Jovian ionosphere, the synchrotron emission

belts, the newly discovered Jovian ring, the satellite Amalthea, and

other smaller satellites are found in the inner magnetosphere.

The middle magnetosphere is taken to have an inner boundary of

>6 Rj, and an outer boundary extending 30 to 50 Rj. The three icy

Galilean satellites reside here. In this region the simple dipole

model of Jupiter's magnetic field no longer provides a satisfactory

description. Models usually show that the magnetic field morphology

of this region is dominated by azimuthal electrical currents con-

tained in a "magnetodisc" which is about 5 Rj in thickness.

In the outer magnetosphere the effects of the solar wind are

prevalent, leading to a day-night asymmetry. This region includes

the bow shock, the magnetopause, and the magnetotail. Depending







45

on local conditions, both the planetary dipole and the magnetodisc

may influence this region.

Hill et al. (1983) point out that one of the more striking

features of the Jovian magnetosphere is the confinement of plasma and

energetic particles into a relatively narrow range of latitudes,

leading to the magnetodisc model. Prior to the Voyager visits, a

competing model of the forces driving the magnetospheric processes

was advanced by Dessler and Vasyliunas (1979). In this scheme many

of the observed periodic phenomena (e.g., the DAM and lo's control of

it) are accounted for by an "active sector" of longitudes in the

Jovian magnetic field. This anomalous region in an otherwise sym-

metric magnetic field is hypothesized to be the result of higher

order multipoles of Jupiter's internal field (the magnetic anomaly

model). Although it now appears that the magnetodisc model accounts

for many of the observations, there are still longitudinal asym-

metries in much of the Jovian phenomena, and proponents of a synthe-

sis of the two models can be found (Hill et al., 1983)

Because we will be concerned with the energy budget of the

Jovian satellites, it is instructive to first consider the genesis of

the orbital resonance found amongst the inner three Galilean satel-

lites. Yoder (1979) explains that the resonant situation arose in

the following manner. Originally the three satellites (Io, Europa,

and Ganymede) began in orbits far from their respective 1:2:4 com-

mensurability. Large tides were raised within Io by the proximity of

Jupiter, but the tidal dissipation on Jupiter by lo caused the satel-

lite to spiral outwards--analagous to the Moon's outward spiral from

the Earth due to oceanic tidal dissipation. As lo's orbit reached








46

commensurability with that of Europa, both satellites' orbits expand-

ed out at a fixed rate. More tidal heating occurred within Io in

order to maintain this resonance. When Europa reached the 2:1 com-

mensurability with Ganymede, further tidal dissipation within Io was

necessary to maintain the three-body resonance lock. This is the

present day situation--the inner three Galilean satellites maintain a

strong orbital resonance at the expense of enormous tidal dissipation

within lo.

Peale et al. (1979) found that the tidal heating rate inside Io

would be an order of magnitude larger than that produced in the

Earth's Moon (a body of similar size and density). They further cal-

culated the tidal heating in Europa to be about 20 times less than

that of lo. Yoder (1979) estimated the tidal dissipation within

Europa to be 0.01 lo's. The effect on Ganymede is even smaller,

which suggests that tidal heating is not a significant source of

energy for the icy satellites. Other energy sources for the Galilean

satellites and Amalthea will be considered in this chapter, but first

we will discuss each of the regions of the Jovian magnetosphere.

With a knowledge of the environment, we may then apply the concepts

of electromagnetic induction to all of the major satellites.

The Inner Magnetosphere

The inner magnetosphere is that region of Jupiter's plasma en-

velope in which the magnetic field from the interior of the planet is

not distorted by the motions of the charged particles. The field

completely dominates the particle motions, causing the magnetosphere

here to corotate with the field and the planetary interior. The

planetary field is known to display non-trivial quadrupole and








47

octupole moments (Acuna et al., 1983). The dipole axis is tilted 9?6

from the rotational axis, toward longitude 2020 (XIII, 1965).

Jupiter's ionosphere has been detected by spacecraft as extend-

ing from several hundred kilometers to several thousand kilometers

above the planetary cloudtops. Profiles of electron density versus

altitude show a pronounced layered structure at lower altitudes, with

a strong possibility of both spatial and temporal variations. A com-

prehensive review of the Jovian ionosphere may be found in Strobel

and Atreya (1983). Beyond the EUV ionizing radiation from the Sun,

other potentially important sources of ionization in Jupiter's upper

atmosphere include galactic cosmic rays, sputtered particles from the

surfaces of the Galilean satellites, and the unipolar action of Io

with Jupiter's magnetic field. Mapping of the Jovian ionosphere was

performed by both Pioneer spacecraft and both Voyagers. Differences

between the sets of observations are explained by the time of

encounters with respect to the solar cycle--the passage of the

Pioneers occurred during solar minimum, while that of Voyagers

occurred near solar maximum.

Although Jovian aurora had been reported previously, Sandel et

al. (1979) identified Voyager observations of the activity as the

projection of magnetic field lines passing through the Io plasma

torus onto the planet's cloudtops. The region displaying aurora

formed a donut shape, centered on the magnetic pole. Barbosa et al.

(1981) have suggested that the mapping of the field lines passing

through the magnetodisc may be an alternative to the Io plasma torus.

The power dissipation in the aurora is several orders of magnitude

greater than in the radio emissions. Broadfoot et al. (1981) estimate








48

that over 1013 watts is supplied to the Jovian aurora by precipitat-

ing electrons. In comparison, Alexander et al. (1981) calculate that

-1011 watts of radiated power arise from all low frequency emissions,

when integrated over all Jovian longitudes.

The next feature encountered as we move radially out from the

planet into the magnetosphere is the newly discovered Jovian rings.

Although their existence had been predicted by magnetometer measure-

ments with the Pioneer spacecraft (Acuna and Ness, 1976), positive

identification required imaging by Voyager. The rings consist of a

"bright band" starting abruptly at 1.81 radii from the planet, which

ends gradually at 1.72 Rj where it blends into a "faint disk." Both

features are enshrouded by a "halo." Two of the satellites discov-

ered by the Voyagers appear to be associated with the rings. Adrastea

(J XV or 1979J1) has a radius of about 10 km, and orbits the planet

at the ring's outermost edge (1.8064 Rj). Metis (J XVI or 1979J3)

has a radius of -20 km and orbits at 1.7922 Rj. Both objects are

dark with albedos like Amalthea (-5%). Burns et al. (1984) review

the observations and theory of these tenuous rings. Although materi-

al in the rings may play an important role in the total electrody-

namic picture of the planet's magnetosphere, they probably do not

contribute to interactions of the larger satellites. Certainly their

presence could not be neglected in a discussion of the minor interior

moons, but that topic is beyond the scope of this dissertation. The

interested reader is referred to Grun et al. (1984).

At Jupiter, the direct analogue of the Earth's (Van Allen)

radiation belts would include the entire inner magnetosphere and to











some extent, the middle magnetosphere (Goertz, 1983). Energetic

particles are transported into the innermost portions of the inner

magnetosphere by radial diffusion from the plasma torus at lo's

orbit. As the charged particles diffuse closer to the planet and

gain energy, they begin to emit synchrotron radiation which results

in the nonthermal decimeter (DIM) emission which is observed by

terrestrial telescopes. These "synchrotron belts" appear to extend

from the planetary cloudtops out to several Jovian radii (~3 Rj),

with a sharp transition at about 1.8 Rj (de Pater, 1981). It is

not known whether this is simply coincidental with the position of

the Jovian rings, or if in fact the two features are physically con-

nected. "Hotspots" of DIM occur at CML 2250 (xIII) and are not yet

fully explained (e.g., Carr et al., 1983); however, this is another

observation that lends weight to the "magnetic anomaly" model of

Jupiter's magnetosphere. Time variability of the DIM emission is

observed (Klein, 1976) by changes in its intensity, and by movement

of the peak of the radiation both toward the planet and shifted in

XTII (e.g., de Pater, 1983). Amalthea's orbit at 2.5Rj is within

the outermost boundary of the synchrotron radiation.

The dominant feature of the inner magnetosphere is the plasma

torus associated with Io. The plasma torus was first discovered by

groundbased optical observations of neutral sodium and ionized sulfur

(Brown, 1974; Kupo et al., 1976). It lies in the planet's magnetic

equatorial plane and engulfs the orbit of Io. Direct plasma mea-

surements in situ (Bagenal and Sullivan, 1981) have shown that the

torus consists of a "colder" inner portion (electron temperature







50

Te-20,0000 K) extending from about 5 Rj to 5.5 Rj, and a "warmer"

(Te-100,0000 K) extending out to -7.5 Rj. Voyager EUV (extreme ultra-

violet) observations indicated that the plasma in the cool inner

torus shows a strong dependence on magnetic longitude (1III), while

only very weak dependence was seen for the hot outer portion (Broad-

foot et al., 1981). The same EUV observations have determined that

the energy loss due to the radiation from the hot.torus exceeds

2 x 1012 watts. The composition of the torus is predominantly S and 0

ions, leading many researchers to conclude that the source of the

material is lo. Bagenal and Sullivan (1981) note, however, that more

sulfur is observed in the torus than would be expected from even

complete dissociation and ionization of the abundant logenic SO2.

Spectrophotometry of the emission in the sodium "D" line requires a

source production rate of -1027 atoms/sec (Brown et al., 1983).

Although it can be assumed that Io is the major source of energetic

S, Na, and 0, Hamilton et al. (1981) suggest that the icy Galilean

satellites and the solar wind may further contribute smaller amounts

of oxygen.

The Middle Magnetosphere

Although first observed by Pioneer 10, the Voyager spacecraft

provided a more complete picture of the "magnetodisc" which dominates

the physics of the middle magnetosphere of Jupiter. Here a plasma

sheet extends radially from roughly lo's orbit out to between 30 and

50 Rj. The sheet appears to be confined to the planetary dipole's

equatorial plane, although observations by the Voyager plasma instru-

ments showed that some deviation away from this plane toward the

rotational equator occurs at larger distances (Bridge et al., 1979a).








51

Scudder et al. (1981) have suggested that the Io plasma torus may be

its innermost boundary and the source of material for it. The plasma

sheet is apparently a permanent feature of the middle magnetosphere,

but large variations in the particle densities have been reported

between the Voyager 1 and Voyager 2 visits (Armstrong et al., 1981).

Beyond maintaining an overall enhanced plasma density, the heavy ions

are found to be enhanced with respect to protons in the sheet (McNutt

et al., 1981). Both electron and ion temperatures are observed to be

cooler than their counterparts away from the disc's plane. Since

protons are a major ion in the middle magnetosphere but play only a

minor role in the inner magnetosphere, Belcher (1983) suggests that

these H+ ions might arise from the Jovian ionosphere or from the

Galilean satellites. Other views on the icy satellites' material

contributions will be discussed.

The resulting magnetic field found in the middle magnetosphere

is the superposition of field lines emanating from Jupiter's internal

field and the field arising from the azimuthal flow of the plasma in

the sheet. The magnetic field of the plasma sheet has a vertical

component which is antiparallel to Jupiter's dipole field. Acuna et

al. (1983) find that this field is comparable in magnitude to the

planetary field at a distance of 15 Rj, and that beyond this point it

becomes radially outward above and radially inward below the sheet.

A model of the current sheet which closely fits many Pioneer and

Voyager observations has been put forth by Connerney et al. (1981).

It is based on a plasma ring extending from about 5 Rj to 50 Rj,

which is flattened in thickness to 5 Rj. At 30 Rj, the resultant








52

field magnitude is twice that expected for the planet's intrinsic

dipole alone, and it is almost entirely due to the radial component

of the current sheet.

Since the plasma composing this sheet is tied to Jupiter's di-

pole field (inclined 9?6 to the planet's rotational axis), it is

coincident with the magnetic equator and thus appears to wobble with

respect to an observer in the centrifugal plane. The icy satellites

therefore traverse a large portion of the thickness of the plasma

sheet during each Jovian rotation. Connerney et al. (1981) calculate

that the local field magnitude at Ganymede's orbit varies by a factor

of two during each rotation. These authors also point out that the

charged particle absorption features due to the sweeping effects of

the satellites, should appear at different locations than those pre-

dicted by use of the planet's internal field alone. This spatial

deviation will include a broadening of the absorption region in rad-

ial extent due to the radial component of the magnetodisc. This

phenomenon has been observed for Ganymede by several Voyager 2

instruments--magnetic field disturbances (Ness et al., 1979); hot

plasma fluctuations (Krimigis et al., 1979b); and plasma dropouts

(Bridge et al., 1979b). The signature of Europa was also identified

in Voyager 1 observations by the low energy charged particle experi-

ment (Krimigis et al., 1979a) and in proton absorption (Vogt et al.,

1979).
The Outer Magnetosphere

Although the outer regions of Jupiter's magnetosphere do not

concern the work presented in this dissertation, we present a brief

overview of the region, for the sake of completeness. Jupiter's








53

outer magnetosphere is that region of space where the cavity formed

by the planetary magnetic field--both the internal and plasma sheet

components--just balances the external pressure from the flow of the

solar wind. This boundary is called the "magnetopause," as it

represents the last set of magnetic field lines which form closed

loops from the northern to southern magnetic poles of the planet's

intrinsic field. Marked cylindrical asymmetry exists in the outer

magnetosphere between the sunward and anti-solar directions, with a

shock cone forming toward the Sun and an extended tail away from the

Sun, where the planetary magnetic field lines are swept downstream by

solar wind particles.

Although the magnetosphere may seem to be tenuous, it presents

itself as a relatively solid object to the flow of plasma comprising

the solar wind. Since the solar wind is most often supersonic, a

shock wave forms at the sub-solar point ahead of the magnetopause.

Hargreaves (1979) explains that a shock front is a discontinuity in a

medium, arising when information about an approaching perturbation is

not transmitted ahead into that medium. At this location, the solar

wind is substantially deflected around the magnetosphere. The region

between this "bow shock" and the magnetopause is usually called the

"magnetosheath." The magnetosheath will generally have properties of

its own, distinct from either the magnetosphere or the solar wind.

Zwickle et al. (1981) observed sporadic bursts of energetic ions dur-

ing the Voyagers' travel through interplanetary space. They con-

cluded that many of these particles originated in the Jovian magneto-

sphere and had simply leaked out through the magnetosheath. While a








54

similar phenomenon is observed for the Earth's magnetosphere, the

composition of the Jovian particles is enriched in heavy elements

compared to the solar wind. The composition of the terrestrial

magnetospheric ejections is very similar to the solar wind.

Apparently the Jovian magnetosphere compresses and expands in

response to variations in the solar wind plasma. Bridge et al.

(1979a, b) report a wide variation in the bow shock and magnetopause

crossing distances for the Voyager spacecraft, and Schardt and Goertz

(1983) place the magnetopause extremes between 46 Rj and 97 Rj.

Lepping et al. (1981) have reported large scale magnetic field struc-

ture in Jupiter's magnetosheath, some of which apparently displayed a

XIlI variation in field orientation. Although this phenomenon was
initially found in Voyager data, Acuna et al. (1983) state that it

was also present in Pioneer 10 data. No corresponding variations are

controlled by the Earth's rotation in the terrestrial magnetosheath.

Directed more or less away from the Sun and representing a re-

markable feature of the magnetosphere is the "magnetotail." The

tail, whose boundary with the solar wind is roughly cylindrical, is a

region where the planetary magnetic field lines are swept along by

the solar particles, where they accumulate on the down-wind side of

Jupiter. The plasma disc found on the sunward side is also present

on the night side, where the azimuthal current sheet continues out to

a distance of at least 100 Rj and is estimated to be about 5 Rj in

thickness (Behannon et al., 1981). As in the Earth's magnetotail,

Jupiter's magnetic field reverses direction across the northern and

southern lobes in the tail. Besides the escape of particles on the











sunward side, several researchers have reported that beyond -150 Rj,

on the night side of Jupiter, measurements of ion anisotropies sug-

gest a "magnetospheric wind" flowing away from the planet (Carbary et

al., 1981; Hamilton et al., 1981; Krimigis et al., 1981). In the

lobe regions above and below the plasma sheet in the night side mag-

netosphere, however, the ion composition is more like that of the solar

wind, 'suggesting a mixture of Jovian and solar plasma in this planetary

wind. As an example of the extent of the Jovian magnetotail, Desch

(1983) has found that it can extend over 5 A.U. to Saturn's orbit,

where it affects the emissions of Saturn's kilometric radiation.

Pioneer 10 observations in interplanetary space reported by

Chenette et al. (1974) revealed Jupiter to be the source of energetic

electron events. The spectral index of these electrons varied with

XIII and reaches a maximum for a subsolar XIII -2400. The power re-
quired to supply this Jovian "cosmic ray" electron source has been

estimated by Schardt and Goertz (1983) to fall in the range of 1013

to 1015 watts. The source location is believed to be the dawn sector

of the magnetopause (Schardt, 1983). Chenette (1983) suggests that

the Jovian MeV electron production is continuous, in contrast to the

impulsive electron source associated with solar activity.
Io

As the innermost major satellite in the Jovian magnetosphere, Io

is subjected to a significantly larger (v x B)-induced electric field

than the outer icy satellites. We have also seen that Io receives

the largest fraction of the energy from tidal dissipation in the

Galilean satellites' orbital resonance. Based on the conceptual








56

background developed in Chapter II and on the description of Jupiter's

magnetospheric environment presented in this chapter, we will discuss

the topics relevant to the electromagnetic interaction of the satel-

lite. We note that Tables III-1, and III-2 and III-3 are provided as

reference for the discussion of Io and the icy satellites. Information

on the orbital, the physical, and the environmental character of the

Galilean moons may be found in these tables.

It is widely believed that lo is the major source of material for

the Jovian magnetosphere (Belcher, 1983), supplying more than 1029

amu/sec toward the -1036 amu total mass of the Io torus. Although lo

is the most volcanically active body known, the transport of material

from the surface of the satellite into the magnetosphere presents some

problems in interpretation (Hill et al., 1983). The vulcanism yields

exhaust speeds of only -1 km/sec, compared to an escape velocity from

lo of -2.6 km/sec. Some material will certainly be stripped away from

a tenuous Ionian atmosphere by the impact of electrons on neutral atoms

and molecules which, when ionized, are carried off by the corotating

magnetic field. Another important mechanism for removal of material

appears to be the sputtering of volcanically deposited solid S and SO2

from Io's surface by charged particle impact. An ionosphere has been

measured for Io (Kliore et al., 1975); however, it is very tenuous (as

discussed in Chapter II). Cloutier et al. (1978) have pointed out the

difficulty in maintaining an ionosphere at lo. They argue that the

small gravitational field of the satellite is not sufficient to retain

even a tenuous structure against the ram pressure of the plasma coro-

tating with the Jovian magnetic field. They propose a model in which











TABLE III-1
THE GALILEAN SATELLITES


Io Europa Ganymede Callisto

Distance from
Jupiter (Rj) 5.9 9.4 14.99 26.33
(10 km) (420) (671) (1070) (1880)

Period (days) 1.769 3.551 7.155 16.689

Velocity relative to
magnetic field
(km/sec) 56.8 104 177 322

Time required to
move a diameter
relative to magnetic
field (sec) 66 30 30 15




















C,

-c c
*i4> o





o
S,







0
+-J
>,



























0
4-)



























L
0
-0





0)


r-

C





O








-4-
c- f r~
d us <-


d-


^f co














CC
**t
on 00 C\1 m


















CIO~

n 0
LL. C:) 0
r- C)J








i-i






Lf> 0













O~ -
cc 0C
o 1-4








cli









LOO







(00
CdO













0~ U
0o0 0












E



E

E


CT>>
-4d





















M 0
.'- to Cd




00 1^
*r















CM rn ) =

X: LU C
V)3U)
~ar ~acwE


CO 0
m CM
m C
c'..


LU


-LJ
I-




LU
l--
co







LU


IC
LU









I-

I-
o/
2;

>-

eC


u a
4-
UC

i -



C4
.- .

0 4-

0)

o W
.- C,
-( to

1- r


LO






0




o



S-
o


4-I




0 M
0u
L






tro
4- 0







4- --
4- -


0) 0
i+- -


o r- Co
CM 0




N O






Co:
o C 0C

Co


4-)






-0
Q.




s-

c
*e- W
I u



eW
3 4-
E 4-

to


S-












4-
4-
k3









-




a 41
O
0-0

01-


a





0 -m












2-
0
























*- m
*..
SCv
CO
0-



C-



Co

W -a



to







C.1


* 0

cnL-

to

0C


0-


-o
Wi-











TABLE III-3
THE GALILEAN SATELLITE ENVIRONMENT


Io Europa Ganymede Callisto

Electron density (cm-3) 3,000 50 5 0.1

Ion density (amu/cm3) 30,000 200 50 10

Magnetic field:
steady component (nT) 2,000 505 125 23
alternating com-
ponent (nT) 800 200 48 9

Alfven velocity
(km/sec) 356 1,080 688 689

Alfven Mach number 0.16 0.1 0.26 0.47

Pressure
(10- ergs/cm2) 1.8 0.31 0.02 0.002

Maximum distance above
magnetic equator (Rj) 1.02 1.63 2.55 4.55


NOTE: Electron densities from Gurnett et al. (1981). Ion densities
from Belcher (1983). Magnetic field values calculated from 100 tilted
dipole of strength 4.2 G R3 Pressures for the icy satellites are
from Krimigis et al. (1981', and are calculated as:

TOTAL = PRAM + THERMAL + MAGNETIC
= pV2 + nKT + B2/2vo.









60

material from Jupiter's ionosphere is transported to lo, where it is

held as the satellite's ionosphere.

In addition to tidal dissipation, another potential source of

energy within Io is electrical heating. Ness et al. (1979) reported

that Voyager measurements near the Io flux tube (the bundle of Jupi-

ter's magnetic field lines which pass through the satellite) indi-

cated an electric current intensity of 2 x 106 amperes, leading to a

power dissipation of more than 1012 watts. They pointed out that

this value is comparable to the power dissipated by interior heating

within Io due to tidal forces. In fact Gold (1979) has suggested

that Joule heating rather than tidal forces played a dominant role in

lo's vulcanism. Joule or ohmic heating occurs because current flow-

ing through a resistor causes a temperature increase in the resistor,

due to electron collisions with the resistive material. Gold found

that "hot spots" seen on lo's surface could be explained as a result

of the focusing of electric current onto localized areas of reduced

electrical resistivity. He predicted that the (v x B)-induced vul-

canism should be confined to the equatorial regions of lo, but that

the tidally-induced theory would display no preference.

The most thorough examination of electromagnetically induced

heating in Io is by Colburn (1980). For a range of possible interior

conductivities, he calculates the heat imparted by the TE mode of

induction within the satellite. The fundamental frequency in the TE

mode is the planet's rotation, due to the tilt of Jupiter's internal

dipole with respect to the planetary rotational axis. He concludes








61


that the largest heat flow due to the TE mode is 3 orders of magni-

tude less than that provided by the tidal heating model. The addi-

tion of a significant stagnation field, which could amplify the

driving field through conservation of magnetic flux, is not likely.

He suggests that the existence of an ionosphere with the satellite

does not provide significant shielding of the interior from TE mode.

The electrical skin depth of such an ionosphere greatly exceeds its

physical thickness. Colburn's description of the TM mode focuses on

the impedance match between lo's surface and its ionosphere. He

argues that most of the current in the flux tube circuit will be

effectively shunted through the ionosphere, and that heating of the

surface due to this source will be negligible. He concludes that the

conditions found at present within the Jupiter-Io electrical inter-

action do not provide significant heating to the planet's surface.

One of the first models of electrodynamic interaction between

Io and Jupiter was provided by Piddington and Drake (1968) and was

based on a (v x B)-induced electrical circuit. They estimated the

Cowling time for Io and the local Alfven velocity near the satellite

(which was later measured by the Voyagers and found to be several

orders of magnitude lower than their estimate). They suggested that

magnetic field lines could be "frozen" into the satellite, and that

an electrical current would flow along the "Io force tube (IFT)" and

generate the disturbances that led to the emission of DAM radiation.

(IFT is now often referred to as the "lo flux tube.") Goldreich and

Lynden-Bell (1969) expanded the (v x B) model and presented a formal

description of the lo-Jupiter interaction. They assumed infinite

conductivity along the IFT; however, they noted that if sufficient









slippage of the magnetic flux tubes prevented current closure

through the satellite, then their d.c. model would no longer work.

The DAM emission mechanism they proposed was the beaming of coherent

cyclotron radiation, arising from instabilities within groups of

electrons streaming along the IFT. They also suggested that an

Europa influence might be detected at frequencies below 9 MHz and an

Amalthea modulation up to 34 MHz. These frequencies correspond to

the electron gyrofrequency at the foot of each satellite's flux tube.

Since the discovery of the plasma torus located at lo's orbit,

new models of the lo-Jupiter interaction have been proposed. Neu-

bauer (1980), Southwood et al. (1980), and Goertz (1980) have dis-

cussed the transport of electrical current along magnetic field lines

by Alfven waves. Since the Alfven velocity in the satellite's near

vicinity is very low ( 300 km/sec) compared to the "infinite conduc-

tivity" models, the circuit caused by lo's motion relative to the

Jovian magnetic field is not electrically closed. By the time the

"magnetic wake" created by Io has made the round-trip along the mag-

netic field lines (including a reflection at the Jovian ionosphere),

Io has moved a distance greater than its own diameter. Closure of

the electrical circuit through the satellite in this case is not

possible.

Several researchers have considered another possibility that

may occur as a consequence of the torus and the Alfven disturbances

from Io--a standing Alfven wave system within the plasma torus (e.g.,

Gurnett and Goertz, 1981; Bagenal, 1983; Walker and Kivelson, 1981).

It appears that the plasma torus impedes the transmission of the

(v x B)-induced electrical current, and lo's relative motion insures









63

that closure through the satellite cannot take place. After reflec-

tion at the Jovian ionosphere, the returning Alfven wave will find

that Io has moved, and it will continue along the magnetic field

lines to a reflection at the opposite pole. Bagenal (1983) finds

that up to 25-30 bounces of an Alfven wave can take place per 13 hour

period of (x1II, YIo). It takes an Alfven wave many minutes to

travel from Io to the Jovian ionosphere, while the satellite moves

its own diameter in 66 seconds. The motion of Io with respect to the

Jovian magnetic equator will also produce secondary effects in the

Alfven wave pattern. As either magnetic pole rotates toward lo, the

satellite will reach its maximum magnetic latitude. When Io is near

a boundary of the torus, the oppositely traveling Alfven waves will

require different lengths of time to reach their first reflection at

the ionosphere. The model of Bagenal (1983) shows a systematic pat-

tern of spacing between the standing waves in the torus. Both Neu-

bauer (1980) and Bagenal (1983) suggest that this may be a mechanism

to explain the Io modulation of the DAM. However, close agreement is

not found between these models and the observed arc structure in the

Voyager PRA dynamic spectra of radio emissions.

If Io possess an intrinsic magnetic field, many of the details

of the preceding discussion would be altered. With a satellite mag-

netosphere, the (v x B) potential could be tapped across a signifi-

cantly larger boundary. In an early report on lo's control of the

DAM, Burns (1968) noted that an antiparallel orientation of lo's and

Jupiter's intrinsic fields could lead to more complex interaction, as

we discussed in the earlier chapter. Neubauer (1978) has modelled

strengths of intrinsic fields for the Galilean satellites, and he










calculates a magnetic intensity of 2200 nT at the surface of lo.

Kivelson et al. (1979) and Southwood et al. (1980) use this computed

magnetic moment to model the standoff distance of an Ionian magneto-

sphere against the plasma corotating with Jupiter. They balance the

thermal and ram pressures of the Jovian magnetospheric plasma against

the magnetic pressure of lo's internal field, and they find agreement

with observations of the Voyager fields and particles experiments.
The Icy Satellites

The remaining three Galilean satellites are often called "icy"

satellites (as are Saturn's moons), although Europa's density of >3

gm/cm3 actually classifies it as a "rocky" body like Io and the

terrestrial Moon. Europa is completely covered by a layer of water

ice though, hence it is reasonable to include it with Ganymede and

Callisto in a discussion of the satellites' electromagnetic inter-

action with the Jovian magnetosphere. Densities for Ganymede and

Callisto are -1.9 and 1.8 gm/cm3, and Morrison (1982) suggest that a

40% rock and 60% water ice (or liquid) composition matches these

densities. The abundance of water in these satellites permits the

rapid transport of heat at much lower temperatures than those usually

found in the interiors of rocky bodies. The cooling of an initially

heated sphere depends on the radius (R) of the body and the thermal

diffusivity (K) in the relationship
Cooling Time (R2/K)

A sphere of silicate composition requires an order of magnitude more

time to cool than does an equally sized sphere of ice (Cole, 1984).

Inter alia, Cassen et al. (1982) and Consolmagno (1983) have

considered the possible sources of heat for icy satellites. They








65

include energy of accretion and continuing impacts; tidal dissipa-

tion; heating from internal radioactive nuclides (e.g., 40K,
232Th, 235U, 238U); and electrical or Joule heating. Generally, we

may presume that the tidal source could be particularly significant

for resonant satellites and those in close proximity to a major

planet. The tidal dissipation within Europa and Ganymede was

described earlier in this chapter, and it may be taken to be at least

an order of magnitude less than lo's. Since Europa has a much larger

fraction of silicates, we might also expect a larger internal heat

flux due to radioactivity. Although Cassen et al. (1982) do not

consider electromagnetic heating important, they suggest that it may

not be dismissed completely. Observations and models related to the

electromagnetic induction process will be discussed in this section.

Europa's surface covering of ice is at least several kilometers

thick and, from analysis of crustal features, there is evidence that

its depth is probably several 10's of kilometers (Lucchitta and

Soderblom, 1982). Compositional models of the satellite suggest an

H20 mass fraction of 5% to 10%. Wu et al. (1978) reported that the

Pioneer 10 ultraviolet photometer detected emission from a cloud

centered on Europa's location, but identification of the source

species was ambiguous. Eviatar et al. (1981) suggested, however,

that this detection may have been contaminated by emission from

the Io torus. In support of the identification of the cloud with

Europa, Intrilligator and Miller (1982) have reported than an exam-

ination of the 1973 Pioneer 10 plasma analyzer observations indicated

density enhancements near the predicted location of the satellite's

L-shell. Europa's signature was only seen in absorption by the low








66

energy charged particle observations of Voyager (Vogt et al., 1979),

suggesting that identification of local plasma enhancements remains

speculative.

If a region of enhanced plasma density existed near any of the

icy satellites, then the possibility of significant electromagnetic

interaction for that body would be greatly increased. Material for a

cloud, torus, or satellite atmosphere could come from external

sources (e.g., Io or the Jovian ionosphere), or it could be removed

from the icy body's surface. The identification of an logenic source

(contributing Na or SO2 products) should be distinct from the local

or Jovian ionosphere source (where H and H20 products would

dominate). In considering the local source, we must hypothesize a

mechanism (or mechanisms) for release and transport of material away

from the satellite's surface. Two methods are often considered--

sublimation and sputtering. The contribution of water products to

the Jovian and Saturnian magnetospheres due to sputtering of icy sur-

faces by charged particles has been discussed extensively (e.g.,

Lanzerotti et al., 1978; Johnson et al., 1981; Johnson et al., 1983).

It appears that sputtered water products from the icy surfaces could

form tenuous atmospheres or coronae (Johnson et al., 1983), however

sufficient material to form a co-orbiting torus does not seem

likely.

If an atmosphere produced by sputtering or by sublimation were

significantly dense, then the surface of the satellite could be pro-

tected from impacting micrometeorites and plasma. The H20 vapor

pressure due to sublimation has been calculated by Wolff and Mendis

(1983). Their model yields pressures for the icy satellites'








67

atmospheres which are comparable to those found for the Moon--10-14

bar (Pollack and Yung, 1980). Kumar and Hunten (1982) characterize

other models for the icy satellites and suggest that significant oxy-

gen (02) atmospheres could arise from sputtering of H20. A ground-

based stellar occultation experiment by Carlson et al. (1973) meas-

ured an atmosphere at Ganymede of 1 microbar surface pressure; how-

ever, a similar experiment by the Voyager 1 ultraviolet spectrometer

placed an upper limit on the existence of Ganymede's atmosphere of

-10-10 bar (Broadfoot et al., 1979). One observatonal consequence of

an atmosphere on an icy satellite has been described by Wolff and

Mendis (1983). Their model of the solar-induced pressure gradients

results in convection which transports H20 away from the sub-solar

point, with a net migration of material toward the poles. Polar

shrouds have been identified in Voyager images of Ganymede (e.g.,

Shoemaker et al., 1982). With regard to our discussion, a more

important consequence of atmospheres on the icy satellites would be

the possibility of ionospheres.

Ionospheres have not been detected for Europa, Ganymede, or

Callisto--no radio occultation measurements were performed by Pioneer

or Voyager spacecraft for these satellites, however such an experi-

ment may be possible with the Galileo probe to Jupiter in the late

1980's. Theoretical arguments concerning the existence of tenuous

ionospheres for the icy Galilean satellites have been discussed by

Wolff and Mendis (1983). Since these satellites travel in and out of

the previously described plasma sheet, a "pumping" of their hypothe-

sized ionospheres occurs, similar to that observed at Venus in re-

sponse to solar wind variations. Their calculations show that only








68

Callisto is likely to have an ionosphere substantial enough to with-

stand the ran pressure of the Jovian plasma sheet, and that any tenu-

ous ionosphere for the other satellites would probably be swept away

(as described in the section on Io by Cloutier et al., 1978).

We should also consider Io as a source of material for the sur-

roundings of the icy satellites. Because material originating on Io

should be readily identifiable, contamination of satellites' surfaces

can yield information on transport processes in the magnetosphere and

on the existence of atmospheres for these other satellites. Color

asymmetries between the leading and trailing hemispheres of the icy

satellites have been known for many years (e.g., Wolff and Mendis,

1983). One clear case for leading-trailing asymmetry is reported by

Lane et al. (1981). Observations made with the earth-orbiting IUE

satellite indicated an ultraviolet absorption feature only on the

trailing side of Europa which they attributed to implantation of SO2

molecules. However, the calculated column density of the molecule

was many orders of magnitude less than that predicted if all of the

logenic material had remained on Europa's surface. Eviatar et al.

(1981) interpreted this observation as evidence for an equilibrium

between the implantation of sulfur dioxide and the preferential

removal of water by sputtering. A recent re-examination of the

problem (Eviatar et al., 1985) leads to a model which balances the

effects of implantation, sputtering, and redistribution of escaping

water on Europa's surface.

From the preceding discussion, it does not appear that substan-

tial atmospheres, ionospheres, or plasma tori exist for the icy

Galiliean satellites. Significant interaction due to electromagnetic








69

induction could still occur if these bodies possessed intrinsic mag-

netic fields. We refer again to the dynamo models of Neubauer (1978)

which are tabulated in Table 111-2. The large fraction of ice in

Ganymede and Callisto leads to very small magnetic moments for these

satellites; however, the diminished plasma density compared to the

inner magnetosphere can result in relatively large magnetospheres

(Kivelson et al., 1979). Wolff and Mendis (1983) suggest that excur-

sions of Callisto above and below the plasma sheet (cf., Table 111-3)

would lead to a more asymmetric magnetosphere for that satellite.

Hypothesized Europa and Ganymede magnetospheres would be more sym-

metric, since the total external pressure (ram, thermal, and magnet-

ic) varies less for these satellites as they travel through the

plasma sheet.

In the absence of intrinsic magnetic fields and material sur-

rounding the icy satellites, we should consider the possibility of EM

interaction with the satellites' surfaces. The only detailed work to

date on this topic has been by Colburn and Reynolds (1985). They

investigate the possibility of electrolytic currents within Europa's

surface layers as an energy source for life forms on that body

(Reynolds et al., 1983). The conductivity of ice (or water) is

orders of magnitude larger than that of solid rock (discussed in

Chapter II), and the addition of dissolved salts further increases

the conductivity of liquid water. Colburn and Reynolds (1985) find

that the TE mode supplies negligible heating to the satellite, even

if it could be confined to a layer of liquid water between the icy

surface and the silicate interior core (hypothesized by Cassen et

al., 1979). In discussing the effects of TM induction, the authors









70

assume that Europa does not possess an ionosphere which could shunt

the induced electrical current (as occurs at lo). If the resulting

current can be confined to the linear cracks seen covering Europa's

icy surface, and if suitable contact can be made to the magneto-

spheric plasma surrounding the satellite, then significant electro-

lysis of water could occur. They calculate that this "battery" could

account for a total electrolysis of 7 x 105 kg/year, with an equiv-

alent energy storage of 1013 joules/year. This model is critically

dependent on the completion of the electrical circuit to the Jovian

ionosphere through the magnetic flux tubes. Measurement of electric

current perturbations to the ambient magnetic field (similar to those

reported by Ness et al. 1979, for lo's flux tube) may be possible

with the Galileo probe. Discussion of the Alfven interaction of each

of the Galilean satellites was included by Neubauer (1980), who

concluded that Io produced the only significant disturbance.

Other observations of Ganymede require mention. Features at-

tributed to that satellite were seen in a variety of Voyager instru-

ments--magnetic field disturbances (Ness et al., 1979); hot plasma

fluctuations (Krimigis et al., 1979b); and plasma dropouts (Bridge et

al., 1979b). Tariq et al. (1983, 1985) have modelled the wake produced

by Ganymede's travel through the magnetospheric plasma. One further

phenomenon associated with the icy Galilean satellites' electromag-

netic nature should be noted. As shown in Table III-2, the radar

reflectivities of these bodies are anomalously large when compared to

lo, the Moon, or the inner planets (Campbell et al., 1977). Moreover,

the circular polarization of the reflected signal is in the same








71

sense as the transmitted signal, while a returned signal of the oppo-

site polarization sense or an unpolarized signal was expected. This

behavior must be associated with the large fraction of ice at the

surface of these bodies, and several models of sub-surface reflecting

facets have been put forth (Ostro, 1982).
Amalthea

The small, irregularly shaped Jovian moon Amalthea (J V) orbits

at 2.55 Rj, and has a period of almost 12 hours. Its figure is not

well-fit by a triaxial ellipsoid, but its volume is equivalent to a

sphere of radius 83 km (Veverka et al., 1981). It is comparable in

size to the Trojan asteroid Hektor; however, Amalthea is colored dark

red, with localized bright markings. The trailing side of the

satellite is actually redder than the leading face. This has been

interpreted as evidence of logenic material (probably sulfur)

diffusing inward from the Io plasma torus. All charged particle

experiments on the Pioneer 10 and 11 spacecraft detected absorption

by Amalthea (Fillius, 1976); the Voyager spacecraft closest

encounters were outside of Amalthea's orbit.

The origin of Amalthea remains a mystery--it may have been

formed in place through accretion, or it could be a captured aster-

oid. Veverka et al. (1981) note that the masking of its surface by

sulfur makes determination of its origin difficult. Although it is

comparable in size and shape to the low density, carbon-rich Trojan

asteroids, a measurement of Amalthea's mass would be necessary for

positive identification. If its density is found to be less than

about 2.5 gm/cm3, the implication would be that it is a captured

asteroid. A higher density (> 3 gm/cm 3) would suggest that it








72

originated in place. Morrison and Burns (1976) note that if Amalthea

were formed in place, it may be very rich in refractory material due

to the intense heating early in the Jovian proto-planet condensa-

tion.

The disk-averaged surface temperature of the satellite is anom-

alously high--1800 K--while solar heating models suggest that the

maximum temperature should be -160 K (Thomas and Veverka, 1982).

Charged particles and Joule heating may provide additional energy be-

yond the solar radiation, but the electrical potential difference

across Amalthea is not very large (only half that of Callisto) due to

its small velocity relative to the Jovian magnetic field (Table III-

4). The estimated cratering rate is many times larger than that

found for the outer Galilean satellites or for the Moon, so the ener-

gy due to impacts and accretion should probably be considered in the

total budget. Tidal forces appear to have made Amalthea's rotation

synchronous with its orbit, with the long axis of the satellite ori-

ented radially with respect to Jupiter.

Thomas and Veverka (1982) discuss the factors affecting the

surface of Amalthea. They suggest that intense bombardment by micro-

meteorites and heavy ions could produce significant amounts of molten

material (glass) in the ejected regolith. This lo-contaminated,

glass-rich regolith could provide an abundance of loose material,

perhaps as deep as several hundred meters. If Amalthea is a captured

asteroid, rich in carbonaceous material, then we might consider the

suggestion of high electrical conductivities in asteroids by Ip and

Herbert (1983). They discuss the possibility of large electric cur-

rents in the metal-rich asteroids, induced by the (vXB) interaction

of the solar wind. In view of the rich plasma environment and












TABLE 111-4
AMALTHEA


Distance from Jupiter

Synodic Period

Dimensions

Velocity relative to Jupiter

Surrounding electron density

Magnetic field (steady component)

Magnetic field (alternating component)

Alfven velocity

Alfven Mach number

Estimated escape velocity

Maximum potential difference

Time required to move a width

Albedo (optical)


2.55

0.489

270 x 165 x 150

4.9

20

23,800

9,800

38,700

0.00013

100

23

34

0.06


Rj

days

km

km/sec

cm-3

nT

nT

km/sec



m/sec

kV

sec


NOTE: Magnetic field strengths as in Colburn (1980). Electron
density and Alfven velocity from Neubauer (1980). Albedo and escape
velocity from Veverka et al. (1981).








74

intense magnetic field found at Amalthea's orbit within the Jovian

magnetosphere, we might expect a sizeable electric current flow with-

in the satellite. Neubauer (1980) includes Amalthea in his models of

Alfven wave generation due to the (v x B) induction, but he concludes

that no significant Alfven interaction will arise from this satel-

lite.

Since lo's electrical influence is seen in the decameter radio

emissions, it would seem reasonable to search these observations for

the signature of Amalthea. Unfortunately, the orbital phase of

Amalthea is nearly equal to one half the observing period for ground-

based telescopes--the ratio of the terrestrial period (23.9344 hours)

to that of Amalthea (11.9577 hours) is (2.0016). This near-resonance

makes unambiguous detection of any modulation by Amalthea virtually

impossible from the groundbased observations. We are aware of only

one examination of groundbased observations of the DAM for Amalthea's

influence to date. Register (1968) found no modulation of the 22.2

MHz emissions in a single apparition (1967) by the satellite's phase.

Continuous monitoring of the emissions by spacecraft could lead to a

definitive answer by eliminating the observing limitation.
Discussion

We have now reviewed theories and observations concerned with

the electrodynamic interaction of the major satellites with Jupiter's

magnetosphere. Following Colburn (1980) and Colburn and Reynolds

(1985), we show in Table III-5 the magnitude of induced electric

fields for each satellite. Both the TE and TM modes are calculated,

and each mode has been normalized to lo's value for ease in












TABLE 111-5
INDUCED ELECTRIC FIELD INTENSITIES FOR THE MAJOR SATELLITES


Io Europa Ganymede Callisto Amalthea


TE 1 0.25 0.11 0.02 0.2

TM 1 0.46 0.18 0.09 1.03



NOTE: This table shows the ratio of each satellite's induced electric
field intensity compared to that of Io.

ETM = wrBTE/2 ETM = (v x BTM)

where BTM is the steady component of the magnetic field

BTE is the alternating component of the magnetic field.

For Io: ETE = 9.81 x 10-5 V/m and ETM = 0.114 V/m.








76

comparison. The time-varying portion of Jupiter's dipole field is

due primarily to the 100 tilt of the magnetic axis with respect to

the rotational axis. This results in a component of the field in the

equatorial plane which varies as exp (-iwt) for each satellite. Con-

ductivity (a) considerations are not invoked here--the numbers in

Table 111-5 represent the maximum electric field (E) which could be

available in modelling the heat flow due to TE induction within each

satellite. A first-order determination of the electrical current (I)

generated within a satellite would be given by Ohm's law (I = oE).

In practice, one solves a set of coupled differential equations for

each radial shell inside the satellite and determines the heat input

at each boundary. For the TM mode, we have seen that electrical

heating will depend on the conducting path in the immediate vicinity

of the satellite's surface, if current flow takes place at all.

TM induction appears to be confirmed for Io. The existence of

an ionosphere and the measurement of current flow within the Io flux

tube provide at least a conceptual basis upon which one can explain

the satellite's modulation of electromagnetic emissions. Intense

heating due to the tidal dissipation seems a likely cause for the

vulcanism, which supplies abundant material to the Jovian magneto-

sphere. Electrical heating, as envisioned by Gold (1979), is

considered negligible within the satellite due to the shunting effect

of the satellite's ionosphere. The existence of the Io plasma torus

complicates the "closed-circuit" models of induction, since the

Alfven disturbance generated by Io's motion with respect to the

magnetic field will not close back on the satellite itself.







77

At Europa, the possibility of a layer of liquid water near the

icy surface seems likely in view of tidal and radiogenic heating.

The existence of liquid water would impact both the TE and TM modes

of induction. A subsurface ocean would effectively shield the in-

terior of Europa from TE induction, however this still does not pro-

vide substantial heating of the satellite's interior. The existence

of a water ocean near the surface could allow significant TM inter-

action though, if escape of the vapor were concentrated to the linear

cracks seen in the icy surface. This mechanism can provide up to

3 x 105 watts of power (Colburn and Reynolds, 1985), however it is

not clear how this could be used to modulate any of Jupiter's elec-

tromagnetic emissions. This value is several orders of magnitude

smaller than either the integrated DAM component (assuming the

radiation is beamed) or the peak power radiated by bKOM (e.g., Carr

et al., 1983). It is therefore doubtful that a search of the

emissions for Europa's signature would prove positive. However, even

a negative detection would confirm the discussion of the past

chapters and show that a qualitatively different interaction occurs

for Europa than that for Io.

The absence of tidal heating and the decreased fraction of

radioactive nuclides within Ganymede and Callisto make the existence

of liquid water a remote possibility. Ganymede has been tectonically

active in the past, evidenced by the grooved terrain seen over its

surface; Callisto's surface is near saturation with impact craters.

The difference in appearance of the two satellites is explained by a

larger fraction of radioactive materials in Ganymede (e.g., Cassen et

al., 1982); however, liquid water within either satellite is not








78

expected. The absence of measured atmospheres lessens the possibil-

ity of significant TM interaction, since the icy surface layers do

not provide a good conducting path. As seen in Table 111-5, both TE

and TM modes of induction are significantly reduced from that of Io,

hence we expect no modulation of electromagnetic emissions by either

satellite.

The electromagnetic interaction of Amalthea with the Jovian

magnetosphere remains speculative. Since we do not have a reliable

estimate for the satellite's composition, we are not able to model

its interior conductivity. The contamination of Amalthea's surface

by logenic material could result in an electrically insulating layer,

since the conductivity of sulfur is relatively low. The dense plasma

of Amalthea's environment probably leads to a situation similar to

that at lo, where completion of the electrical circuit through the

satellite is not possible. TM mode induction of the abundant rego-

lith could lead to electrostatic effects on the satellite (e.g.,

Thomas and Veverka, 1982), however no mechanism is known for transfer

of this energy to the planet's electromagnetic emissions. In gener-

al, electrostatic forces are comparable to that of gravity for parti-

cles of micron size and smaller. Electrostatic levitation of charged

dust grains from the surface of Amalthea could be expected (e.g., Grun

et al., 1984), particularly since the TM mode electric field inten-

sity for that satellite is equivalent to lo's (see Table III-5).

Amalthea's signature has been searched for in the groundbased DAM

(Register, 1968; Kaiser and Alexander, 1973; Thieman, 1979), but no

effects have been reported. We include Amalthea in the first search

of the lower frequency Jovian KOM emissions for satellite signatures

(Chapter V).













CHAPTER IV
THE 11.9 YEAR PERIODICITIES

As will be shown in Chapter V, in order to ascertain the role

of the Galilean satellites in the decameter emissions one must have a

sufficient body of monitoring to defeat the effect of the orbital

resonance of these moons. We must also understand any other long

term effects in the database which might influence the study of the

satellites. This chapter presents such an investigation.

While analyzing the groundbased DAM observations, we have dis-

covered a previously unreported effect of the 11.9 year period of

Jupiter. The effect is subtle; however it can be verified by obser-

vations from several different locations. In this chapter we first

describe the groundbased DAM observations used in the study. We then

describe the long term effects which are known to be present in these

data. The dominant 11.9 year effect found in the observations is

called the "DE effect"--where DE refers to the declination of the

Earth, as seen from a Jovicentric coordinate system. Changes in both

the DAM detection probability and the polarization characteristics of

the radio bursts due to DE are discussed in this chapter. Special

emphasis is placed on the newly discovered secondary DE effect.

Because spacecraft observations at higher Jovicentric latitudes are

now available, we compare this new finding with relevant Voyager data.
Observations

A rich archival database exists for synoptic groundbased detec-

tion of the Jovian decameter noise. The University of Florida Radio










80

Observatory has monitored many different portions of Jupiter's

decametric spectrum with a variety of instruments over the past 25

years. Four of the discrete observation frequencies span a signifi-

cant fraction of this period, and three are paralleled by observa-

tions from the Observatorio Radioastronomico of the Universidad de

Chile at Maipu, Chile. These are 18.0 MHz, 22.2 MHz, and 27.6 MHz.

The fourth frequency monitored by Florida and used in this study is

15.0 MHz. Observations through the 1974 apparition were compiled by

Thieman (1977); more recent data from UFRO and Chile have been added

to his work. In addition to the Florida/Maipu observations, we have

included data from the University of Colorado. Decametric emissions

from the Sun and Jupiter were monitored there with a sweep-frequency

interferometer each year from 1960 until 1975 (Lee and Warwick, 1964;

Warwick et al., 1975).

The location and instrumentation of these three observatories

are complementary--Northern versus Southern Hemisphere, and discrete

frequencies versus a continuous spectrum. Table IV-1 outlines the

gross statistics of the observations. The discrete frequency

observations from Florida have been combined with the matching data

from Chile to form three larger datasets (at 18.0, 22.2, and 27.6

MHz). Observing methods at each station are such that no new

information would be available by combining unlike frequencies, hence

the spectral information in the observations has been preserved. We

note that the data used in this study has been gathered by low-gain











TABLE IV-1
OBSERVATIONS USED IN


THIS STUDY


Frequency Total Hours' Total Hours' Number of
(MHz) Observation Activity Apparitions


Florida 15.0 9,817 867 16 (1961-81)
18.0 19,090 1482 23 (1957-81)
22.2 20,236 899 22 (1958-81)
27.6 14,202 127 15 (1958-73)

Chile 18.0 12,357 513 14 (1960-74)
22.2 11,950 253 12 (1960-74)
27.6 11,212 86 12 (1961-74)

Colorado 7.6-41 32,113 2285 14 (1960-75)


NOTE: Summary of the observations used in this study. The Colorado
instrument was a radio spectrograph.











antennas. Desch et al. (1975) showed that lo's control of the

radiation is a function of intensity and that earlier conclusions

related to lo's frequency control were biased by a selection effect.

We include this remark because it is important for the reader to

understand that at least one component of the DAM (the non-lo-related

Source B emission) is not detected by low-gain antennas. An investi-

gation of satellite effects in this less intense DAM component would

require long term monitoring with a more sensitive instrument. In

Chapter VII we discuss this possibility further.

The Florida/Chile observations were obtained with steerable

Yagi antennas (Register, 1968; Thieman, 1974). The observing process

is conducted similarly at both observatories. The output of each

antenna/receiver is connected to both a paper chart recorder and a

loudspeaker. Identification of Jupiter activity is made aurally by

the observer on duty, whose responsibility also includes annotating

the paper chart for future reduction. Times of monitored quiescence

and activity are later tabulated, to be further reduced in terms of

Jovian longitude and satellite geometry.

The University of Colorado radio spectrograph was operated in

the swept-frequency mode from 15 41 MHz during 1960; from 7.6 41

MHz during 1961 through 1968; and 7.6 80 MHz during 1968 through

1975. The antennas were two trihedral corner reflectors, each with

an effective area of about 500 m2. (For comparison, the effective

area of a Florida five element Yagi operating at 18.0 MHz is about

205 m2.) During the day, the Colorado instrument tracked the Sun;

most of the observations of Jupiter were conducted at night. How-

ever, within 30 days of conjunction both Jupiter and the Sun were









83

in the main antenna beam. Also, within a few months of conjunction,

Jovian activity could be detected when the planet was in a side lobe

of the antenna pattern. After the 1964 discovery of the Io effect,

special observing sessions were instituted at times of propitious

(III, Io) geometry.
The Solar Influence

In Chapter I we mentioned that known (and perhaps unknown) long

term effects influence the groundbased DAM observations. Such effects

are evident from apparition-to-apparition, in contrast to the period-

icities which are hours (Jovian rotation), or days (Io's period), or

even weeks in duration (Galilean satellite resonance). At present,

there are two of these long term modulations which are recognized--

the solar activity cycle and the Jovicentric latitude of the ob-

server. The phasing of both is shown in Figure 4.1

Groundbased observations of decametric phenomena are usually

limited to nighttime sessions. This is because the electron density

of the terrestrial ionosphere increases during the daylight period.

The electron density has both a diurnal period (increased density

during local daytime) and a solar cycle period (increased density

even at night during sunspot maximum). The enhanced ionospheric

electron density affects the decameter observations in two ways.

First, man-made radio signals and naturally occurring terrestrial

radio noise (from thunderstorms) can not escape into space, since

they can be reflected multiple times between the surface of the Earth
and the terrestrial ionosphere. For a groundbased radio telescope,

this rise in the level of background noise makes unambiguous detec-

tion of Jovian radio bursts difficult or impossible. The effect is

worse for the lower decametric frequencies (_18 MHz and below)
































Figure 4.1. Solar activity and DE. Graphs showing: (a) the solar
activity cycle (the relative mean sunspot numbers plotted monthly)
as compiled by the World Data Center in Boulder, Colorado; and (b)
monthly values of the Jovicentric declination of the Earth (DF)
from the 1957-1980 editions of American Ephemeris and NauticaT
Almanac.


1



























































0 C C
o N
i0 0 0cS dNns
C" 0 (sshOCT) 3 gn9 C]

d~amN IOcSNFiS








86

than for the higher frequencies (e.g., above .22 MHz). The second

effect of the ionospheric electron enhancement is that Jovian (or

other cosmic) signals below a certain frequency may not be able to

penetrate the terrestrial ionosphere. For a given ionospheric

electron density, both effects can occur for radio waves below a

certain frequency (depending on the wave's angle of incidence from

above or below the terrestrial ionosphere).

Effective groundbased decameter observations of Jupiter can be

made only at times when the terrestrial ionosphere's maximum electron

density is low. This modulation of the annual monitoring, which has

the period of the solar cycle, is shown in Figure 4.2. This figure

may be compared to Figure 4.1, which showed the monthly values for

sunspot number over the two decades of interest. (Sunspot number is

only one indicator of solar activity. The occurrence of spots on the

Sun is not believed to be the direct cause of enhanced electron

density in the Earth's ionosphere, but it is correlated with the

enhanced levels of untraviolet radiation emitted from solar active

regions.) Two things should be noted in Figure 4.2. The 18.0 MHz

observations include both Florida and Chilean observations between

1960 and 1974. This explains why the observing peak in 1965 is

significantly larger than that in 1976. The second remark concerns

the Colorado observations shown in the same figure. Recall that

the Colorado instrument was a spectrograph covering 7.6 MHz to over

40 MHz. Monitoring at ANY frequency by the Colorado instrument has

been included in Figure 4.2, so we remind the reader that detection

of the Jovian bursts at the lower decametric frequencies (below -18

MHz) was not possible during years of increased solar activity.






























Figure 4.2. Observing time for Florida/Chile 18.0 MHz and
Colorado. Plots of observing time versus year for Florida/Chile
18.0 MHz and Colorado datasets (all data are included). The effect
of the solar activity cycle on the state of the terrestrial
ionosphere is apparent. When solar activity is high, fewer
observations can be made at the decameter frequencies (compare to
Figure 4.1).






















co
=



=_1


58 68 78


Year


=
'/4


,-1


58 68 78


Year


Colorado


I I I I I I I


I t