Long-Term Periodicities in The Jovian Decametric Emission

LONG-TERM PERIODICITIES
IN
THE JOVIAN DECAMETRIC EMISSION
By
LEONARD NORMAN GARCIA
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
1996
Ad Majorem
y para
Dei Gloriam
las Tias
ACKNOWLEDGMENTS
There are so many people that have entered
have directly or indirectly contributed to
them enough.
want
in
to
my
thank
this
my
work
life,
that
have enriched
could not
so greatly and
it
them
list
all
or thank
have been blessed with good friends, good family and good instructors.
my
Thomas D.
advisor, Dr.
knowledge of
the field
my
and
Carr, for his patience with
skills as a researcher.
work with someone who was so instrumental

radio astronomy and
such a
is still
Jupiter's radio emissions.
vital
would
in the
It
me
was
as
truly
development of the
field
grew slowly
an honor to
of planetary
and respected influence on our understanding of
like to
thank the other members of
my
committee,
Humberto Campins, Charles Hooper, George Lebo and Alex G. Smith. This work
Drs.
has benefited greatly from their helpful suggestions and recommendations.
want
Levy of
my warmest
to express
the
UF
thanks to Francisco Reyes,
planetary radio group.
the observatory, building antennas or
occasionally had
its
spent in his office
something
moments. To Francisco
when
had written.
and
his limitless
think
will
want
to
to
Jorge
miss the times spent out
the lawn even the
express
could not understand a paper or
want
my
I
all
thanks for the hours
needed his opinion on
found
real support
from
his
encouragement
enthusiasm for the subject. Dr. James R. Kennedy was very helpful
make some sense of
possible solar effects on the Jovian magnetosphere.
advice and encouragement are greatly appreciated.
had the opportunity
to
work with
at
at
night observing
thank Kazumasa Imai for his helpful suggestions
during the time he spent with us in Florida.
trying to
mowing
Wes Greenman and
NASA/ Goddard
iii
am
very thankful for the people
Space Flight Center.
wish
to
in
His
I
thank
James Green, James Thieman, Shing Fung and Robert Candey of the National Space
Drs.
Science Data Center for the opportunity to consult with them on
my
their interest in
work.
but
in this dissertation
help
at
me
am
that all of those
who gave
meet
I
life
in this
has been enriched by
department and
can not be certain that
education without them.

radio group and
in the
around ideas, when
NASA
Graduate Student Researcher's
all
in this dissertation.
of the graduate students
hope they
will forgive
me
if I
good
friends, for
being there for
was stumped on a problem or when
My
by the work of the astronomy department
me
to
throughout
writing
have
my
my
who
family
helped
me
just
in the
lifted
my
spirit
know during
love, Sarah.
"free-time."
me when
and inspired
my
colleagues
needed
needed
to toss
to complain.
astronomy department
Ann
office staff
Elton,
Debra
through a never ending bureaucratic maze.
who have always been
educational career.
time
me when
a tremendous source of support to
the last
few months of
had the prayers of many people from around the country.
for the blessings
is
my
thank
to
of them here.
to withstand the strain of a graduate
would have been able
was made
want
list all
must thank Chuck Higgins and Liyun Wang,
great officemates and friends.
had the opportunity
my
Hunter and Glenda Smith
do not
They have been

easier
of their time to
wish to acknowledge the National Solar Observatory/Sacramento
Peak for the monthly average sunspot numbers used
My
much
so
to thank the Office of University Programs
wish
for their financial support through the
Program fellowship.
research and for
of course solely responsible for the conclusions presented
do hope
be disappointed.
will not
Goddard
my
was uncertain of myself.
dissertation
know
these prayers
am
very thankful
have received. Foremost among those whose prayers have touched me
She has shared
She has been there
in
for
my
difficulties
me.
iv
and suffered through
my
guilt-ridden
TABLE OF CONTENTS
page
ACKNOWLEDGMENTS
LIST
OF TABLES
LIST
OF FIGURES
iii
vii
ix
ABSTRACT
xvi
CHAPTERS
1.
INTRODUCTION
1
.2
1.3
3.
Polarization of
DAM
Frequency Dependence of Jovian Radio Emissions
DE
.4
The Jovian Emission Beam Model and
.5
Observing Jovian Decametric Radiation
11
Discussion of the Problem
12
1.6
2.
The Discovery of Jovian Decametric Radio Emissions
the
Effect
METHODS
14
2.1
Data Collection
14
2.2
Data Reduction
15
2.2.1
Standard Format
15
2.2.2
Occurrence Probability
16
2.2.3
Weekend
2.2.4
Elongation Effect
20
2.2.5
Galactic Background and Distance Effects
24
2.2.6
Calibration Procedure
34
versus
Weekday Observers
18
RESULTS
3.1
3.2
38
A New
Determination of
DAM
Source Locations
in
CML
and Io
Phase
38
Occurrence Probability and Average Flux Density
38
3.2.1
Source Widths and Locations
44
3.3
Verification of the
3.4
The
3.5
The
3.5.1
3.6
4.
DE
DE
Low
Effect on the
Intensity
High Probability Non-Io-B Source
47
54
Non-Io-A Source
Effect on the Other
DAM
59
59
68
76
Sources
Io-B
3.5.2
Io-A and Io-A'
3.5.3
Io-C and Non-Io-C
3.5.4
Summary
81
Models of the Non-Io-A Occurrence Probability versus Time

Model I: Based on Assumed Variation of OP with De
3.6.1
3.6.2
Model 2a: OP Varies with both D E and Sunspot Number ...
3.6.3
Model 2b: Same as Model 2a Except for Assumed Decrease in
OP at High D E
3.6.4 Model 3: OP is Proportional to Source Width
3.6.5
Is There a Solar Effect on DAM?
SUMMARY AND CONCLUSIONS
83
83
86
88
91
94
98
APPENDICES
A.
A SAMPLE CALIBRATION OF THE YAGI ANTENNA SYSTEM ....
105
B.
TESTS FOR D E DEPENDENCE
109
REFERENCES
112
BIOGRAPHICAL SKETCH
117
vi
LIST
OF TABLES
page
Table
DAM
sources
1.1:
Various characteristics of the
2.1:
Seasonal averaged galactic background temperatures for
20 and 22
18,
MHzYagi
2.2:
30
Seasonal averaged occurrence probabilities for 18
MHz
Yagi
31
1957-1975
2.3:
Seasonal averaged occurrence probabilities for 18
MHz
Yagi
32
1975-1994
2.4:
The procedures used
to obtain galactic
background-limited occurrence
34
probabilities
3.1:
Source locations
in
These ranges span
CML
and Io phase taken
the sources at
at the
8 through 22
10%
MHz
of peak level.
and
all
values of
DE
3.2:
The
39
division of apparitions into groups of similar Devalues.
gives the
De
The
table
ranges and the apparitions which are combined for each of
44
these sets
3.3:
3.4:
An
overview of the 26.3
Confidence of
MHz
De dependence on non-Io-A
first column is Spearman's rank correlation

column is the degrees of freedom
3.5:
The
source characteristics. The

coefficient, the
second
58
slopes and errors for those non-Io-A source characteristics
demonstrating a very likely dependence on
3.6:
49
array data used in this study
Confidence of
De dependence on
De
58
Io-B and non-Io-B source
67
characteristics
vii
3.7:
The slopes and
errors of those characteristics of the Io-B source
demonstrating a very likely
De dependence on Io-A
3.8:
Confidence of
3.9:
The slopes and

very likely
1:
The
errors of Io-A source characteristics demonstrating a
De dependence on
75
Io-A' source characteristics
75
De dependence on
80
Io-C source characteristics
slopes and errors of those characteristics of the Io-C source
demonstrating a very likely
A.
74
De dependence
3.12: Confidence of
The
74
source characteristics
slopes and errors of Io-A source characteristics demonstrating a
very likely
3.13:
67
De dependence
3.10: Confidence of
3.1
De dependence
Antenna
80
De dependence
characteristics for the
MHz
20
MHz
and 22
MHz
Yagi
107
antennas
A. 2: Line losses and gains and antenna
failures
1965-1985. Losses and gains
107
are in units of decibels (dB)
A.3: Line losses and gains and antenna failures 1986-1994
Vlll
108
LIST OF FIGURES
page
Figure
1.1:
Occurrence probability versus System
III
(1965.0) Central Meridian
Longitude for data spanning the years 1957-1994. The three histograms
are from the 18,
20 and 22
MHz
Yagi antennas.
Of
the three
frequencies, the highest overall occurrence probability
The
1.2:
three strongest decametric sources A, B, and
plot of occurrence probability at 18
MHz
is at
18
MHz.
are labeled
for the years
spanning
1965-1992. The Io-related sources are labeled. Non-Io A is the vertical

line at central meridian longitude of approximately 252 Non-Io C is
.
weakly visible
at
about 324
CML. The CML
these sources are given in Table
1.3:
and Io Phase range for

5
1.1
Average radio power flux density spectrum with band designations from
[Barrow and Carr, 1992]. High and low end cutoff frequencies are
explained
2.
in the text
sample of the format developed by Higgins for the reduction of the
Jupiter synoptic monitoring
2.2:
bias
found
in the
program data
16
amount of activity observed during weekday

in weekend watches. The points are labeled
watches versus that found
with the season over which the data
1987-88 season. The 45

detected during
averaged. Label 8788
amounts of
fit
to data
The amount of weekday
is
for the
activity
weekend and weekday watches. The lower
represents the best
2.3:
is
line represents equal
line
from 1985-1994
activity time in the apparitions
through 1994 has been divided by the slope of the best
19
from 1984
fit
line to the
data over this period. These bias corrected occurrence probabilities are
plotted with data taken
from 1975 though 1983
ix
20
2.4:
plot of the
MHz
monthly occurrence probability for 18
Yagi
observations from 1957-1994. Data are binned by the number of
months before or
1
after the opposition date for
each year. Error bars are
standard deviation for the scatter around the
mean
represented by the
22
point in each bin
2.5:
Plot of
monthly occurrence
probabilities at
MHz centered on the month
of opposition for those years during which the sunspot number was low.
2.6:
Plot of
monthly occurrence probabilities
month of opposition
at 18
for those years during
MHz
centered on the
which the sunspot number

23
was high
2.7:
23
plot of the average
20 and 22
MHz
background antenna temperature measured
The
with Yagi antennas.
largest
at
18,
peaks occur when
Jupiter crosses the galactic equator near the galactic center, the smaller
peaks occur when Jupiter crosses the galactic equator
in the anti-center
25
direction
2.8:
Number
distribution of
DAM
MHz
bursts in 1986 at 18
in
terms of the
temperature of each burst above the galactic background antenna

temperature. This season had a low average galactic background
29
antenna temperature
2.9:
A
18
plot of the occurrence probability for all
MHz
Yagi.
The curves
the data collected with the
are the uncorrected O.P.,
and the O.P.

33
corrected for distance and galactic background effects
2.10:
Sample
on June
calibration curve.
1,
1987 for the 18
The points
MHz
are taken
Yagi.
The
polynomial. The attenuation was off of a
from a calibration done
best
HP
fit
line is a fourth-order
461 A noise generator with

36
a noise temperature of 62.7 million Kelvin
2.
Plot of antenna temperature for the highest peak every 5 minutes. This
storm was observed on June
1,
1987 with the 18
solid triangles represent the galactic
MHz
Yagi antenna. The
background antenna temperature
determined for the beginning and end of the storm interval.

polynomial was found from these two points and
2.12:
The
galactic
is
shown
first
order
in the figure.
36
background subtracted antenna temperature of a Jupiter
decametric storm. The antenna temperature of the peak for every 5
minute interval
is
subtracted from the interpolated galactic background
over that interval
37
3.1:
An
3.2:
Comparison of occurrence probability and
illustration of the
source regions defined
meridian longitude and Io phase
at 18
40
Table 3.1
in
flux density against central
MHz
during the 1987-88
42
apparition
3.3:
The De dependence of the non-Io-A occurrence probability

The occurrence probability for apparitions from 1957-1994
degree bins of
3.4:
given
3.5:
De dependence
The
from 1965-1993
definition of average flux density
is
43
in the text
Contours of constant occurrence probability
CML
groups are the same as those listed
as Figure 3.5 but
flux density.
in the Io
(horizontal axis) plane. This plane
each of 3 frequencies (columns) and 5
Same
of the non-Io-A flux density
flux density for apparitions
degree bins of De-
axis) versus
3.6:
MHz.
43
MHz. The average
are in
18
De
histogram illustrating the
at 18
at
are in
now
The De groups
in
De
phase (vertical
is
represented for
ranges (rows).
The De
Table 3.2
the contours for each represent average

are the
same
as in Table 3.2 except
46
excluding seasons before 1965
3.7:
26.3
MHz
Desch
et
Array, 1973 apparition. This plot
a!.,
45
is
of the same data used in
49
1975
3.8:
26.3
MHz
Array, 1974-75 apparition
50
3.9:
26.3
MHz
Array. 1975-76 apparition
50
3.10: 26.3
MHz
51
3.11: 26.3
MHz
51
3.12: 26.3
MHz
52
26.3
MHz
52
3.14: 26.3
MHz
Array, 1981 apparition
53
3.13:
xi
3.15:
comparison of occurrence probabilities between
MHz
Yagi antennas and the 26.3
from 1973
to 1981.
The
probability and the bottom
contours are
3.16:
at
CML
The high
row
top
row
CML
histograms of occurrence
CML-Io phase contour
are
10, 20, 40,
edge of the non-Io-A source
maximum
full-width-half
are
occurrence probabilities of
The high and low
apparition.
CML
at 18
3.17:
The
CML
The
plots.
60 and 80%. ...
MHz
for
each
edges are defined by the
The
amount of
54
non-Io-A source during each apparition
of the peak of the non-Io-A source at 18
MHz
for
each
55
apparition
3.18:
The low
CML
3.19:
The
fit
best
lines to the
lines are the
same
and the errors
non-Io-A
CML
20 and 22
the dates of highest
A CML-Io
MHz
MHz
55
peak, midpoint and edges.
The values
of these lines are presented
in the slopes
probability at 18,
at 18
as those in Figures 3.16-3.18.
3.20: Plots of seasonal averaged sunspot
3.21:
53
longitude of the occurrence probability.
error bars are proportional to the inverse square root of the

listening time for the
MHz
20 and 22
18,
array for the integrated apparitions
number,
De and
for non-Io-A.
in
The
for the slopes
Table 3.5 ..
56
occurrence
The
vertical lines
mark
De
57
phase occurrence probability contour plot of Io-B for the
1965-66 apparition
at 18
MHz. The
histograms
in
CML
and Io phase
can be interpreted as projections of the contour plot along the two
60
coordinates
3.22:
Same
now contours
sum of the flux
as Figure 3.21 but
The histograms
are the
Units are in Janskys (IO"
3.23:
An
26
m2
represent average flux density.
density received within each bin.
60
Hz)
enlargement of the contour plot of Figure 3.21 illustrating recurrent
features of the source
3.24:
61

axis) versus
CML
Io-B source
is
ranges (rows).
(horizontal axis) plane.
in the Io
phase (vertical
The region containing
the
represented for each of 3 frequencies (columns) and 5
The De groups
are the
xii
same
De
as those listed in Table 3.2
62
3.25:
Same as Figure 3.24 but now for contours of average flux density. The
Dg groups are the same as in Table 3.2 except excluding seasons before
63
1965
CML
peak, midpoint and edges
64
phase peak, midpoint and edges
64
3.26:
The
best
fit
lines to the Io-B
3.27:
The
best
fit
lines to the Io-B Io
3.28:
Plots of seasonal averaged sunspot number,
probability at 18,
dates of highest
3.29:
20 and 22
for Io-B.
'
CML
source
De
occurrence
vertical lines
mark
the
66
in the Io
phase (vertical axis)
the Io-A, Io-
De
is
ranges (rows). The
Same
Dg and
The
De

versus
3.30:
MHz
De
as Figure 3.29 but
now
groups are the same as
in
Table 3.2
for contours of average flux density.
The
Table 3.2 except excluding seasons before
in
70
1965
3.31:
MHz
Plots of occurrence probability versus Io phase at 18
ranges of
degrees.
DE
These
plots are integrated over
The Io-A peak
is at
CML
for various
from 200
to
290
240 degrees Io phase and the Io-A' peak
is
71
between 200 and 220 degrees
3.32:
The
best
fit
lines to the
Io-A Io phase peak, midpoint and edges
3.33:
The
best
fit
lines to the
Io-A
CML
3.34: Plots of seasonal averaged sunspot
probability at 18,
dates of highest
3.35:
69
20 and 22
MHz
72
72
number,
for Io-A.
De and
The
occurrence
vertical lines
mark
the
73
De

axis) versus
CML
Io-C source
is
in the Io
phase (vertical
the
ranges (rows). The
De
CML
in
3.36:
The
best
fit
lines to the
Io-C
3.37:
The
best
fit
lines to the
Io-C Io phase peak, midpoint and edges
XIII
De
Table 3.2
77
78
78
3.38:
dates of highest
3.39:
number, De and occurrence
Plots of seasonal averaged sunspot
20 and 22
probability at 18,
The
best
fit
MHz
for Io-C.
The
vertical lines
lines to the low,
high and peak
The
CML
vertical
locations for each of
dashed
peaks or edges which could not be proven to have a
The
best
fit
the
79
the sources discussed in this chapter.
3.40:
mark
De
lines to the low,
De
lines represent
dependence.
81
high and peak Io phase locations for most
of the sources discussed in this chapter. The horizontal dashed lines
De
represent peaks or edges which could not be proven to have a
82
dependence
3.41:
The non-Io-A occurrence probability values from 1957 through 1994 at

8 MHz. The three sections are divided by dates of minimum D E value.
.
3.42:
Pde
for sections
expresses the
3.43:
3.
Pde
the probability factor which
is
85
of the occurrence probability
triangles indicate the
and
DE
2 and
3.
line)
3.45: Pssn for sections
1,
85
observed points
in seasonal
comparison of the time variations
number (dashed
averaged sunspot
from 1957-1994
P S sn
is
87
the probability factor
which
87
expresses the sunspot dependence of the occurrence probability
3.46:
Model 2a of
source.
P DE
the occurrence probability for the 18
The constant of proportionality
sections.
3.47:
The
for sections
and
2.
is
The drop
in probability at
maximum
is
equal to 0.23 for
the constant of proportionality,

for each section
is
all
three
88
De >
2.7
is
used
all
at 18
three
to
.
probability at 18
89
MHz. The
sections. The
90
observed points
Model 2b of the non-Io-A occurrence
non-Io-A
occurrence probabilities.
3.48: Model 2b of the non-Io-A occurrence probability
constant of proportionality
MHz
equal to 0.23 for
observed points
model the "notch" seen during the
3.49:
84
of the occurrence probability for the 18 MHz non-Io-A source.

Model
The constant of proportionality is equal to 0.23 for all three sections.
The
3.44:
2 and
1,
De dependence
MHz now allowing
A, to vary between sections. The value of
shown. The triangles indicate the observed points.
xiv
90
3.50:
non-Io-A occurrence probability based on the source
Model
3 of the
width.
The constant of
proportionality
0.23.
is
The
triangles indicate
92
the observed points
3.51:
model of the non-Io-A 18
MHz
source width and sunspot number.

three sections
3.52:
0.23.
is
The
occurrence probability based on
The constant of proportionality
in all
93
triangles indicate the observed points
superposition of the occurrence probability data from the 1970s on
the occurrence probability data of the 1980s (12 years later).
dashed
line illustrates that the
declining phase of the 1970s
and early 1990s. This drop
was higher than
in
The
occurrence probability during the

that seen in the late
occurrence probability
may
1980s
be related to
95
the sharp rise in solar activity at the time
4.
latitudinal slice
MHz
18
through the non-Io-A emission beam based on data
from 1957-1994.
An
envelope
illustrating a possible shape to the beam.
upper edge of the

is
4.2:
in the
An
illustration,
higher De-
phase
is
that
is
at
drawn around the points
The dashed
sampled during
line represents the
that time in
which Earth
100
notch as described in the text
not to scale, of the effect that
the Io-B source.

at
beam
is
Cones
at progressively
The De range
is
De
has on the Io-phase of
lower Io phase intercept Earth
represented by the vertical bar and Io
increasing towards the right
XV
102
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
LONG-TERM PERIODICITIES
THE JOVIAN DECAMETRIC EMISSION
IN
By
Leonard Norman Garcia
December, 1996
Chairman:
Thomas D. Can-
Major Department: Astronomy
Observations of the decametric radio emissions from Jupiter have been

for 39 years at the University of Florida
large data set that has been accumulated
of long-term variations
emission.
years,
in the
The predominant
which
is
from these observations
periodicity that
is
used
in
was detected was on a time
beaming
an investigation
scale of about 12
was found
effect
caused by the periodic variation
in
were verified and further developed. The Dg
to be greatest for Jovian source non-Io-A.
this
source was developed which provided a good
De
effect
was
this dissertation, the very
the period of Jupiter's orbital motion about the Sun. Previous suggestions
the Jovicentric declination of Earth,
effect
regularly
occurrence probability and average flux density of the radio
that this periodicity is a geometrical
De,
Radio Observatory. In
made
fit
to the
rotating
observed
beam model
effect.
The
for
periodic
also found to be present in the Io-related sources Io-B, Io-A and Io-C,
but to a considerably smaller extent than in non-Io-A.

indicated a possible relationship with sunspot number,
jcvi
The
analysis of this data set also
which has
a periodicity varying
between 10 and 12 years, but more observations

can be verified. Other investigations included a
previously established boundaries (in
will be necessary before this relationship
much needed
revision of the widely used
terms of central meridian longitude and Io phase)
of the major Jovian decametric sources, and a measurement of the elongation effect,
the variation of occurrence probability with respect to Jupiter's elongation angle
the Sun.
xvii
i.e.
from
CHAPTER
INTRODUCTION
1
1.1
The Discovery of Jovian Decametric Radio Emissions
The study of planetary magnetospheres outside of our own began

discovery of radio emissions from Jupiter
Nebula
at
a frequency of 22
They found during
and K. L. Franklin [1955].
always occurred
at the
same
sidereal time.
in sidereal time.
Over
to
announcement of radio emissions from
that Jupiter activity
cosmic background noise records
months
the course of three
in the
this source
sky [Franklin, 1959]. Soon
Jupiter, C.
confirm these observations from ones he had made
found
by Bernard Burke
After checking an ephemeris, they found that the position
of Jupiter coincided with the location of this source

after the
1955 with the
a drift scan of the sky near the Crab
that a source of intermittent interference in the
slowly drifted
MHz
in
at 18.3
A. Shain [1956] was able
MHz
had been present on records he had made
from Australia, and
in 1950.
Shain found
that the occurrence of Jovian activity tended to recur at intervals about equal to the so-
called
System
II
period used by optical astronomers. Gardner and Shain [1958] further
refined the analysis of the longitude effect and identified one primary source
secondary sources of Jovian decametric

fixed in longitude relative to
drift in
to obtain apparently precise
drift rate
II
emission.
The
and two
three sources remained
All three sources, however,
one another.
longitude relative to the System
measurements of the mean
(DAM)
were found
to
period. Subsequently, several observers used
over intervals of several apparitions of the planet
measurements of the
rotation period.
the rotation period of the radio sources in data taken at 18 and
Gallet [1957] found
22
MHz
to
be about
2
10 seconds shorter than the System
et al.
II
period of 9 hours 55 minutes 40.6 seconds. Carr
obtained a similar value from their observations. They defined a rotating Jovian
longitude system called System
with System
System
II at
Oh
UT
period to 9h
III
which has a period of 9h
III,
on January
55m
1,
55m
1957 [Carr et ai, 1958]. They
System
III
(1957.0) period as 9h
55m
Observing Jupiter through a telescope

there
is
no solid surface on which
to track
only "feature" of some permanence
cyclone for
at least
devised.
The
constant
rate.
to specify
300
years, but
and zones of
belts
cloud locations
to
it
29.37s.
all
one can see of Jupiter are bands of clouds:
permanent features as the planet
the Great
Red Spot which has
and dark colors
light
rotates.
move
rotate differentially
and never
in
in the equatorial region.
in the equatorial
by the System
9 hours 50 minutes.
planet generally follow the System
By
II
in
at
which
regions of latitude, but none of the clouds remain
two systems. Clouds
at a rate specified
The
existed as a distinct
any longitude system that can be
drifts erratically in
in different
through 360 of longitude
than
is
Union [1962] approved
Observers have classified for convenience two longitude systems
stationary in either of the
assumed
later revised the
29.35s [Carr et ai, 1961]. Using data from decimetric as
well as decametric observations, the International Astronomical
the
and coincided
28.8s,
region of the planet are
longitudes.
These features
rotate
The more temperate regions of
the
longitudes and have a period some 5 minutes longer
specifying a zero longitude along the central meridian
on some date and using the appropriate longitude system the central meridian longitude
(CML), or
the sub-Earth longitude, for any future date can be determined.
of these systems was there any
body of
the planet
way
was spinning.
to be sure at
what
rate the inner,
With neither
supposedly
solid,
The discovery of
the
be used as a "landmark"
gave observers
Jupiter.
Once
for the
it
DAM and the
(i.e.,
first
a relatively permanent and regularly appearing radio feature)
time a means to study something located
was determined
of the planet, a fixed
field
landmarks could be used
main source (source A) could
realization that the
that the
DAM
had
its
of
origins in the magnetic field
which was generated within the
to establish a constant
in the interior
interior of the planet, these
and meaningful
Rotation
rotation rate.
periods based on Jovian decametric observations have been found to be very stable for
more than 30 years [Higgins,
0.3
1996].
'
!
Figure
1.1:
Occurrence probability versus System
III
(1965.0) Central Meridian Longitude for
data spanning the years 1957-1994. The three histograms are from the 18, 20 and 22
antennas.
Of
the three frequencies, the highest overall occurrence probability
three strongest decametric sources A, B, and
plot of a
to
18
MHz
Yagi
MHz. The
are labeled.
measure of the likelihood of detecting
shows three peaks which have come
is at
DAM
versus Jovian longitude
be known as "sources." Current theories assume
that they are not distinct sources, but instead
may be due
to rotating
beams from
a single
source, or perhaps from
two sources located
have been designated A, B and

as
The sources
opposite Jovian hemispheres.
in
in
some
literature they
may be
referred to
Main, Early and Third, respectively. The standard way of characterizing the location
of these sources
in
is
terms of the central meridian longitude (CML). The measure of
likelihood of detecting
computed by dividing
DAM
is
The
DAM
satellite Io in its orbit.
activity detected
probabilities of detecting
more than 20%.
probability are given in Chapter 2.
1.1.
DAM
number of hours of
the
of the sources are rarely
Figure
referred to as the occurrence probability.
The occurrence
hours of listening.
in
although
More
DAM
It
can be
by the number of
even
in the center
on computing the occurrence
details
plot of occurrence probability versus
CML
is
given
can be further characterized by the location of the Galilean
Bigg [1964] discovered
that the 3
major
DAM sources are strongly
influenced by the departure of Io from superior geocentric conjunction, or the Io orbital
phase from superior geocentric conjunction. The
known
Io are
DAM
sources which are influenced by
and Io-C. Each source also has a non-Io related component
as Io-A, Io-B
DAM sources
as well as an Io-related one.
plot of the
Also visible
in
CML-Io phase diagrams
in
Figure
MHz
is
1.2.
the Io-D source.
somewhat higher
Carr et
al.
Io-D
is
until after
is
given
frequencies below about 18
at
[1958] noted the existence of a very pronounced eight-day periodicity of
terrestrial
responsible.
this effect
by means of a correlation with
observing conditions turned up negative.
Ganymede on
orbital period of this satellite
Ganymede was
CML-Io phase space
Io phase.
sunspot number or
satellite
at
located at similar longitudes to the Io-B source but
Jovian radio emissions. All attempts to explain
by the
in
is
the Jovian radio emission
possible effect
was also considered, since
the
about 8 days, but no convincing evidence was found that
The explanation of the 8-day
periodicity
was not forthcoming
Bigg's momentous discovery of Io control. The 8-day cycle of Carr etal. turned
out then to be the beat frequency between the roughly 10 hour Jovian rotation period
and the 42 hour Io
orbital period.
CML
Figure
1.2:
plot of occurrence probability at 18
Io-related sources are labeled.
approximately 252.
Non-Io
Non-Io
is
range for these sources are given
weakly
in
is
MHz
for the years
Table
CML. The
results
field outside
of Earth.
Phase
is
its
strong circular, or
one event was seen with mixed
left-
and
Gardner and Shain pointed out that these polarization
imply the presence of a magnetic
magnetic
Io
Franklin and Burke [1956] found most of the radiation
to be right-hand circularly polarized, although
right-hand circular polarization.
and
DAM
Another characteristic of the decametric radio emission

elliptical polarization.
CML
1.1
1.2 Polarization of
sometimes
spanning 1965-1992. The
the vertical line at central meridian longitude of
visible at about 324
It
is
field.
This was the
now known
first
that the
elliptically or circularly polarized, in the right-hand sense for
indication of a planetary
emission
sources
is
almost 100%
and B, and often
in the left-hand
sources
is
X-mode
fast
sense for source C.
given in Table
at
1.1
The dominant
The emission
polarization sense for each of the
a frequency just above the local electron gyrofrequency.
The C source
tends to be right-hand polarized at the low longitude edge of the source but
increasingly left-handed with longitude [Kennedy, 1969].
right-handed over
Table
1.1:
all
The
and
becomes
sources remain
the frequencies studied.
Various characteristics of the
CML
Source
DAM
believed to be radiated in the right-hand
is
range
DAM
lo
sources.
Phase
range
Designation
Maximum
Dominant
frequency
Polarization
(MHz)
Io-D
0200
Io-B
105 185
80
non-lo-B
80200
lo-A
non-Io-A
30020
225260
38
RH
and
LH
300 360
0360
32
RH
and
LH
Io-C
non-Io-C
95 130
18
LH
39.5
RH
0 360
38
RH
200270
205260
38
RH
230280
0 360
38
RH
10
from Carrel al, 1983.
.3
The discovery of
Frequency Dependence of Jovian Radio Emissions
a decimetric wavelength nonthermal
component of
spectrum (DIM) provided further insights into the Jovian magnetic

1959]. This radiation
flat
spectrum and
to
was found
to
have a large
linearly polarized
the Jovian radio
field
[Sloanaker,
component, a relatively
have an emission region centered on Jupiter but extending about 2
Jovian radii on either side of the planet.
It
was soon recognized
implied a population of high energy electrons trapped
in
that this emission
a magnetic field and emitting
through the synchrotron process. The discovery of these Jovian radiation belts occurred
nearly at the
et ai,
same time
1958].
the linear
of 10.
that Earth's
Van Allen radiation
DIM
characteristic of the
component
emission
is
varies sinusoidally with about a 10 hour period
This variation
axis of the planet. There are
the axis of
field,
many review
which
articles
is
tipped by 10 from the rotation
on the synchrotron component of
Jovian radio emission [Roberts and Komesaroff, 1965, Carr
component had a maximum frequency of nearly 40
component only reached 28 MHz. Assuming

extraordinary
magnetic
mode
at frequencies just
field strength at the
magnetic
field intensities
source could be estimated.
of
at least 14
The magnetic
opposite that of Earth. These two

that the north
the radiation
Gauss
in the
north pole
maximum
is
while the left-hand

in the right-hand
Gauss, the 40
MHz
is
first
approximation
in
and 28
maximum
northern magnetic hemisphere and

located in the northern hemisphere,
field intensities
end of an assumed magnetic dipole
good
MHz
emitted
Since the gyrofrequency
field strength in
together with the observation
tipped by about 10 toward the System
longitude of 200 can define an offset tilted dipole
was
is
the right-hand and left-hand components, respectively, imply
10 Gauss in the southern.
III
the
Gulkis, 1969, Carretai,
above the local electron gyrofrequency, the Jovian
approximately 2.8 times the magnetic
is
MHz maxima for
this
and
Studies of the polarization of the decametric radiation found that the right-hand
polarized
MHz
and an amplitude
of the polarization plane was evidence of a
in the orientation
predominantly dipolar magnetic
1983].
were discovered [Van Allen
belts
that the plane of polarization of
(OTD) model of
the field. Although
to the true field distribution, later in situ
measurements
by Pioneer and Voyager spacecraft demonstrated that there are also strong quadrupole
and octupole components (and probably higher multipole components as well) [Acuna
et ai.
1983].
The Jovian decametric emission

cutoff of 39.5
MHz
now known
to a rather indistinct
MHz. A spectrum
to extend
1.3.
low sunspot number
detected in the
is
just
is
medium. At
the high frequency
starts to
about
is
given in
observations have been
made
The lowest frequency
end of the decimetric emission region thermal emission
1
i
regions
above the plasma cutoff frequency of the interplanetary
1
i
merges
at
the prevalent source of radio noise.
become
1
i
(KOM)
and by the Voyager flyby missions.
from Jupiter
it
ground-based observers can detect during years of
that
MHz. Lower frequency
about 10
satellites
KOM band
10"'"
where
of the Jovian radio emission illustrating the intensities and extent
The lowest frequency
by Earth-orbiting
MHz
low frequency cutoff of 3
of the decimetric, decametric, hectometric, and kilometric
Figure
from a sharp high frequency
emission [Carr et al, 1983]. The peak flux occurs
(HOM)
into the hectometric
is
1
1
mi|
'
DAM
HOM
10"
10"
10"
Burst Emission
10"
i
-J
u
V
<
10"
DIM
10"
Synchrotron Continuum
'
10"
'
'
10*
10"
10
10
10
10'
10"
Frequency (Hz)
Figure
1.3:
Average radio power
flux density
spectrum with band designations from [Barrow and
Carr, 1992]. High and low end cutoff frequencies are explained
in the text.
.4
The Jovian Emission Beam Model and
the
De
Effect
After 1961, occurrence probability histograms appeared to indicate a rotation period

for the decametric sources that
was about
[Douglas and Smith, 1963, Smith
able to
show
that of
System
III
(1957.0)
1965]. Gulkis and Carr [1966] recognized that
of the sources had both a secular and a cyclic component.
this drift in the longitude
they assumed that the
et al,
second longer than
When
cyclic error component had a period of about 12 years, they were
that the secular
component of
measurement was about
the rotation period
+0.3 seconds. They demonstrated that the cyclic error was most likely due to the fact
that
De, the Jovicentric declination of Earth, was
and ending times of the

rotation period
significantly different at the beginning
interval (several years long) over
was measured. The cyclic
which the average value of the
component was
error
first
thought to be due
to changes in solar activity since the period and the phase of this drift
correlation with the solar cycle (about
1 1
years).
orbital period
is
The period of
One
to the orbital period of Jupiter (1 1.86 years).
was
the drift
was
3.1 to a perpendicular to
its
orbital plane. Jupiter's orbital plane
vary in Jovicentric declination by

Jupiter as well as the
The combination of
3.3.
DE
the
varies with the
year period of Earth in
its
also close
quantity that varies with the Jovian
the Jovicentric declination of Earth or De- Jupiter's axis
Earth's orbital plane [Bishop, 1995].
in near anti-
orbit.
The
two
1
is
is
inclined by
inclined by 1.3 to
effects causes Earth to
1.86 year orbital period of
DE
effect
is
a geometrical
one, resulting from the fact that the ground-based radio observers see different parts of
a narrow emission
planet,
beam
that corotates with Jupiter during different apparitions of the
and therefore an error
source peaks
at the first
and
will
last
from Gulkis and Carr [1966] was
be made
in
assuming
that the longitudes
apparitions were the same.
An
of the radio
important conclusion
that the cyclic error in rotation period
measurements
can be largely eliminated by selecting a pair of apparitions having mid-dates that are
10
about
2 years apart for the observations, so that the average
will be
At
approximately the same.
completely rule out a solar anti-correlation on the
The
drift in the
Dulk found
[1970].
et al.
in the
(A) source and the early (B) source were
dynamic spectra
the sources with
latitudes.
De was due
to
is
first
magnetic
that the drift of the
He concluded
main
that the
The divergence
of
changing perspective as Earth moved into higher Jovian
field
from a perfect
dipole,
not oriented such that any edge of the cone reaches Earth [Dulk, 1967].
thorough investigations of higher-order multipole-components of the Jovian
in the
the Pioneer 10 and 11 missions.
came with
field
made
the
first in situ
Jovian system. The Pioneer
(Rj) of the planet [Acuna et
al.,
In 1973, Pioneer 10
measurements of the magnetic

1
spacecraft in particular
detailed information on the Jovian magnetic field since
to
to
southern hemisphere cone should also be visible from Earth according to
1974, Pioneer
plasmas
apparitions
upon by Dulk [1967]
thin, conical sheet.
Dulk's model, but due to asymmetries of the Jovian magnetic
cone
two
decametric sources.
opposite directions.
in
decametric emission was sharply beamed into a
The
for the
were insufficient data
results of Gulkis and Carr were substantiated and enlarged
and Carr
this
De
the time however, there
it
field
and
in
and embedded
was able
to provide
passed within 1.6 Jovian
radii
1983]. Both spacecraft found the Jovian magnetosphere
be an immense structure of about 100 Rj on the sunward side. The magnetotail was
detected by the Voyager spacecraft out to about the orbit of Saturn, 5

1983]. If the magnetosphere
be larger than the
full
Moon. The
distorted outside of 6 Rj
disc.
were
due
visible,
it
AU
distant [Desch,
would appear from an observer on Earth
mainly dipolar shape of the magnetic field
to trapped charged particles
This disc causes the field lines to have a
much
is
to
severely
forming an equatorial current
larger radial
component along
the
magnetic equator. From the Pioneer and Voyager spacecraft data several models of the
Jovian magnetic field have been
built.
Since
all
of the spacecraft were on flyby missions,
from complete. This
the sampling of the magnetic field over latitude and longitude
is
allows for a family of different models to be proposed
of which
fit
the data.
One of
the models proposed that best
The
model included terms up
model
field
and Voyager
is
the
may
adequately
was
the observed data
the
O4
to octupole for the field generated within the planet.
external magnetic field due to the current disc
magnetic
far
Using data from the magnetometer on board Pioneer
model [Acuna and Ness, 1976].

11, this
fit
all
was assumed
negligible.
newer
06 model which uses weighted observations from Pioneer
[Connerney, 1992]. This
is
also a multipole expansion to the octupole
term but provides for the existence of higher orders [Connerney, 1993]. This model
in the
made
process of being modified due to

in the infrared
[Connerney
Perhaps the
how
first
The
thing one notices
is.
interplanetary
1993].
The
when making ground-based observations of DAM
a scintillation effect
DAM
and there are large variations
in intensity
emission.
The
DAM
by
during one event.
upon
the emission
Earth's ionosphere will
and out over several seconds.
also cause scintillation of the
DAM
is
radio emissions appear in "storms" of perhaps an hour
medium impresses
so the radio noise will fade in
two types of
is
observations of the Io flux tube footprint
Observing Jovian Decametric Radiation
.5
bursty the emission
or less.
et ai,
new
its
nature
is
intermittent
Observers have classified
emission bursts, long or L-bursts and short or S-bursts.
L-bursts
appear as wide bandwidth (few MHz), amorphous emission regions when displayed on
frequency-time spectrograms, or dynamic spectra. L-bursts usually
and when the radio noise

a beach.
is
last several
seconds
played over a loudspeaker sounds like waves crashing on
L-bursts often display the effects of terrestrial and interplanetary scintillation
as well as modulation lanes. Riihimaa [1970] discovered that in
some dynamic
spectra,
12
modulation lanes appear as alternate stripes of increased and decreased emission that
drift
al.
by about 150 kHz/s, and can have either positive or negative slopes.
[1992] have proposed that modulation lanes
in
Imai et
observed dynamic spectra represent
frequencies and times of constructive and destructive interference produced as wide-
band decametric emission passes through an interference screen located near

1
Io's orbit at
80 from superior geocentric conjunction. L-bursts observed near the planet might be
several seconds to minutes long, their
of the
DAM
nonuniform interplanetary plasma on

broken up by interplanetary
that
we
dynamic spectra often exhibiting
On
seen in dynamic spectra taken by Voyager.
actually observe.
its
way
the arc structure
propagating through the
to Earth, these initial long L-bursts are
scintillation into the shorter L-bursts (1 to 10
The modulation lanes
are
still
second duration)
present, but have large gaps
caused by the fading of intensity between successive L-bursts. S-bursts have only been
observed with the Io-related sources. S-bursts are repetitive sequences of short-duration
frequency-drifting pulses having lifetimes less than
significant structure on time scales
private communication).
down
The frequency
MHz/sec always from high
to a very
drift
100 msec each.
Each pulse has
few microseconds (Carr and Reyes,
of these S-burst pulses varies from 5 to 45
low frequencies. S-burst dynamic spectra on timescales of
to
milliseconds can exhibit remarkably varied structure [Flagg et al, 1991].
1.6
The emission beam model

of the
DAM
edges of a
is
thin,
Discussion of the Problem
that is
now used
to explain
many
very similar to that developed by Dulk [1967].
hollow cone fixed to
and rotating with the planet
in
field lines
little
of the characteristics
The "sources"
are the
running through the center of the planet
under 10 hours.
The
latitudinal effect provides
information on the east-west width of the cone and the cone thickness.
The
and B
13
sources are the trailing and leading edge of a cone located
and
just
above the ionosphere of the
hemisphere cone but that
The
latitudinal effect
is still
uncertain.
DAM
on the
northern hemisphere
in the
The C source may be one edge of a southern
planet.
provides information on the north-south extent
of the emission beam. This effect depends on the slowly changing declination of Earth
in Jovicentric coordinates.
is
Since one Jovian orbit about the Sun
necessary to study several dozen years of data
control of the Jovian decametric emission.
is
unique
in
is
The
DAM
is
how
DAM
changes with the Jovicentric
result of this analysis will be an
the shape of the emission beam.

to enable
one
improved understanding of
second consequence of studying these long term

to find solar influences
with the roughly decade long solar sunspot cycle.
In
on
DAM
in the
is
DAM
given
due
to
viewing geometries and changing galactic background levels are removed. Chapter
a study of the latitudinal control of the
effects.
DAM
which may vary
Chapter 2 a description
of the data reduction method and of the method whereby variations

orbital
is
it
being able to provide a data set which spans several Jovian years (12 Earth
declination of Earth.
nearly 12 years
order to understand the latitudinal
The University of Florida Radio Observatory
years each) and allows one to thoroughly study
variations in
in
DAM
Chapter 4 will summarize what these
emission
beam and
of a solar cycle.
and a search for any long-term solar
results
may mean
the changes the source region
for the
shape of the
may undergo over
the course
CHAPTER
METHODS
Time: 330 Started watch period. Occasional very faint
interference.
portion
Set clock with recorder.
of log from
observation taken
Thomas D.
at the
The radio observations used
December
in Chile.
The
31, 1956.
1957 during the
Carr, January 8,
station
first
week of
Data Collection
in this
University of Florida Radio Observatory
(MRAO)
No
University of Florida Radio Observatory
2.
Observatory
static-like bursts.
Killed scorpion.
study were taken by antennas located
(UFRO) and
Radio observations
station
was located on
at the
at the
the
at the
Maipu Radio Astronomical
University of Florida began on
campus of
the University of Florida
near the current location of the University Public Safety office. In 1962 the observatory
was moved
the
UF
to the northern shore of
Bivens
Radio Observatory was moved
The data from

The
MRAO
UFRO
18,
are
added
20 and 22
Collins 75S-1 receivers.
The
Arm
lake in Gainesville. In the
to a site near
Old Town
to the Florida data in the
MHz
in
fall
1959 through 1965 seasons.
data were taken using Yagi antennas connected to
receiver bandwidth
was 5 kHz. The audio output
sent to Texas Instruments chart recorders, Brush chart recorders
and
to
WWV.
important component of the data collection process
up
is
the observer.
signal
was
audio loudspeakers.
Timing marks were recorded automatically from signals received from
to listen to the signal
of 1967
Dixie County, Florida.
It is
The most
to the observer
over the loudspeaker and discriminate between Jovian decametric
radio noise and interference of various kinds.
When
Jupiter
is
recognized over the
loudspeakers the observer will note the deflections due to Jupiter on the chart record.
Immediately
after the Jupiter noise
"storm" the observer will calibrate against a noise
14
15
generator
all
the channels in
since the 1965 apparition
found
to
HP
The
is
which Jupiter noise was present. The noise generator used

a Hewlett-Packard
have a very broad and

461 A
is
flat
HP
461 A amplifier. The
461 A has been
noise spectrum at the decametric radio frequencies.
occasionally calibrated against a pair of type 5722 noise diodes that
serves as the standard noise source.
The observer keeps
a log of the start and stop times
of the watch, observing conditions, and other information pertinent to the maintenance
of the observatory and the further reduction of the chart records.
Every half hour the
observer will indicate the observing conditions for each frequency channel. The standard
notation for observing conditions has been a scale from 5.0, for perfect conditions with
no interference,
to 0.0,
where interference
is
so high that practically no Jupiter radio noise
could be detected over the interference. Determining observing conditions for moderate
interference has been a subjective matter quite dependent on individual observers.
The
importance of observing conditions to the determination of occurrence probabilities will

be addressed
in the
data reduction section.
2.2
2.2.
Data Reduction
Standard Format
The format used

over the years.
column
for reading files into
The use of punch cards
computer programs has changed considerably

for
programming and data
format for the Jupiter data reduction the best choice.
designations for various stages of the data reduction can be found
Later data sets were reduced without the need for punch cards.
a chart record of a night's observation
the start
and stop times
for the watch,
during the night. For nights
in
is
entered as a computer
any Jupiter
activity
entry
made an 80
review of the column

in
Thieman
[1977].
The method by which

file
begins by entering
and any interference intervals
which Jupiter was detected, the peak Jupiter deflections
16
for every five minute interval as well as the deflections due to the galactic
immediately before and after the Jupiter radio storm are noted.
background
The deflections due
various attenuations of a standard noise generator are also recorded.
to
new format was
used for data taken from 1981 through 1992 [Higgins, 1996]. Earlier data sets have also
been converted into
this
one format.
sample of the data format
is
given
in
Figure 2.1.
fl23081. 18y
123081
18y
808
1137
1
809
829
1
1100
1130
X
18.00
1
809
70.00
filename
dace
frequency 4 ancenna
begin watch hour minute (all times UT)
end watch hour minute
number of storms
begin storm
end storm
number of interference intervals
begin interference interval
end interference interval
type of interference (x=static, s=stations, b=buzz)
galactic background deflection at start of storm
calibration number
number of 5 minute intervals
start of first 5 minute interval
deflection of highest Jupiter peak in interval
814
32.00
819
22.00
824
28.00
15.00
galactic background after storm
C1123081 .18y
123081
18y
1
59
3.00
calibration filename
date
frequency 4 antenna
calibration number
number of calibration steps
first decibel attenuation of noise generator
chart deflection corresponding to decibel attenuation
37
14.00
35
20.00
32
31.00
29
47.00
26
68.00
23
93.00
Figure 2.1:
A sample
of the format developed by Higgins for the reduction of the Jupiter synoptic
monitoring program data.
2.2.2 Occurrence Probability
One
the
of the fundamental parameters used to study the Jovian decametric radiation
amount of
activity detected
over a given time period. The occurrence probability
is
is
17
common
indicator of this activity. Occurrence probability
OP =
given as
is
Number of hours
of Jupiter detected
-t
r
Number oj hours ojflistening
,
_
(2-1)
For an accurate estimate of the occurrence probability one needs to know when the
observers were actually "listening." Listening
instruments are operating; the observer
and interference,
if
Jupiter noise
factors.
if
is
present,
is
is
defined as the condition where
listening to the speakers for
all
any Jupiter noise
not bad enough to prevent the observer from determining
How much
present.
If the Jupiter
is
noise storm
prevent an observer from detecting
interference
is
much" depends on many
"too
very strong, almost no amount of interference will
is
With moderate interference and moderate
it.
levels
of Jupiter noise, listening becomes dependent upon the skills of the individual observers
to distinguish Jupiter noise.
For those
who
reduce the Jupiter data, these factors must
be taken into account for accurate occurrence probabilities to be computed.
Personnel
reducing the data will use the observing condition records written in the log book and the
chart records themselves in order to determine the
which
the data sets
conditions were given as less than
that
amount of
1.0,
For
listening.
include "interference intervals" are time periods
conditions where the interference
no Jupiter could have been detected had
it
this study,
when observing
is
so intense
been present. This might be an overly
conservative estimate of the interference intervals but because no systematic scaling of
more moderate
levels of interference
of the observers,
that has to be
it
is
had been made and because of the varying
very difficult to gauge where the cutoff would be.
made when
discussing occurrence probability
is
An
skills
assumption
that the listening times
chosen were completely independent of Jovian longitude and Io orbital location. Since
ephemerides can provide times when certain configurations of central meridian longitude
and Io phase
will take place, this information
can not be used to plan observations
occurrence probabilities generated from these data are to be free of bias.
if
the
18
To build up accurate
statistics
on the occurrence probability of Jovian decametric
emission, several conditions must be met.
made by instruments with
observations
this report
The
condition
first
is
that the catalog be of
similar sensitivities. This condition
is
satisfied in
by making use of data taken only from Yagi antenna systems. The antennas,
the cables, the receivers and the recorders have been of a consistent type and quality over
the span of the data set used in this study.
in
which the data
The
rest
of the conditions deal with the manner
are reduced. Besides accurate estimates of the
amount of
listening time,
accurate estimates of the completeness of the data in terms of the intensity of the Jupiter
noise must be made.
The
likelihood of detecting
weak
Jupiter pulses
factors including the listening conditions at the observatory.
activity affect the density of the Earth's ionosphere
listening conditions.
times
when
it is
will raise the
weak
Weak
pulses
in intensity.
Both the
a function of
local time
The
if
Jupiter
galactic
is
and solar
observed during
background radio noise
detectable flux density and subsequently reduce the
The detected
many
over the observatory and thereby the
also not be detected
located closer to the galactic plane.
minimum
Jupiter pulses detected.
dependence
may
is
radio emission
amount of
from Jupiter also follows a -r
During any one observing season the distance
to Jupiter
may
vary by nearly 2 astronomical units and can also contribute to changes in the likelihood
of detecting weak Jupiter bursts. All of these factors need to be accounted for before a
meaningful discussion on variations

2.2.3
Weekend
An
versus
in the
occurrence probability can be made.
Weekday Observers
inspection of data taken during the late 1980s at the
differences between the occurrence probabilities
tions
the
turned up significant
computed using only weekend observa-
and those using data taken during the week.
new weekday
UFRO
It
appears that from 1984, when a
observer was hired, through the end of the data set used
weekend observers detected more
in this project
Jupiter radio emissions per unit listening time than
19
WEEKEND
OCCURRENCE PROBABILITY
Figure 2.2:
found
that
is
averaged.
bias found in the
weekend watches.
in
Label 8788
is
amount of
activity observed during
The points
the
observer.
this disparity, a plot
line.
1994
is
the data
line represents the best
fit
An example
shown
unbiased apparitions.
which was used
probability
identical for
In order to correct for the bias introduced
was made. On
this plot all
The biased apparitions have
linear best-fit to the biased data
as a correction factor to the data.
weekday observer recorded each apparition

the inverse of the slope of this best
fit
seasonal occurrence
weekend and weekday observers should
of this plot using data taken at 18 and 20
Figure 2.2
in
identical.
of weekend average occurrence probability for each season ver-
weekday occurrence
probabilities which were
45
The lower
Previous to 1984 the occurrence probabilities for weekend and
weekday observations were nearly

by
which
for the 1987-88 season. The 45 line represents equal amounts of
from 1985-1994.
weekday
sus the
weekday watches versus
are labeled with the season over
activity detected during weekend and weekday watches.
to data
02
0.15
0.1
0.05
line.
is
The
MHz
lie
on a
from 1975 through
a different distribution than the
from 1984-94 provided the slope

total
amount of
activity that the
multiplied by a correction factor which
is
This would increase the weekday occurrence
20
probabilities for the apparition to a level roughly equivalent to the
probability.
In this correction
reflect accurately the true
it
is
assumed
that the
weekend occurrence
weekend occurrence
probabilities
occurrence probability for decametric emission detected by our
ground-based instruments. The correction factor for the weekday activity for
22
MHz
is
2.013.
Figure 2.3
bias corrected apparitions
is
18,
20 and
a plot showing apparitions from 1975 to 1983 and the
from 1984
to
1994.
HI
s
0.15
Q
HI
x
X
O
o
>-
3
o
O
x
0.05
0.
o
z
LU
X
X
U
o
0.05
0.15
0.1
OCCURRENCE PROBABILITY
Figure 2.3: The amount of weekday activity time
been divided by the slope of the best
fit
WEEKEND
in the apparitions
line to the data
over
from 1984 through 1994 has
this period.
These bias corrected
occurrence probabilities are plotted with data taken from 1975 though 1983.
2.2.4 Elongation Effect
The beginning and ending dates of the observing seasons

have generally been based on the date of opposition of
at the
Jupiter.
UF Radio
Observatory
The longest nighttime
observations of Jupiter occur on the night of opposition and the season will extend
about 4 to 5 months before and after
this date.
plot of occurrence probability versus
21
months
until opposition at
18
MHz
is
given
Figure 2.4.
in
The
figure
shows
that the
occurrence probability increases as Jupiter approaches opposition but that the degree to
which the occurrence probability drops away from opposition

zero.
The months before opposition
those months after opposition.
hours when
These good observing conditions
after sunset
generally have a higher occurrence probability than
in the early
and radio interference
will often last until sunrise.
is at
morning
a minimum.
After opposition Jupiter
evening. Observing conditions can be quite poor in the hours immediately
when
the terrestrial ionosphere
The
stations can be strong.
is still
quite dense
and interference from radio
density of the terrestrial ionosphere
by the sun. The average electron number densities
is
strongly influenced
ionosphere show a day-night
in the
and a periodicity based on the solar cycle.
periodicity
not symmetrical about
Before opposition Jupiter transits
the terrestrial ionosphere has thinned
transits in the
is
During times of high sunspot
denser than during low sunspot number.
In Figures 2.5
and
number
the ionosphere
2.6 the
monthly occurrence probabilities for years of low and high sunspot number
are plotted.
is
The seasons of low average sunspot number
are those
from 1961-1966,
1972-1978, 1983-1988 and 1992-1994. The seasons of high average sunspot number
are
1957-1960, 1967-1970, 1978-1982, and 1988-1992.
plots
is
the larger
amount of asymmetry between
probabilities in the plot
made during
The
The
interplanetary
The
DAM
This asymmetry
has on the observing conditions
at
these
difference in the overall levels of activity between the two plots will
be discussed in Chapter 3
sion.
significant difference in the
years of high sunspot number.
illustrates the strong influence solar activity
frequencies.
pre- and post-opposition occurrence
in the
medium
scintillates as
context of the
De
effect
on the
DAM.
also influences the detection of Jovian decametric emis-
it
passes through this
medium and
causes the bursts to
be modulated strongly with a period of a few seconds. Genova and Leblanc [1981]
find
22
that the
amount of
scintillation is
to the interplanetary
reduced around the time of opposition.
plasmas moving roughly along the
velocities are very small.
Any
density inhomogeneities
This
is
due
line of sight so their tangential
would take longer
to cross the
observer's line of sight.
0.12
i
~\
0.1
0.08
0.06
0.04
0.02
3-2-10123
Months
Figure 2.4:
1957-1994.
plot of the
L_
Until Opposition
monthly occurrence probability for 18
MHz
Yagi observations from
Data are binned by the number of months before or after the opposition date for
standard deviation for the scatter around the mean represented by
each year. Error bars are

the point in each bin.
23
i
1
1
0.15
LOW Sunapot Number
0.1
0.05
-2-1012
Months
Until Opposition
Figure 2.5: Plot of monthly occurrence probabilities at 18

opposition for those years during which the sunspot
MHz
centered on the
month of
number was low.
0.06
0.04
0.02
-2-1012
Months Until Opposition
Figure 2.6: Plot of monthly occurrence probabilities at 18

opposition for those years during which the sunspot
MHz
number was
centered on the
high.
month of
24
2.2.5 Galactic
The
Background and Distance Effects
sensitivity
background
level.
limit of our instruments
This limit
is
primarily determined by the galactic
often expressed in terms of the
is
antenna temperature. The effects that changes
in galactic
on the minimum detectable antenna temperature need
DAM
analysis of long-term variations in
detectable
accounted for before an
to be
can be performed.
threshold temperature must be set that will allow one to

activity levels free of influence
minimum
background and distance have
In other words,
compare seasons
some
for Jupiter
from the galactic background. Another influence on the
apparent galactic background temperature comes from the orientation of the antenna with
the ground. Since the data used in this study are
from Yagi antennas which track Jupiter
across the sky, the configuration of the antenna with respect to the ground will change
and
this will
change the apparent galactic background
storm takes place while Jupiter
at
is
levels.
In other words,
a low altitude, part of the antenna
pointing towards the ground so the background levels will be less than
takes place while Jupiter
As
Jupiter orbits the
is
Sun
near
if
a radio
beam may be
when
the storm
transit.
the ground-based observer sees Jupiter set against different
parts of the galactic background.
Over one 12-year period Jupiter enters regions of
sky close to the galactic center and to regions near the galactic poles.
the
The average
value of the galactic background across the beam-width of the Yagi antenna will reflect
these changes.
Figure 2.7 illustrates the large changes
background antenna temperatures measured
The
at
18,
in the
20 and 22
seasonal average galactic
MHz
with Yagi antennas.
highest peaks in 1972 and 1984 occur during passage of Jupiter near the galactic
center and are nearly four times the
1990 occur when
minimum
values.
The secondary peaks
in
1978 and
Jupiter passes through the galactic plane in the anti-center direction.
25
1975
1970
1965
1985
1980
1990
1995
YEAR
Figure 2.7:
MHz
plot of the average
with Yagi antennas.
The
background antenna temperature measured
largest
near the galactic center, the smaller
peaks occur
when
peaks occur when
at 18,
20 and 22
Jupiter crosses the galactic equator
Jupiter crosses the galactic equator in the
anti-center direction.
Besides the twelve year period of galactic background changes there

variation in the Earth-Jupiter distance.
The
relative distance of the
by almost 2 astronomical units over one observing season. Both the changes
and galactic background
is
a yearly
two planets can change

in distance
are expected to influence the occurrence probability.
As
the
distance from Jupiter increases over one season, a certain fraction of those bursts which
would have been detected around opposition would drop below

of the Yagi antennas. The occurrence probability
galactic
levels
background
levels since
were low now would be
weak
Jupiter that
lost in the noise.
the detection threshold
would be lower during years of high

would be detected when background
26
If
Ta
is
the galactic
the antenna temperature for a given frequency channel due to Jupiter
background and Tg
is
the contribution
then the contribution due to Jupiter

receiver
is
detecting Jupiter activity
is
is
Tj =
negligible in comparison to Tg-
Ta
The
system sensitivity must remain constant.
standardized value of
the apparition during
apparitions one
whenever Tj
is,
is
The minimum
minimum
which the
would
to the galactic
Tg-
The
It
noise temperature of the
sensitivity of the radio telescope for
less than
therefore necessary to establish a
T min
antenna temperature
the Jovian activity
detectable antenna temperature
shows a number
galactic
galactic
background
at 18
MHz
background temperature.
low antenna temperature
is
assumed
to
less than
until a
will
to distinguish these pulses as
other
all
Tm n
That
be zero.
plotting a
number
one season.
Figure
was estimated by
above
This figure shows that the counts increase towards
peak
all
is
reached. This turnover point in our distribution
bursts.
be hidden
in the
certain fraction of the bursts at such a
noise and the observers
DAM. The minimum
may
not be able
estimated from this curve to be about 10,000 Kelvin.
In
defined for
1986 season. This season had a very low average
for the
where we can no longer detect
galactic
distribution for the antenna temperature of Jupiter bursts
lower antenna temperatures

is
Tm n
background temperature was highest. In
galactic
distribution of antenna temperatures for bursts over the course of
2.8
which
comparison of occurrence probability,
was
detectable Jovian antenna temperature,
reject all pulses of Jovian
and
background alone,
thus limited by the galactic background temperature,
is
different in different apparitions. For long-term
the
due
This value
is
for years of
is
low
Table 2.1 are the average galactic background temperatures for each season
20 and 22
by which
MHz
using a Yagi antenna of beamwidth approximately 50.
we account
for both the galactic
background changes and
at 18,
The method
the changing Earth-
27
Jupiter distance
is
designated
Tmax
is
is
The average of
as follows.
T max
the average of the temperatures
is
the average of the 1986, 1967-68
how much
and 1975-76
apparitions.
higher the galactic background temperature
ratio multiplied
background temperatures
The
ratio is
is
20 MHz. The high
Kelvin
at
Yagi
33,740 Kelvin. For 22
minimum
galactic
The
3.132 for the 18

will
MHz
assume
background temperature cutoff
MHz the minimum
describes
ratio
Yagi. This
for the years of
Kelvin and the high galactic background cutoff
is
about 10,000
for the
is
T min
on average during years when
by 10,000 Kelvin represents the cutoff we
high galactic background. The
Tm n
is
from the 1972, 1984 and 1985 apparitions.
Jupiter appears near the galactic center.
is
the three highest
Similarly the average of the three smallest temperatures
20
MHz
is
5,000
15,270 Kelvin. For those seasons in
which no calibrations were performed, namely before the 1966 season and
after the
was estimated based on seasons 12 years
earlier or
1992-1993 season
later.
one
the corrected O.P.
Jupiter will appear with the
orbital period earlier or
same region of
was
background as
it
had
12 years.
In order to correct for distance effects the
for each pulse
the galaxy in the
background subtracted antenna temperature
multiplied by a distance correction factor. That
is
(2-2)
where
and
rmin
is
is
the distance to Jupiter in Astronomical Units at the time of each Jupiter storm
equal to 4 A.U.
By
selecting only those Jupiter bursts in every season
distance corrected antenna temperature exceeds this cutoff,

the effects of high galactic
we
whose
will effectively simulate
background noise levels for every season and correct for the
changing Earth-Jupiter distance. This will of course reduce the amount of data used
compute
to
the occurrence probability. Table 2.2 gives the occurrence probabilities for the
28
18
MHz
Yagi using
all
of the activity data each season and the occurrence probabilities
using the temperature cutoff of 31,320 Kelvin. Figure 2.9 shows the seasonal occurrence
probability for
bursts detected by the 18
all
MHz
Yagi and the occurrence probability
with the distance and galactic background correction. The incentive to check the effect
of galactic background levels on occurrence probability
that the rise
and
fall
of the occurrence probability with
came when
De was
occurrence probability tends to rise sharply with increasing
The peak
after the
in the galactic
minimum
in
De
background occurs during the sharply
De- Figure 2.9 shows
it
was recognized
asymmetric. That
is,
the
and drop more gradually.

rising period about
that with the correction the trailing
year
edges
of the occurrence probability curves are reduced slightly more than the leading edge.
The asymmetry was not completely removed however. The procedure

the background-limited occurrence probabilities
was found
in the
program used
to
correction
was not performed on
was about
5%
compute
the
is
the antenna temperatures
too high. Both of these errors caused too
was corrected and
however
all
in
of the tables and plots
is
The
error
distance
pulses to be rejected and
small however, less than
in this
An
2.3.
and the temperature cutoff
many
7%. This
chapter have been corrected
the occurrence probabilities used in the following chapters
for distance
for determining
Table
background corrections.
reduced the occurrence probabilities. The difference

error
summarized
and have a cutoff temperature somewhat too high.
were not corrected
Figure 2.8:
Number
distribution of
DAM
bursts in 1986 at 18
MHz
in
terms of the temperature
of each burst above the galactic background antenna temperature. This season had a low average
galactic
background antenna temperature.
30
Table 2.1: Seasonal averaged galactic background temperatures for
Season
Q MriZ
\AXJv
lo
TIq
ZU MriZ
18,
20 and 22
2.1
MHz
1965-66
108628
80489
84958
1966
77156
103818
56378
1967-68
771 12
57107
19046
968-69
941 13
51484
39785
1969-70
105132
76532
56219
1970-71
177637
128563
96070
1972
305839
168039
104293
1973
134429
102242
65616
1974-75
91334
70490
40964
1975-76
84103
66185
35058
1976-77
86771
67644
40224
1977-78
104053
67603
41007
1978-79
92493
60818
32386
1979-80
93634
57406
34364
1980-81
102598
82104
44988
1981-82
10138
53787
15781
Yagi.
MriZ Iq
88354
1983
164269
141848
1984
256902
224300
1985
183008
97038
1986
76836
49432
34200
1987-88
83445
63003
40888
44430
10077
57713
1988-89
89341
67353
1989-90
103128
79084
47596
1990-91
81442
42978
40596
1991-92
87572
93690
39456
1992-93
12401
55735
56473
Table 2.2: Seasonal averaged occurrence probabilities for 18
Season
Uncorrected 0.
P.
MHz
Yagi 1957-1975.
Corrected O.P.
T min = 31.320K
.03006
1957-58
.03133
1958-59
.00820
.00812
1959-60
.05404
.02656
1961-62
.02309
.01945
1962-63
.12873
.11498
1963-64
.15170
.11530
1964-65
.07853
.05961
1965-66
.08658
.06456
1966
.07911
.07601
1967-68
.05000
.04872
1968-69
.04338
.04317
1969-70
.03362
.03263
1970-71
.00979
.00971
1972
.02790
.02790
1973
.06268
.06137
1974-75
.11146
.09666
1975-76
.15960
.13417
Table 2.3: Seasonal averaged occurrence probabilities for 18
Season
Uncorrected O.P.
MHz
Yagi 1975-1994.
Corrected O.P.
T min = 31,320K
1975-76
.15960
.13417
1976-77
.11883
.095
1977-78
.12757
.11307
1978-79
.06612
.05890
1979-80
.05931
.05324
1980-81
.03919
.03654
1981-82
.02824
.02454
1983
.01720
.01582
1984
.02035
.01874
1985
.12662
.10330
1986
.15577
.11051
1987-88
.12304
.09891
1988-89
.09734
.07196
1989-90
.04639
.03665
1990-91
.08080
.06902
1991-92
.05077
.04244
1992-93
.05029
.04621
1994
.03207
.02660
33
Figure 2.9:
The curves
effects.
plot of the occurrence probability for all the data collected with the 18
are the uncorrected O.P., and the O.P. corrected for distance and galactic
MHz
Yagi.
background
34
Table 2.4: The procedures used to obtain galactic background-limited occurrence probabilities
Method
according
1.0
2.
for obtaining seasonal averaged occurrence probabilities
Determine the number of hours of
/.
For each
galactic
to the
five
listening.
observer's log and
when
This includes deleting those times when the conditions went below
the
equipment was not functioning properly.
minute interval of activity, determine whether the antenna temperature exceeded the interpolated
background found from the galactic background values taken before and
after the radio storm,
from
this
determine the galactic background subtracted antenna temperature, Tj.

3.
Correct for changes in distance by multiplying each Tj by a correction factor.
4.
Correct for changes in the galactic background levels, for seasons from 1966 through 1993.
minute intervals of activity
five
exceeds the threshold

5.
in
Sum
which the distance corrected antenna temperature of the Jupiter
only those
burst, T'j
level.
For the seasons from 1984 through 1994. multiply the weekday
by 2.013 to correct for observer
activity time
bias.
6.
Divide the sum of activity over the season from step
the listening determined in step

7.
I.,
this is the galactic
3.
(or
if
the year
For the seasons outside of the range 1966-1993 find a season 12 years
determine the
ratio
this correction.
is
between 1984-1994) by the sum of
background-limited occurrence probability.

later
which does
fall in
the range and
of the galactic background-limited occurrence probability to the occurrence probability without
Multiply the occurrence probabiltiy for seasons outside this range by this factor.
2.2.6 Calibration Procedure
The
first
step in the calibration process
into decibels of attenuation
Jupiter
converting
all
of the chart record deflections
from a standard noise generator. Each chart record
was present should contain
to decibel attenuation
is
a calibration scale.
The least-squares best
in
fit is
which
found
and chart record deflection response. This provides a functional
form for which one can determine for any deflection the corresponding decibel attenuation
against the standard.
The program used
second order polynomial
study
in this
to the calibration curve.
tries
Whichever
to
fit
fit
a fourth, third and
had the smallest sum
of squared residuals was chosen to describe the calibration curve. Figure 2.10 shows a
typical calibration
and the best
background deflections
fit
curve. This function was used to convert the galactic
into decibels.
Each peak
deflection for every 5 minute interval
during the storm was also converted into decibels.
In order to
compute the antenna
35
temperature due to the galactic background and Jupiter the attenuation and gains in the
line
between the antenna and the calibration point must be known. Appendix
line losses
To
and gains for the
background
is
MHz
found by
end of the storm
fitting a line to the
interval.
two
to Jupiter alone
is
interval is subtracted
flux density
is
in
shown
Figure 2.1
Figure
in
2.
lists
the
for a typical storm.
galactic
background temperatures
galactic
The antenna temperature of
the antenna temperature of the highest peak each 5 minutes
background are shown
Yagi antennas from 1981 through 1992.
background noise from the Jupiter noise an interpolated
subtract the galactic
either
22
18, 20,
at
the galactic background,
and the interpolated galactic
The antenna temperature due
2 where the antenna temperature for each 5 minute
from the interpolated
galactic
background for that
interval.
The
found from the background subtracted antenna temperature, the antenna
effective area and
Boltzmann's constant. Appendix
by which the Jupiter storms were
calibrated.
A provides
an example of the method
36
40
'
'
'
100
80
60
40
20
DEFLECTION
Figure 2.10: Sample calibration curve. The points are taken from a calibration done on June
MHz Yagi. The best

HP 461 A noise generator with
1987 for the 18

off of a
line is
fit
580
a fourth-order polynomial.
The
attenuation
1,
was
a noise temperature of 62.7 million Kelvin.
600
620
640
660
UT (minutes)
Figure
2.
Plot of antenna temperature for the highest peak every 5 minutes. This storm
observed on June
1,
1987 with the 18
MHz Yagi antenna.
The
background antenna temperature determined for the beginning and end of the storm
first
order polynomial was found from these two points and
was
solid triangles represent the galactic
is
shown
in the figure.
interval.
37
3xio*
600
580
'
'
620
-i
640
660
UT (minutes)
Figure 2.12: The galactic background subtracted antenna temperature of a Jupiter decametric
storm.
the
peak for every 5 minute
interpolated galactic background over that interval.
interval
is
subtracted from the
CHAPTER
RESULTS
A New
3.1
Determination of
DAM
Source Locations
Before proceeding with a study of the effect of

emission
it is
sources in
range of
CML and Io phase space.
CML and
De changes on
Table
MHz
these sources are defined at
The
3.
data set from 1964-1994 and the 22
final
10%
as well,
This definition of the decametric source boundaries

than
is
MHz
from 18
Yagi data set from 1957data set from 1957-1994.

at
each
values for the source locations were chosen to span the sources at
this data set
will enclose the source regardless of
it
MHz
of the peak value for each source
each of the three frequencies mentioned above. Since
DE
the Jovian decametric
provides a consistent and uniformly defined
data used for these definitions are the 18
1994, the 20
of
and Io Phase
Io phase coordinates for all the sources apparent in the range
The edges of
frequency.
CML
enlightening to investigate the average location of the Io- and non-Io-related
MHz. The
to 22
in
the widely used
one reproduced from Carr
is
any
spans the entire range
shifts in
CML
or Io phase.
based on a more extensive data
et al.
[1983] as Table
set
1.1
3.2 Occurrence Probability and Average Flux Density
Two methods
make an
for specifying the "activity" of a source are considered.
One
is
to
estimate of the occurrence probability of the Jovian emission over an indicated
time period. The other
is
to estimate the average Jovian flux density (with the galactic
background subtracted out) over the indicated time period. Results are presented
section of a test
made
to
in this
determine which of the two methods of specifying Jovian activity
38
39
is
more useful
the
in investigating
long-term effects such as the activity variations with
respect to De-
Table 3.1: Source locations
span the sources
in
CML
18 through 22
at
and Io phase taken
MHz
and
all
at the
10%
of peak
level.
Source Designation
CML
Io-D
75 200
105 136
Io-B
75200
60 105
non-Io-B
105 180
Io Phase
Range
Range
Io-A
190288
225 270
Io-A'
200288
186 225
non-Io-A
205 288
Io-C
288 -^0
non-Io-C
288 350
The occurrence
210 260
probability at the specified frequency
is
defined as the ratio of the
time during which there was detectable Jovian activity to the
total
"listening" for activity.
The average
These ranges
values of De-
flux density at the specified
total
time of effective
frequency
is
defined as
the time average of the background-subtracted Jovian flux density over the total effective
listening time.
It
should be noted that the average flux density, so defined,
flux-density-weighted occurrence probability;

carry
more weight
in
computing
i.e.,
is
simply the
the emissions of higher flux density
the weighted occurrence probability than
do those of
lower flux density.
It
was necessary
because
it
was not
to
to simplify the data reduction process
make approximations
feasible to
measure every Jupiter
burst.
Instead, the highest value
of antenna temperature was read and recorded for each 5-minute interval of listening
time.
For the results presented
flux density
of De-
The
were
in this section, the
calculated in turn for 5 bins of
calculation procedure for each bin
is
occurrences probability and average
CML,
5 bins of Io phase, and 1 bins
summarized
here. For a given bin,
let:
40
CML
Figure 3.1:
An
illustration of the source regions defined in
Table 3.1
41
n =
the total
= TA
Tj,
number of 5 minute intervals of listening time and
<
< n
Tg, the peak value of Jovian antenna temperature

(adjusted for line loss) for 5-minute interval
where
T Al =
antenna temperature and
total
i,
Tg =
temperature of galactic background
= peak value of Jovian
S,
flux density for 5-minute interval
where
= (2/A e )Tj where k = Boltzmann's constant and
S,
Ae
= antenna
effective area.
(5)
Si
all
Tmin
average flux density,
galactic
minimum
standardized value of
temperature,
Pi
if
T A( ^
i.e.,
the average of the
peak values over
of the 5-minute intervals of listening time.

detectable Jovian antenna
the value determined for the apparition of highest
i.e.,
Tmi n Pi = 0
;
if
T Al < T min
P=
Ep =
What
be called the occurrence probability
will
in this
background limited occurrence probability as defined
chapter
in
is
Chapter
actually the galactic
2.
The occurrence
probability and the flux density are of course not entirely independent of
The
stronger the bursts are the
3.3 however, non-Io-B
source. Figure 3.2
is
more
shown
shows a
likely
it
is
In Section
be a low flux density but high occurrence probability
to
plot of occurrence probability
and Io phase for the 1987-1988 apparition

a strong correlation between the
one another.
that they will be detected.
two
at
18
and
MHz. As
pairs of curves.
flux density against
expected, there
is
CML
clearly
For the analyses of long-term
variations of Jovian activity to be presented in subsequent sections of this dissertation, a
decision had to be
the
measure of
which of
the
made whether
activity.
two
The
quantities
to use occurrence probability or average flux density as
plots in Figures 3.3
would be
the
and 3.4 were used as a
more
test to
determine
suitable for revealing the sought-for
42
effects. Figure 3.3 is a plot
of the
latter,
of occurrence probability as a function of
and Figure 3.4
of occurrence probability.
is
the
It
is
smoother variation with respect

flux density appears to
will therefore be
same except
that average flux density
for 1 intervals
is
used instead
clear that the occurrence probability plot indicates a
to
De
than does the average flux density plot. Average
be more subject
to
random
fluctuations.
Occurrence probability
used more often than average flux density as the activity indicator
the plots
which follow. The reason
density
is
not known.
system
losses, or alternatively
It
for the larger
might be largely due

it
might
result
random
violent intensity increase
random
is
and
from an actual randomness

latter idea,
in correcting for
in the radiation
an unexplained sudden
often observed during a noise storm; such events
would tend
fluctuations in average flux density.
in
fluctuations in average flux
to errors in calibration
intensity during periods of activity. In support of this
to increase the
De
<
'
O.P.
Flux density
200
100
300
CML
Flux density
A_ A
f
200
100
>
lo
L.
300
Phase
Figure 3.2: Comparison of occurrence probability and flux density against central meridian
longitude and lo phase at 18
MHz
during the 1987-88 apparition.
43
0.25
0.2
non-Io-A
XI
a 0.15
O
0.1
0.05
Figure 3.3: The
De dependence
MHz.
The
flux density at 18
MHz.
degree bins of De-
The
of the non-Io-A occurrence probability at 18
occurrence probability for apparitions from 1957-1994 are in
degree bins of De-
4x10
3x10
5x10
u
-j>
,0
>
<
Figure 3.4:
The average
histogram illustrating the
DE
flux density for apparitions
definition of average flux density
is
given
dependence of the non-Io-A

from 1965-1993 are
in the text.
in
44
Source Widths and Locations
3.2.1
In order to investigate the variation in source location
of
DE
and
to
improve the
seasonal averaged
De
each
De
statistical
and width over the range
confidence, the apparitions were grouped by their
value. Table 3.2 lists the apparitions that
were combined
group. Figure 3.5 illustrates the changes in occurrence probability for
to
all
form
of the
Io-related and non-Io-related sources with frequency and De. All the sources appear to
be influenced by
to
De
at
each frequency and most of
this
chapter
Table
The
3.2:
division of apparitions into groups of similar DEvalues.
ranges and the apparitions which are combined for each of these
De range
IH
MHz
57-Si iyoD.
n
20
MHz
wit,
69-70. 70-71,
81-82, 83, 92-93,
81-82. 83. 92-93,
94
94
-2.5
<
DE <
-1.0
59-60, 68-69,72.
79-80, 80-81, 84,
68-69, 72, 79-80.

80-81, 84, 91-92
91-92
-1.0
1.0
>
is
what extent the source characteristics are influenced by changes
<
<
2.5
DE
DE
<
<
1.0
2.5
61-62. 67-68. 73,
67-68, 73, 78-79,
78-79, 85. 90-91
85, 90-91
62-63, 65, 66, 74,
65, 66, 77-78, 86,
77-78, 86. 89-90
89-90
63-64. 64-65,
64-65, 75-76.
75-76, 76-77,
76-77, 87-88,
87-88, 88-89
88-89
devoted to studying
in
The
De-
table gives the
sets.
11 MHz
^ame^^^TTTTz
De
45
18
MHz
Figure 3.5: Contours of constant occurrence probability
CML
5
De
(horizontal axis) plane. This plane
ranges (rows).
The De groups
22 MHz
20 MHz
is
in the Io
phase (vertical axis) versus
represented for each of 3 frequencies (columns) and
are the
same
as those listed in Table 3.2
46
Figure 3.6:
The
Same
90
270
180
same
The
now
180
270
360
90
'80
270
360
the contours for each represent average flux density.
as in Table 3.2 except excluding seasons before 1965.
Four points have been used

Io phase histograms.
90
360
as Figure 3.5 but
E>e groups are the
22 MHz
20 MHz
18 MHz
first
to describe the
point
is
source width and shape in the
the location of the
CML
and
peak of the source. The height
47
of the peak
used to determine the two full-width half
is
edge and low edge. The width of the source

For the non-Io-A source and
MHz
all
Merging
is
the data
from the three
activity recorded at 18
frequencies would not be
maximum
peaks and half
in 3 bins
The 26.3
MHz
antenna
is
Low
filled
than at 20 or 22
1.1.
The
CML
and Io phase
was smoothed and rebinned
Intensity
array
to 1.5.
High Probability Non-Io-B Source
composed of 640
dipoles. each dipole oriented
in the east-west direction. There are 16 dipoles in the east-west direction and
The bandwidth of the array
north-south direction.
observations
latter value
is
it
is
is
about
MHz,
the overall system bandwidth.
The
et ai, 1975].
For
this study, the
26.3
sensitivity limit of the array
MHz
at the
in
to the format
used for the Yagi data as shown
in
in the
MHz. The
is
about 100
same frequency
array data format, referred to as the
Thieman's dissertation [Thieman, 1977], was converted
raw data card
40
but for Jupiter
usually used with a receiver having a bandwidth of 0.25
Janskys compared to 60,000 Janskys for the Yagi antenna system
[Desch
MHz.
but in order to improve our ability to identify the
points the data
3.3 Verification of the
MHz
useful since the location of the
sources are frequency dependent as can be seen in Figure
histograms were originally
points, called the high
between these two points.
determine source width and locations. The reason
to
more
that there is
maximum
the difference
of the other sources studied in this report the data at 18
have been used exclusively
for this choice
is
Figure 2.1.
into
one similar
All of the data for each
apparition were used without regard to the storm characteristics or credibility tags given
in the
raw data
method of
the
file.
26
The dramatic
relatively
MHz
thorough discussion of the instrumentation and data reduction

array data
is
given
in
Desch's dissertation [Desch, 1976].
increase in sensitivity of this instrument
weak components of
makes possible
the Jovian decametric radiation.
From
the study of
the data obtained
48
during the
first
apparition in
1973 apparition), Desch et

that
which
al.
was used
the large array
for Jupiter observations (the
[1975] discovered a high probability non-Io-B component
had not previously been recognized because of the low peak
bursts.
These
flux densities
were found
to
be considerably
flux densities of
its
less than those for the bursts
of Io-B, Io-A, Io-C and non-Io-A. However, there has yet been no confirmation of the
Desch
al.
et al.
finding,
paper. This section
the
al.
MHz
high-De CML-Io phase

at
Desch
et
array data obtained during the 7 apparitions
paper. Table 3.3 summarizes the large array data used.
should be noted that the Yagi-obtained data
non-Io-B region
De
a report on a complete confirmation of the
based on the large 26.3
following that of the Desch et
It
is
plots a very
MHz. There
is
in
Figures 3.5 and 3.6 indicate for
low probability and low
essentially
groups. At the higher frequencies, there
is
no
flux density for the
activity for the
virtually
corresponding Iow-
no detectable non-Io-B
Figures 3.7 to 3.15 for the large-array data, on the other hand,
show
activity.
relatively high
show
for each
of the apparitions a plot of contours of constant occurrence probability on the
CML-Io
occurrence probabilities for the non-Io-B data. Figures 3.7 through 3.14
phase plane, together with histograms of occurrence probability
vs.
CML
and occurrence
probability vs., Io phase. Figure 3.15 contains similar plots for the aggregate of
data, together with the
corresponding Yagi-obtained plots for comparison.
figure best indicates that the conclusions of
Desch
et al.
[1975] were valid.
all
This
the
latter
49
Table 3.3:
An overview
of the 26.3
MHz
array data used in this study.
Begin
End
Listening
Activity
Occurrence
Date
Date
(hours)
(hours)
Probability
1973
4-16-73
12-30-73
529.333
82.583
0.156
1974-75
1-23-74
2-28-75
476.583
81.167
0.170
1975-76
6-28-75
4-27-76
484.667
81.917
0.169
1976-77
9-27-76
5-19-77
435.167
59.000
0.136
1977-78
10-24-77
5-4-78
214.167
48.583
0.227
1978-79
11-13-78
4-10-79
339.250
42.917
0.126
1979-80
11-21-79
4-10-80
281.750
24.417
0.087
1981
1-5-81
4-28-81
236.500
18.583
0.078
Season
System
Figure 3.7: 26.3

ai,
1975.
MHz
III
Longitude
Array, 1973 apparition. This plot
is
of the same data used
in
Desch
et
50
Figure 3.9: 26.3
MHz
Array, 1975-76 apparition.
51
40
System
III
Longitude
Figure 3.10: 26.3
MHz
Figure 3.11: 26.3
MHz
52
In:
U
Occurrence Probability
;
-j
'
'
i
-0
iii
0
60
120
System
180
III
240
...
300
Longitude
Figure 3.12: 26.3
MHz
Figure 3.13: 26.3
MHz
308
20
...
40
...
60
..
80
53
MHz
Figure 3.14: 26.3
Array, 1981 apparition.
MW
26 MHz Arroy
22 MHz YAGI
20 MHz YAGI
'
^^^^
0
90
180
CML
Figure
3.1 5:
and the 26.3
270
360
90
180
CML
270
380
90
360
90
180
270
360
CML
A comparison of occurrence probabilities between 8, 20 and 22 MHz Yagi antennas

MHz array for the integrated apparitions from 1973 to 1981. The top row are CML
1
histograms of occurrence probability and the bottom row are

contours are
270
180
CML
at
occurrence probabilities of 10, 20, 40, 60 and 80%.
plots.
The
54
3.4
Of
all
the sources
non-Io-related
shows
changes
that the
The De Effect on
shown
in the
in
strong dependence on De-
fit
best
fits
it
is
the
source width and location for non-Io-A are very large over the
De
CML
The occurrence
edge of the non-Io-A source has a very
probability of the non-Io-A source varies in
as seen in Figure 3.20. Figure 3.16 through 3.18
lines of the different sections of non-Io-A.
best
plots of Figure 3.5
source which suffers the most dramatic changes with De- Figure 3.19
roughly 7 degree range of De- The high
close step with
Non-Io-A Source
the
to the source using data
The
fit
lines are
the data and
lines in these figures represent the
from every apparition
accurately be determined. These best
show
in
which these locations could
reproduced
in
summary
in
Figure 3.19.
300
280
I)
-ao
Ed
_:
ap
X
w
260
-J
240
-4
-2
CML edge of the non-Io-A source at 18 MHz for each apparition. The
CML edges are defined by the full-width-half maximum longitude of the occurrence
Figure 3.16: The high

high and low
probability.
The
error bars are proportional to the inverse square root of the
time for the non-Io-A source during each apparition.
amount of
listening
55
Figure 3.17: The
CML
260
of the peak of the non-Io-A source
at 18
MHz
non-Io-A
<i>
00 240
T3
w
s
c
-J
a
220
"
200
0
Figure 3.18: The low
CML
at 18
MHz.
for each apparition.
56
300
High Edge
Peak
Mid-Point
Low Edge
220
0
-2
-4
3E
Figure 3.19:
the
same
The
best
fit
lines to the
of these lines are presented
The degree
depends on
De
to
is
hypothesis of no
De dependence
peak, midpoint and edges. The lines are

for the slopes
and the
errors in the slopes
which the shape of each of the sources and the occurrence probability
Spearman's rank correlation coefficient [Press
performed
De
to see
dependence
in the data.
De
in
effect
3.4
etal., 1992].
whether the data allow for the rejection of the

favor of the alternative that there
Spearman's rank correlation coefficient
Appendix B. Table
been tested for the
CML
The values
Table 3.5
in
tested using
statistical test is
detail in
non-Io-A
as those in Figures 3.16-3.18.
lists
is
is
null
some degree of
described
in
more
each of the characteristics of non-Io-A that have
and the likelihood of a
De
dependence. The three columns
labeled Very Likely, Possible and Inconclusive expresses the degree of confidence in
the rejection of the null hypothesis of
no De
effect.
We
perform a Student's
the probability of obtaining a nonzero correlation coefficient
zero.
A De
dependence
that
is
very likely
is
chosen when
when
we can
the null hypothesis can be rejected in favor of the alternative of
test
the true correlation
be
95%
of
is
confident that
some De dependence.
57
possible
De dependence
A De
null hypothesis.
dependence
80%
even the
null hypothesis at
one where
is
that
we
can be 80-95% confident of rejecting the
inconclusive
is
is
confidence interval. The
rank correlation coefficient, the second column
one where we cannot

first
column
is
rs
the degrees of freedom. Table 3.5
is d.f.,
The
gives the slopes of those lines which had a very likely correlation with Deline to the longitude of the
both
fits
in
reason for
non-Io-A peak was computed twice and the slope
Table 3.5. The second

this is that there is
excluded data with
fit
evidence presented
De
source has a different dependence on
reject the
Spearman's
De
is
best
greater than about 3.
later in this section that the
fit
given for
The
non-Io-A
high end of the range.
at the
non-Io-A
_
'
I 200
0
f 100
-fl
1
i
hz
2
a"
0
-2
0-3
h-
18
0.2
>:
'
'
/A
MHz
o
0.1
"'\
>
*\
0.15
a-
/ \
20 MHz
0.1
0.05
>
:
,
0.13)
I
1
"-T*
X K
X.
h-
22 MHz
0.1
*-\
cu
o
0.05
V-K
1990
1980
1970
1960
Year
Figure 3.20: Plots of seasonal averaged sunspot number,
20 and 22
MHz
for non-Io-A.
The
vertical lines
mark
De and
occurrence probability at 18,
the dates of highest De-
58
Table 3.4: Confidence of
is
De dependence on non-Io-A
Spearman's rank correlation
coefficient, the
source characteristics.
second column
is
The
first
column
the degrees of freedom.
De Dependence
d.f.
rs
non-Io-A
Probability
Probability
Probability
@
@
@
CML
of Peak
High
CML
20
22
MHz
MHz
MHz
Edge
Low CML Edge
Very Likely
Possible
7J
on Q<cr.
/o
OU-7J
.87
31
.80
25
.76
31
.51
26
.69
26
-.27
26
10
Inconclusive
SOU 10
Table 3.5: The slopes and errors for those non-Io-A source characteristics demonstrating a very
likely
dependence on De-
De Dependence
degree binned data
Slope
Standard
0.0209
0.0056
non-Io-A
Deviation
Probability
CML
CML
18
MHz
of Peak
2.13
3.88
of Peak without
5.45
0.71
CML
4.15
1.94
notch
High
Edge
59
The
3.5
Effect on the Other
DAM
Sources
Io-B
3.5.1
A CML-Io
plot
DE
show
phase plot of Io-B reveals a complicated structure.
that of the
is
1965-66 apparition shown
in
typical Io-B contour
Figures 3.21-3.23. Nearly
all
a relatively sharp cutoff of the Io-B source at high Io phase and a
drop toward lower Io phase.
CML
source Io-B in the
The
features labeled /
[1978]
The
CML was
bifurcation in
No
of the Io-B source but
//
more gradual
appear as a bifurcation of
and Io phase coordinates. Both of these features were noted
by Leacock and Smith [1973], Thieman, Smith and
and Barrow [1981].
and
apparitions
May
[1975] and Thieman and Smith
noted by Gallet [1961], Smith, Carr and Six [1962]
physical explanation has been presented to explain the splitting
it
is
a recurrent feature confirmed by our data.
Feature
/// is
a bridge of enhanced occurrence probability apparently connecting the Io-B source to
Feature IV
the non-Io-A source.
DE
and high
but
the Io-D source
is
is virtually nonexistent in the
close-up of Figure 3.21
now showing
which appears weakly
20 and 22
MHz
at
18
MHz
Figure 3.23
data.
is
only the Io-B source. These plots are integrated
over several apparitions and therefore some of the finer details have been smoothed. The
contours
in
Figure 3.22 represent the average flux density. The overall shapes match quite
closely the occurrence probability plots of Figure 3.21 but the relative importance of the
central peaks of the Io-B source
is
the bifurcation of the Io-B source.
Similar to the
source.
made
Non-Io-B
to study the
for the peak,
source, there
is
DE
very
greater and the
The
is
flux density
CML
histogram shows more clearly
histograms have not been smoothed.
also an associated non-Io related
weak compared
dependence of
this
to the other
DAM
component
sources.
No
source but for completeness the
low and high edges of the source are shown
in
Figure 3.39.
to the
attempt was
mean
values
60
30
I
*o
25
y\
| 20
.o
Occurrence ^robao^tv
0
ISO
120
40
70
MO
100
System
Figure 3.21:
A CML-Io
apparition at 18
Hi
160
190
223)
,n
20
30
40
50
Longitude
phase occurrence probability contour plot of Io-B for the 1965-66
MHz. The
histograms
in
CML
and Io phase can be interpreted as projections
of the contour plot along the two coordinates.
40
70
100
130
160
System
Longitude
190
9
22e5.0u.'oH%iQ% a'sio
,
III
Figure 3.22:
Same
histograms are the

(IO"
26
Hz)
as Figure
sum
3.21
but
now
contours represent average flux density.
of the flux density received within each bin.
Units are
in
The
Janskys
61
18 MHz 1965-66
40
60
Figure 3.23:
An
of the source.
80
100
120
140
160
180
200
220
240
enlargement of the contour plot of Figure 3.21 illustrating recurrent features
62
1
50.0
22 MHz
20 MHz
MHz
18
1
'
37.5-
0.0
40
85
CML
175
38D
85
130
175
Table 3.2
2HE
85
in the Io
130
'
175
220
The region containing the Io-B source is represented for each of

De ranges (rows). The De groups are the same as those listed
3 frequencies (columns) and 5

in
130
Figure 3.25:
are the
same
Same
as Figure 3.24 but
now
for contours of average flux density.
as in Table 3.2 except excluding seasons before 1965.
The De groups
64
200
Io-B
High Edge
Peak
150
Mid-Point
100
Low Edge
D,
Figure 3.26:
The
best
fit
CML
lines to the Io-B
'
1
'
Io-B
100
High Edge
<L>
q0
a.
Peak
Alid Point
ao
\Low
.
1
,
70
Figure 3.27: The best
fit
Edge
lines to the Io-B Io phase peak,
midpoint and edges.
65
Table 3.6
lists
source. Figure 3.26
increasing De-
De dependence
shows
that the
for both the
low
The De dependence of
CML
CML
of the peak point
are the
low and high edges. The midpoint
and
small slope reflects the opposite slopes of the two edges.
its
edges of the Io-B
edge tends towards lower longitudes with
CML
the
low and high
is
computed from
the source drifts toward lower Io phase with increasing De-
The
is
not as certain as
low and high edges
the
Figure 3.27 shows

slightly higher slope
of the low edge might be due to the apparent growth of feature / with
De
lines aid the
rise in
De
as seen in
Figure 3.28 compares the time series of the seasonal averaged sunspot
Figure 3.24.
number,
De
and occurrence probability for Io-B

eye
in
matching features
in the
at
18,
20 and 22 MHz. The
curves to the dates of
maximum
vertical
De- The
appears to be matched with a rise in occurrence probability except for a drop
in probability at the
Figure 3.28
is
maximum De
computed from
the
values.
monthly
The seasonal averaged sunspot number
R number
in
obtained from the National Solar
Observatory /Sacramento Peak. Only those months which span the observing season are
used to compute the average.
20 and 22
MHz
for Io-B.
The
vertical lines
mark
and occurrence probability
at
18,
67
Table
3.6:
Confidence of
Dg dependence on
Io-B and non-Io-B source characteristics
De Dependence
d.f.
rs
Very Likely
Possible
Inconclusive
>95%
80-95%
<80%
lO-D
Probability
Probability
Probability
CML
18
20
22
MHz
MHz
MHz
of Peak
Io Phase of
High
@
@
@
CML
Peak
Edge
Low CML Edge
.59
31
.33
25
.55
31
.26
27
-.56
26
X
X
X
X
X
.68
27
-.40
27
High
Io
Phase Edge
-.88
21
Low
Io
Phase Edge
-.62
21
Table
likely
3.7:
The slopes and
errors of those characteristics of the Io-B source demonstrating a very
De dependence
De Dependence
degree binned data
Slope
Standard
0.048
0.023
-1.3
1.2
Io-B
Deviation
Probability
Io Phase of
Peak
18
MHz
CML Edge
Low CML Edge
High
Phase Edge
High
Io
Low
Io Phase
Edge
2.6
1.2
-2.5
1.2
-1.1
0.6
-1.8
1.2
68
3.5.2 Io-A
and Io-A'
The Io-A source on

source.
Both sources tend
CML-Io phase diagram shows some
and 3.29, both sources appear

Io-A often appears
to
similarity with the Io-B
have a steep high Io phase edge. Comparing Figures 3.24
to
to
become
less
have secondary peaks
asymmetric, more rounded with lower De-
at
lower Io phase. These peaks
at
times are
of such high occurrence probability to be nearly equal to the main Io-A peak in detection
likelihood.
Included
among
these secondary peaks
may be
Io-A'
itself.
simply an extension of Io-A. The Io-A' region
the Io-A' region
is
several peaks
of which are of nearly equal occurrence probability.
all
is
It
may be
that
ill-defined, with
As De
increases
the non-Io-A component becomes stronger and isolating a single peak as the main Io-A'
peak becomes increasingly
Figure 3.31 shows Io-A, Io-A' and non-Io-A
difficult.
in Io
phase. This figure illustrates the difficulty in identifying permanent Io-A' landmarks and
at least in the Io
studying trends
phase component.
From Table
peaks outside of the
3.1,
range from about 180 to 270 degrees are designated the non-Io-A emission.
Io phase of the
edge seems
Io-A source with
to indicate a definite
De
is
Only the high
not clear in this data.
dependence with De- From Table
3.8, the
drift in
Io phase
low Io phase
edge shows a possible dependence and the peak shows no dependence on De- The
De
dependence on
the high
CML
probability of Io-A
The occurrence
a drop at the
CML
shows
that
edge of Io-A can vary by over 30 degrees over the 7 degree span of De-
The occurrence
3.34.
for the Io-A source appears to be quite strong. Figure 3.33
is
also clearly dependent
probabilities tend to rise with higher
maximum De
values.
on De
De
as seen in Figure
but there
is
evidence of
69
10
180
208
235
22 MHz
20 MHz
18 MHz
262
280
208
235
262
280
CML
each of 3 frequencies (columns) and 5

listed in
Table 3.2
De ranges
.2
in the Io
the Io-A, Io-A
(rows).
208
235
262
290

'
source
The De groups
is
represented for
are the
same
as those
Io
Phase
Figure 3.31: Plots of occurrence probability versus Io phase
DE
These plots are integrated over
degrees Io phase and the Io-A' peak
CML
is
from 200
to
at
18
MHz
for various ranges of
290 degrees. The Io-A peak
between 200 and 220 degrees.
is
at
240
73
Io-A
20 and 22
MHz
for Io-A.
The
vertical lines
mark
De and occurrence
the dates of highest
DE
probability at 18.
74
De dependence on Io-A
source characteristics
De Dependence
rs
1
Probability
Probability
Probability
@
@
@
Very Likely
Possible
Inconclusive
>95%
80-95%
<80%
18
MHz
.46
31
20
MHz
MHz
.38
25
22
CML of Peak
Io
d.f.
rY
Phase of Peak
.52
31
.77
26
-.14
17
CML Edge
Low CML Edge
.86
26
.29
26
High Io Phase Edge
.58
17
Low
-.36
17
High
Table
Io Phase
3.9:
Edge
The slopes and
X
X
X
X
errors of Io-A source characteristics demonstrating a very likely
dependence.
De Dependence
degree binned data
Slope
Standard
Io-A
Deviation
Probability
18
MHz
.02
.04
CML
of Peak
4.5
1.7
High
CML
5.7
2.6
High
Io Phase
1.2
1.4
Edge
Edge
De
75
De dependence on
Io-A' source characteristics
De Dependence
d.f.
rs
Very Likely
Possible
Inconclusive
>95%
80-95%
<80%
Io-A
Probability
Probability
Probability
CML
High
Low
@
@
@
20
MHz
MHz
22
MHz
18
of Peak
CML Edge
CML Edge
.79
31
.75
25
.68
31
.58
21
.81
21
.14
21
Table 3.11: The slopes and errors of Io-A source characteristics demonstrating a very likely
dependence.
De Dependence
degree binned data
Slope
Standard
Io-A'
Deviation
MHz
.03
.01
CML
of Peak
3.4
3.7
High
CML
6.6
1.8
Probability
18
Edge
De
76
3.5.3 Io-C
and Non-Io-C
The Io-C source seems

probability with Dg.
It
to exhibit little
change
shape and occurrence
in position,
does seem to show a strong frequency dependence, however.
Figure 3.35 illustrates the large differences in the Io-C source with frequency and the
mild dependence with De- Figure 3.36 also

is
Figure 3.37 on the other hand indicates a clear
weak.
The occurrence
but a
22
test
MHz
probability plots of Figure 3.38
De dependence
do not appear
to
there
at
is
20
CML
in
in Io phase.
show any
of the Spearman's rank correlation coefficient indicates that
Apparently
to
De dependence
illustrates that the
correlations
at least at
18 and
sufficient confidence in the correlation to reject the null hypothesis.
MHz
due
and insufficient data
to a poorer correlation
it is
not possible
make any conclusions on the correlation of the occurrence probability with DeThe non-Io-C source
non-Io-A source.
typically
had an occurrence probability about half
The occurrence probability had a
high according to the criteria set earlier
in this
was
correlation coefficient sufficiently
chapter to have a very likely dependence
on De- The slope of occurrence probability versus

reported in Table 3.13. Because non-Io-C
De was found
relatively
non-Io-C peak and edges are
The
shows
results of
in
for this source
and
is
weak, a study of the correlation
of the peaks and edges of the source was not performed.
shown
that of the
The mean
location of the
Figure 3.39.
Thieman and Smith [1978] from observations of Io-C
at
10 to 30
MHz
a drift in the peak of the source towards higher longitude with lower frequency.
This result
is
confirmed
These
Garcia, 1996].
Io-C source
18 to 22
is
in
recent data obtained by the
results
show
that
shifted by about 40-50 relative to
MHz.
Wind
spacecraft [Kaiser
between the frequencies of 2 and 14

its
MHz
and
the
location in the frequency range of
77
MHz
18
200
290
305
22 MHz
20 MHz
335
290
305
320
335
390
CML
3 frequencies (columns)
in
Table 3.2
305
320
335
350
320
The
in the Io
region containing the Io-C source
is
represented for each of
and 5 De ranges (rows). The De groups are the same as those
listed
78
High Edge
Io-C
360
340
-J
_Mid-Point
u
Peak
320
300
Low Edge
i
Dr
fit
260
CML
lines to the Io-C
'
1
'
'
'
Io-C
^High
250
Edge
0)
Peak
a.
Mid-Point
-^Low
Edge
230
fit
lines to the
Io-C Io phase peak, midpoint and edges.
79
Io-C
20 and 22
MHz
for Io-C.
The
vertical lines
mark
De and
at
18,
De dependence on
Io-C source characteristics
De Dependence
d.f.
rs
Very Likely
Possible
Inconclusive
>95%
80-95%
<80%
Io-C
MHz
MHz
MHz
.47
31
25
.38
31
of Peak
-.08
25
Phase of Peak
-.87
21
Probability
Probability
CML
Io
@
@
@
-.01
Probability
High
CML
18
20
22
X
X
X
.12
25
-.25
25
Edge
-.88
21
Phase Edge
-.62
21
X
X
.82
31
Edge
Low CML Edge

High
Io Phase
Low
Io
non-Io-C
Probability
18
MHz
Table 3.13: The slopes and errors of those characteristics of the Io-C source demonstrating a
very likely
De dependence
De Dependence
degree binned data
Slope
Standard
0.024
0.021
-0.94
0.63
Io-C
Deviation
Probability
Io
non-Io-C
18
MHz
Phase of Peak
Phase Edge
High
Io
Low
Io Phase
Probability
Edge
18
MHz
-0.84
0.42
-1.4
0.9
.0079
.0065
81
3.5.4
Summary
The De dependence of
the shapes of the sources described in this chapter
and Figure
rized in Figure 3.39
The
3.40.
lines are for the
low
CML,
high
is
CML
summa-
and peak
locations of each source as defined previously. Solid lines represent those features of the
source which are very likely to have a

features
which could not be proven
the best
fit
lines
and are drawn
to
at the
De
dependence. The dashed lines are for those
have any
mean
De
dependence. The dashed lines replace
value of the edge or peak of the source.
Io-C
Io-A
Io-B
Io-A'
CM
I
3
non-Io-A
non-Io-B
non-Io-C
CM
?J
I
_l
_l
L_
-J
250
200
150
100
300
350
CML
Figure 3.39:
The
best
fit
discussed in this chapter.
be proven to have a
lines to the low, high
The
vertical
De dependence.
dashed
CML
locations for each of the sources
lines represent
peaks or edges which could not
and peak
82
Figure 3.40:
The
best
fit
lines to the low, high
sources discussed in this chapter.
could not be proven to have a
and peak Io phase locations for most of the
The horizontal dashed
De dependence.
lines represent peaks or
edges which
83
Models of
3.6
3.6.1
Model
I:
Based on
the
Non-Io-A Occurrence
Based on Assumed Variation of
OP
accumulated evidence presented
the
Probability versus
with
Time
De
in this
chapter
we began
to
model
the long-term variations of the non-Io-A source under the assumption of a direct, linear
relationship with De- This relationship can be expressed in the formula
OP = A PDe
where
is
the constant of proportionality
of the data to the model.
Pde
is
(3-1)
which minimizes
the
sum
of squared residuals
given by
(3-2)
expressing the
of
1.
De dependence
Figure 3.41
source.
is
a plot of the occurrence probability at 18
into three sections based
The data have been divided
De- Pde for sections

based on Equation
of the peaks
2 and 3
1,
3.1
probability peak which
asymmetry of
sufficiently to
is
shown
in
and 2 but
fails to
In sections
is
MHz
for the
value
non-Io-A
on the dates of minimum
reach the
and 2 there
is
at
fit
matching the locations
minimum occurrence
probabilities
a clear "notch" at the occurrence
not accounted for in the model.
The model does not match
the occurrence probabilities peaks around the "notch"
match section 3 well.
maximum
Figure 3.42. Figure 3.43 shows the model
and Figure 3.42. The model does well
in sections
around 1970 and 1982.
the
of the occurrence probability and has a
and
fails to
drop
Figure 3.41: The non-Io-A occurrence probability values from 1957 through 1994 at 18
The
three sections are divided
by dates of minimum De
value.
MHz.
85
0.8
0 6
0.4
0 2
0
D.
Figure 3.42: Pde for sections
dependence of the occurrence
1,
2 and
3.
probability.
Pde
is
which expresses the De
86
3.6.2
It
Model
OP
2a:
De and Sunspot Number
Varies with both
can be seen from Figure 3.44 that throughout nearly the entire data set the seasonal
averaged sunspot number was
very close to anti-phase with De- Only
in
in the last
few
years has the sunspot cycle drifted so that the two curves are almost 90 out of phase.
It
was
before
in this
most recent
De had begun
visible in
some
this steep
drop
at the
De
to drop.
The
This drop
is
apparent
at all three
frequencies studied and
We
of the other sources but none as clear as non-Io-A.
is
not coincidental.
to include an inverse relation
relation
is
are proposing that
occurrence probability and the rise in average sunspot number occurring
in
about the same time
be modified
cycle that the occurrence probability dropped sharply, well
The model developed
as
Model
will
now
of occurrence probability with sunspot number.
De and
between occurrence probability,
sunspot number (SSN) can be
expressed as
OP = A Pd e
where Pssn
is
shown
in
is
the
same
Model
as in
occurrence probability was
first
sunspot number to adjust our

set.
This model
The
1.
Figure 3.45. Figure 3.46
is
proportionality, A, equal to 0.23 for
data
25Q
relation
between Pssn
all
three sections.
The
fit
ar>d
sunspot number
is
setting the constant of
influence of solar activity on
recognized in the data from the late 1980s but using the
fit
has improved the
now matches
sunspot number
is
included in the
fitting
over
all
three sections of the
the data in section 3 quite well and also
probability points around
(W-SSN)
an example of the curve
low occurrence
in
(3-3)
given by
.
PssN
iiiv
and Pde
Pssn
fit.
in the late
1970 better than the
The model appears

1970s.
to
first
matches the
model now
that the
overcompensate for the drop
Figure 3.44:
line)
and
DE
comparison of the time variations
in seasonal
averaged sunspot number (dashed
from 1957-1994.
0.2 -
50
100
150
200
250
SSN
Figure 3.45: Pssn for sections
1,
2 and
3.
Pssn
sunspot dependence of the occurrence probability.
is
which expresses the
88
O.J
Section
Section 3
Section 2
Modeled
using D t + SSN
1990
1980
1970
1960
Year
Model 2a of
Figure 3.46:
18
the occurrence probability for the

is
equal to 0.23 for
MHz
three sections.
all
The
non-Io-A source.
The
observed points.
3.6.3
Model
The
is
relationship
maintained
sections
in
2.
as
Model 2a Except
Pde
is
De >
2.7.
in the
is
is
fit
now
shown
asymmetry
Assumed Decrease
to section 3
for
De
used
in
in
OP
Model
where no obvious notch

by assuming a drop
High De
at
is
in the
and 2a
For
seen.
occurrence
Figure 3.47 shows the probability factor Pde for sections
given by
accounting for
in
fit
now accounted
Pd e =
The
for
between occurrence probability and
Model 2b
and 2 the notch
probability for
and
Same
2b:
De
+ 4), D
7^(^
6./
<2.7
and sunspot number as well as a notch
Figure 3.48. Including the sunspot number
in the
(3-5)
occurrence probability peaks although
observed points. The occurrence probability tends
to
in the
it
is
fit
in sections
and
allows for a degree of
not a perfect match to the
be slightly higher before the "notch"
89
than after. If the constant of proportionality
be
made
to the
fit,
an example
is
shown
in
is
allowed
to vary
Figure 3.49.
i
I
-2
Figure 3.47: Pde for sections
"notch" seen during the
and
maximum
2.
The drop
some improvement can
in probability at
occurrence probabilities.
DE
>
2.7
is
used
to
model
the
90
0 3
Section
Section 3
Section 2
Modeled
using D e + SSiV
0.2
0.1
1990
1980
1970
1960
Year
Figure 3.48: Model 2b of the non-Io-A occurrence probability at 18

proportionality
is
equal to 0.23 for
all
three sections.
The
Figure 3.49: Model 2b of the non-Io-A occurrence probability

constant of proportionality, A, to vary between sections.
shown.
The
observed points.
MHz.
at
18
The constant of
observed points.
MHz now
The value of A
allowing the
for each section
is
91
3.6.4
Model
OP
3:
is
Proportional to Source Width
Occurrence probability
emission beam
is
is
proportional to the length of time in which the
pointed towards Earth.
for a longer time as
it
DAM
wide rotating emission beam will be detected
sweeps past Earth than would a narrow beam.
It is
to
be expected
therefore that occurrence probability should be correlated with source width.
Figure
3.19 illustrates clearly the drastic change in source width that occurs with De- Given a
relationship between occurrence probability and source width similar to that expressed
in
Equation
3.1
namely,
OP = A PWidth
we
(3-6)
use Pwidth to describe the dependence that occurrence probability has on source width.
Pwidth
is
of data.
simply the source width divided by the

Figure 3.50
maximum
width. Figure 3.51 uses both source width and sunspot
probability.
There
is
good
number
as that for
De
fit
model
as well as
fits
Assuming occurrence
first
approximation. In
be higher
in the
reason that the
fit
probability
this
would simply be equal
model we assume
center of the
might be poor
during seasons with
model
the occurrence
little
all
model does not appear
is
that the notch appears
three sections.
The
fact
does provides insight into the non-Io-A emission beam.
it
time Earth enters the source until the time

to
to
for this
but an interesting point in this model
naturally to occur at the high occurrence probability peak in
that the
in that section
a gap in the data in the early 1970s and early 1980s where a clear
determination of source width could not be made. The

to be as
source width
a model of the occurrence probability based solely on source
is
is
data or
beam
due
when
it
the
beam
leaves.
to the
will
source width
is
sound
be uniformly active from the
The occurrence
probability
cross section than towards the edges.
to the inability to
is
likely
Another
determine accurate source widths
the sources could not clearly be resolved. Despite
92
these assumptions and errors
the occurrence probability
is
it
appears from this model that most of the
De
effect
on
geometrical and due to changes in the width of the beam.
0.3
Section
r-
Section 3
Section 2
Modeled
using width only
0 2
D
a
o.i
L
1980
1970
1960
1990
YEAR
Figure 3.50:
Model
3 of the non-Io-A occurrence probability based on the source width.
is
0.23.
The
triangles indicate the observed points.
The
93
0.3
Section
-h
Section 3
Section 2
Modeled
using width
SSN
0 2
8 01
l_
1990
1980
1970
1960
Year
Figure 3.51:
model of the non-Io-A 18
MHz
occurrence probability based on source width
and sunspot number. The constant of proportionality
The
in all three sections is 0.23.
triangles
indicate the observed points.
The notch seen

the figure.
in sections
and 2 of Figure 3.41 was not
Our model 2a could not
but simply hidden.

dissertation of St.
The
first
fit
section 3
if
a notch
mention of a "notch"
Cyr [1985]. Evidence of
this
in the
notch
in
visible in section 3 of
was assumed
UFRO
to
data set
other data sets
is
be present
was
seen
in
in the
Figure
7.13 of Physics of the Jovian Magnetosphere [Carr et ai, 1983] where data from the
7-40
MHz
and from
NASA
University of Colorado
at
plotted versus time. St.
Cyr found from plotting the
MHz
is
amount of right-hand
to
observing stations at 16.7
ratio of the
left-hand non-Io-A emission that during the period of highest
DE
this ratio
other words, the amount of left-hand activity had fallen relative to the
hand
activity.
If
the left-handed bursts
source were beamed
and right-handed
in slightly different
peaked.
amount of
bursts detected in the
In
right-
non-Io-A
directions then the drop in left-hand polarized
emission could be due to Earth moving out of that component of the
beam which
is
94
left-handed. At the
lower longitude
most extreme positive range of
in the
non-Io-A peak shown
narrowing of the non-Io-A source
when Earth
that
is at its
at the
in
3.6.5
is
Is
detected
highest Jovicentric declination that
when Earth
The Sun can influence

local observing conditions,
DAM
ram pressure and

(less than
about a
the
is,
De
it
begins to intercept radiation
beam maximum
for the
DAM?
changing the density of plasma

at the
DAM
source.
in the interplanetary
activity
with the magnetic sector structures in the solar wind, solar wind
solar
wind
density.
All of these correlations were found in short-
1978, Terasawa et ai,
1978,
Barrow
et ai,
1986].
in solar activity
and
Our model assumes
DAM
activity.
little
the occurrence probability value
12 years (one
De
present
in
would explain
no lag time
The dashed
cycle) earlier.
occurrence probability after 1987 coincided with a sharp

if
to
exists
Figure 3.52 illustrates the change
occurrence probability with year for the non-Io-A source.
notch,
study of the
(DIM) emission
found a positive correlation with solar wind ram pressure with a lag
time of 2 years [Bolton et ai, 1989].
between changes
peak expected
medium
Researchers have reported positive
one month) variations of the occurrence probability of the non-Io-related
sources [Barrow,
in
non-Io-A
of 2.7.
correlation between long-term (years) variations in the Jovian decimetric
and solar
by assuming
the detection of Jovian decametric radiation by changing the
and by changing conditions
term
is at
There a Solar Effect on
correlations of
there appears to be a turn towards
highest De- This can be explained
from just north of the beam maximum. That

source
De
Figure 3.17. This could be evidence for a
rise in the
The
line
shows
large drop in
sunspot number. The
the decrease in occurrence probability in
1990 never occurred rather the probability continued
1989 but the
to drop.
95
MHz non-Io-A
10
i
'
Sunspot No. -
0.4
0.3
0.2
/
'
'
V
l
//
m=A
1980
1975
&
*'
\\
^JL
'
m.
0.1
'
'
1970
I.
1995
1990
1985
Year
Figure 3.52:
superposition of the occurrence probability data from the 1970s on the occurrence
probability data of the 1980s (12 years later).
The dashed
probability during the declining phase of the 1970s
and early 1990s. This drop
in
line illustrates that the occurrence
was higher than
may be
that seen in the late 1980s
related to the sharp rise in solar
activity at the time.
If
there
is
geometry or the emission source or both.
amount of
DAM
a solar influence on long-term variations in
activity
may
or
may
change
in the
not have changed but the
it
may
effect the
beam geometry means
beam has changed
its
beam
that the
orientation
with respect to the ground-based observer and therefore the observer sees a change
the
amount of emission.
A change
at the
emission source would
mean
the source has changed. There are several characteristics of the
which vary with the solar
cycle.
Of
these
some
are
more
are in phase with sunspot
number.
that the activity of
Sun and
the solar
likely to influence the
geometry while others may influence the emission source; not
all
in
wind
beam
of these characteristics
96
The dynamic
minimum
solar
or
ram pressure of
at solar activity
maximum
the solar wind, for example, has been found to have a
maximum. The peak
[Schwenn, 1990, Richardson
due
pressure occurs about 2 to 3 years after
1995].
The
along which the emission cone
is
aligned are moving
wind pressure. This movement
to the reduction in solar
hemisphere may allow ground based observers

the north than usual, well into the notch
wind ram pressure
solar
One
falling after 1987.
and the occurrence probability are both

field lines
in
et ai,
possibility
in the
that the
for a cone in the northern
to see a section of the
mentioned
is
sunward direction
beam
farther to
During years of highest
earlier.
De
in the 1964 and 1976 seasons the occurrence probabilities dropped, but with the change
in orientation
of the cone due to the decreased
The De
larger.
degrees higher than during the previous

higher than normal
in
beam
at
in
beam
in the
the drop in 1988
is
even
about 1991 the
problem with
this
1960s and 1970s. These
oriented such that ground-based
a higher than normal latitude
occurrence probability.
pressure peaked
De maximums
values and the emission
De
observers intercept the
drop
ram pressure
values reached in the years from 1987 to 1989 were actually about 0.5
model
may combine
is
that
when
to explain the
the solar
wind
beam should have been pushed back and ground-based
observers should have intercepted the peak again. This did not occur. The peak should
have been seen

but
it
in 1990,
by
this
model the peak may have appeared
The paper by Richardson
etal.
ram pressure follows
the sunspot
although during solar
maximum
higher standard deviation
Terasawa
fast
later
than expected
should have appeared.
et al.
in the
[1978] and
[1995] shows the standard deviation of the solar wind
number more
there
pressure
is
is
more
closely than the average
scatter in the data.
an increase
Kennedy [1976] speculate
in
shock fronts
that
ram pressure
One cause
in the solar
of a
wind.
shocks from the collision of
and slow solar wind streams may enhance the non-Io-related emission source. The
97
Jovian magnetosphere however
many shocks
in
succession
may have
may have
a characteristic time constant in
less of an effect
DAM
on the
than
which too
more widely
space shocks [Kennedy, personal communication]. The high rate of shocks encountered
during the solar
maximum may
therefore reduce the
amount of activity seen
Another possible means by which the Sun may influence the

by way of
10.7
UV
cm
flux
photons.
in this period.
emission source
and the core-to-wing
ratio
of the
Mg
UV
h and
II
flux
would
lines,
an index of solar
likely increase the electron
density of the Jovian ionosphere, similar to what occurs on Earth.
proposed
to explain the
within a narrow range.

the emission source.
is
good correlation has been found between sunspot number,
[Heath and Schlesinger, 1986]. The
flux
number
UV
DAM
DAM
Theories
emission mechanism require electron number densities
The increased number
densities
may have
the effect of saturating
CHAPTER
SUMMARY AND CONCLUSIONS
This study of over 30 years of decametric observations has been a process of
wheat from the
winnowing
the
changes
galactic
in
background
The chaff
chaff.
changes
level,
includes such bothersome details as
in
the distance
occasional modifications of the radio telescope system

these effects and others had yielded, by the
end of Chapter
2,
from the source and
Efforts at correcting for
itself.
a set of galactic-background-
limited occurrence probabilities and average flux densities. Chapter 3 presented what had
been gained from the

in
some
effort, a clear indication
aspects of the decametric emission,
parts and a
few parts
that
The non-Io-A source
of periodic behavior and general trends
somewhat
less certain periodicities in other
remained nearly constant over the entire span of observations.
is
confirmed to be strongly influenced by
probability and the average flux density of the source both
effect
though the occurrence probability was found
to be a
D E The
.
occurrence
showed evidence of a
more
reliable indicator
DE
of
this effect. Studies showing a D E effect in source width are also confirmed. The paper
by Carr
[1970]
et al.
edge; the low
CML
showed a widening non-Io-A source mostly due
edge was virtually independent of
beam has been confirmed
of non-Io-A
at
18,
to the high
CML
This figure for the non-Io-A
with the addition of data from over two additional orbits of
Jupiter about the Sun. Hypothesis tests using

the null hypothesis of no
DE
DE
20 and 22
Spearman's rank correlation coefficient of
dependence were performed. The occurrence probability
MHz
to reject the null hypothesis at the
all
showed
95%
sufficient evidence of a
confidence interval. The
98
CML
DE
dependence
peak, midpoint
99
and edges of the non-Io-A source were tested and except for the low
found
all
to
From modeling
results,
the
change
in source
in
width was found to be one of the
came from
was especially effective
mid
in the
to late
1980's.
The
in
beam
effect
cross section
including a factor based on the
yearly averaged sunspot number. This factor served to diminish the

probability and
The De
occurrence probability.
therefore primarily a geometrical effect caused by changes in the
seen from Earth. Improvements in our model
noted
edge were
show a De dependence.
primary reasons for the long-term variations

is
CML
modeling the drop
in
modeled occurrence
solar activity cycle appears to play a secondary
role in influencing the long-term variations of the Jovian decametric activity, at least
for non-Io-A.
"notch" seen
maximum D E
and
De
DE
At
seen.
is
in
over 39 years.
DE
The
plots of occurrence probability versus time at the points of
confirmed. Figure 4.
is
a polar plot of non-Io-A occurrence probability
The general trend in occurrence probability
to increase with
values greater than about 2.7 however a drop in occurrence probability
latitude range available to
ground-based observers
emission beam appears
to take
UFRO
made about 6 sweeps
data. Earth has
in the polar plot
changes
is
in the
it
just north of the
could be due to changes
in latitude
is
sampling the non-Io-A
beam maximum. Over
the span of the
through the beam.
in the orientation
emission source over this period.
in
of the
beam
The
scatter
as well as to
100
non-lo-A De dependence
Figure 4.1:
latitudinal slice through the
from 1957-1994.
beam. The dashed
in
which Earth
is
search for
An
envelope
in the
De
notch as described
A
De
listing
DAM
of the
at
18
MHz
the points illustrating a possible shape to the
upper edge of the beam that
is
sampled during
that time
in the text.
effects in the other decametric sources has turned
generally not as pronounced influences.
of the major
non-Io-A emission beam based on data
drawn around
is
line represents the
up similar though
The De dependence of the occurrence
probability
sources was studied using Spearman's rank correlation coefficient.
DAM
sources in terms of this coefficient
is
(in
order of decreasing
dependence), non-Io-A, non-Io-C, Io-A', Io-B, Io-C and Io-A. The Io-A and Io-A'
sources, like the non-Io-A source,
showed
edge of the source but strong dependence

a dependence on
show
strong
De
in
little
in the
both the low and high
De dependence
to
no dependence on
high
CML
Io-phase as well.
in
CML edge.
edges.
CML
and
Io
The
low
CML
The Io-B source shows
The Io-B and Io-C sources
set
of source boundaries were
phase locations of the major sources.
source was also found to display recurrent features

several previous papers.
in the
Using nearly 40 years of data from
the University of Florida Radio Observatory, a uniform
determined, refining the
De
in
CML-Io phase
significance of these features
is still
plots,
The Io-B
confirming
not entirely clear.
101
As
reported by various authors and confirmed in Chapter 3 of this study,
the phase of Io at
the
DAM
which
DAM
is
first
detected and at which
it
De
influences
ceases for several of
sources [Lecacheux, 1974, Thieman et ai, 1975, Bozyan and Douglas, 1976,
Boudjada and Leblanc, 1992]. Desch [1978] described the influence De has on
phase of the
at the
DAM
source and the direction to the observer (Earth).
opening angle the emission
is
beamed towards
Earth.
the observer changes, and for a particular cone,
angle. The cone from which the radiation
one
at
the Io
sources by defining 3, the angle between the magnetic field vector
is
When 8
When De
is
equal to the cone
changes, the direction of
8 may no longer equal
most
the
likely to be detected as
a longitude farther from the Earth-facing longitude.
The
cone opening
De
increases
is
toward smaller Io
drift is
phase for the Io-B and Io-C sources. Desch found that the degree and the sense towards
which the Io-B and Io-C sources
in Io
phase.
Figure 4.2 illustrates with a simple geometrical drawing
phase of the Io-B source

by the vertical bar
Cones
circle.
at
De
is
consistent
Chapter 3 reports no conclusive evidence for an Io-A dependence on
with his model.
De
phase over the range of
drift in Io
may be
influenced by De-
in the figure, the particular
Over
the range of
how
the Io
De, represented
cone which intercepts Earth
will change.
2 and 3 are represented by circles with their labels located at the centers of each
Cone
is
located at a higher Io phase, than cone 2 and 3 and intercepts Earth
a lower De, represented by the
towards the right
in the figure.
Earth at a slightly higher De.
number
Cone 2
is
next to the vertical bar. Io phase increases
at
a slightly lower Io phase and intercepts
102
Phase
Io
Figure 4.2:
source.
An
Cones
illustration,
not to scale, of the effect that
at progressively
The
ities
is
in the late
at
The De range
higher De-
is
1980s could be due to a variety of different characteristics of the Sun

to the solar cycle.
by changing the beaming geometry or the

ations in the solar
UV
flux, the solar
amount of
DAM.
This
list is
as possible
DAM
were presented
any of these suggestions are the correct ones. This solar

by the unexpected early
10 years. Previous to
rise in
this,
alter the
DAM
the standard deviation of
means by which
not exhaustive and although
each of these factors may influence the
in the last
The Sun could
activity at the source or both. Vari-
wind ram pressure and
wind ram pressure were discussed
influence the
attention
has on the Io-phase of the Io-B
increasing towards the right.
solar effect believed responsible for the premature drop in occurrence probabil-
which vary with period roughly equal
the solar
De
lower Io phase intercept Earth
represented by the vertical bar and Io phase
the
Sun could
some suggestions on how

it
is still
effect has
not clear which
if
been brought to our
sunspot number with respect to
the secondary influence of the
Dg
that occurred
Sun could not be
103
separated from the primary
confirm whether
to
Another
emission
6 Jovian
De
study of long-term variations of the decametric
this
in
the role of volcanic activity on Io.
is
radii
is
filled
is rich in the material
body
by which Io
atoms
injects these
volcanic activity influences the
between volcanic
at the
activity
found
these data sets and the
Io.
UFRO
The Voyager spacecraft on
[1981] and
make up
is still
uncertain.
Some
The degree
This
Another program running coincident with
Lick Observatory.
at
MHz
data sets might be
their flyby mission to Jupiter
The Voyager 2
1
is
a program
this
one
is
were able
to
measure
DAM
spacecraft approached Jupiter at a
approached Jupiter
at
about 3.5. Barrow
St.
is
no report of any tapering of
Cyr however notes
that the
the
drop
Dg
effect to these
non-Io-A beam as reported
in left
hand emission
at
18.4
detected by both Voyager spacecraft and reported in a paper by Alexander et
1976. Currently, the
MHz
an
correlation study between
fruitful.
[1981] might be confirmation of the effect seen in the
2-14
which
and Carr [1984] note that the longitude of the non-Io-A source from
There
in this dissertation.
to
insight into the relationship
Voyager spacecraft data are consistent with an extrapolation of the
higher latitudes.
in
the torus but the exact process
be gained from a study of observations made
approximately 7 while Voyager
Maeda
This volcanically
this material is Io itself.
unknown.
is
may
higher Jovian latitudes than Earth.
latitude of
the
Io's orbit at about
Infrared Telescope Facility in support of the Galileo mission.
Io Plasma Torus monitoring program
at
to
into the torus
DAM
DAM
and
monitor IR hot spots on
to
The region around
with a complex mixture of neutral and ionic species of Sodium,
Potassium, Sulfur and Oxygen. The source of

active
data during this unusual period will help
this solar effect really exists.
parameter
free
More
influence.
Wind
spacecraft
[Kaiser and Garcia, 1996].
is
If
detecting
it
UFRO
DAM
18
MHz
al.
polarimeter data
over the frequency range of
survives another 10 years
it
will
be able
to
104
study the changes in
DAM
over the entire range of Dg. This long-term monitoring over
wide range of frequencies may help resolve what variations are due
the emission beam and what are due
to the
to variations intrinsic to the source.
geometry of
APPENDIX A
A SAMPLE CALIBRATION OF THE YAGI ANTENNA SYSTEM
Calibration of Jupiter radio bursts
is
done by matching the antenna temperature, T'A
at the calibration point to the
temperature of a noise standard, T'v
from both of these sources
assumed
The
is
line loss in decibels
given as L/ and the
preamplifier in the line
At
is
from
to
have been attenuated
The
slightly
signal
by
coming
line losses.
the active element of the antenna to the calibration point
line loss
from the noise standard
is
Lc
There
may
coming from the antenna which may add a gain of
the calibration point the observer will adjust an attenuator in the line
also be a
decibels.
coming from
the noise standard until
T'A
The antenna temperature
at the calibration
= T'N
point
(A-l)
is
given
in
terms of
TA
the temperature
at the active element of the antenna by
T'A
= TA W^r-
and the noise temperature of the standard
T'N
where
is
(A-2)
at the calibration point is
XyyTO
given in terms of 7/v by
(A-3)
io
the attenuation in decibels inserted in the line by the observer. Equating the
two expressions and solving
for
TA we
get
TA = TN 1
(L
-L c -a-D)
-o
.
105
A-4)
106
As an example,
the galactic
Jupiter radio storm of June
the calibration
done
1,
background levels for the 18 Yagi before and
1987 had a deflection on the chart recorder of
L/
from the antenna
was
6.8 decibels,
is
35.1 decibels, the value of D.
observatory so there was very
little
to
be zero because the calibration was done
line
in the
cable between the noise standard and the calibration
is
temperature of the galactic background from equation A.4
is
point.
The
given in tables A. 2 and A. 3. In 1987,
to the calibration point are
Lc was assumed
From
corresponding decibel attenuation of the noise
after the storm, the
standard necessary to produce a deflection of 15
losses
after the
15.
There was no preamplifier on 18 Yagi
in
1987 so
zero as well.
The antenna
(6.8-0-0-35.1)
TAaB = TN
.
The noise source used
temperature of 62.7 x 10
in the calibrations is
Kelvin.
(A-5)
ro
an
HP461A
which has a noise
amplifier
the galactic
background
is
93,147 Kelvin. During one five minute interval of the storm a Jupiter burst was recorded
with a deflection of 29. The noise standard matches
of 31.7 decibels. The
this deflection
antenna temperature for this burst
TAj+aB
is
with an attenuation
202,596 Kelvin. This
temperature contains a contribution due to Jupiter as well as the galactic background.
The
interpolated galactic
background
is
at
background temperature
for this storm
is
equal to the galactic
the beginning of the storm. The difference between the two temperatures
equal to the galactic background subtracted antenna temperature, Taj_ gb
The
flux density of the Jupiter burst
S =
where k
The
Ae
is
Boltzmann's constant and
Ae
is
2kTA aB
/-
is
(A-6)
the effective area of the
values for the Yagi antennas used in this study
density for the Jupiter burst
is
8.6
given by
x 10-21 Watts
m2
is
18 Yagi antenna.
given in Table A.I. The flux
Hz. Usually the flux density
is
107
expressed
is
in
then 8.6
Table A.
terms of Janskys where
10
Antenna
Jansky = 10~ 26 Watts
m2
Hz. The flux density
Janskys.
characteristics for the
MHz
Number
Frequency (MHz)
Yagi
20
MHz
and 22
MHz
of Elements
Yagi antennas
Effective Area
18.0
351.0
20
20.0
218.0
22
22.2
229.0
18
(m 2 )
Table A.2: Line losses and gains and antenna failures 1965-1985. Losses and gains are
in units
of decibels (dB).
Year
Antenna
Li
Le
Comments
1967-90
18Y
6.8
Losses and gains for 18Y and
typical
20Y
5.5
20Y based on measurements
22Y
3.5
taken on
1965-67
18Y
4.8 est.
Bivens
1970
18Y
9.8
Dec. 7-18, -3
1982
22Y
1983
18Y
Fixed
Fixed
Dec
18Y
est.
est.
@
@
dB
hybrid ring
T-3 Jan. 5-20
T+5 May
6-16, T-3
1-3,
May
18
Apr. 2-Jun. 5 -3dB hybrid

ring, Jul. 5-Sep.
preamp
&
-3
12
+10 dB
dB hybrid
ring,
@ T-3 Apr. 24-30

Fixed @ T-l Aug. 21-24,
@ Trans. Aug. 25-26,
@ T-0.6 Aug. 27-29,
@ T+l. 5 Sep. 4-25
fixed
22 Y
1988.
Arm
T+l May
1985
6,
108
Table A. 3: Line losses and gains and antenna failures 1986-1994
Year
Antenna
1986
22Y
Li
Lc
Comments
Fixed
@
@
1987
Fixed
20Y
Fixed
22Y
Fixed
@
1990
T+2.5
18Y
@
1988
@
@
until
Jul 30-31
15-19?
end of season?
May
T-4
26-Jun. 3
Trans. Aug. 12
T+l Aug. 25-27
T+4
T+7.5 Sep. 17-19
Sep. 20
T-2 Dec.
18Y
5.5
Fixed
4.5
Lines replaced for 18Y
Fixed
Nov.
@
18Y
13.0
22Y
18Y
Fixed
3.0
13.0
30,
@
@
@
1-2.
& 20Y
T-3 Oct. 10?-
T+l Dec. 3-11
12Feb.
1?
T-2 Nov. 22
T-5 Jan. 1-Feb. 1?
T+5 Mar.
Fixed
T-5 Dec.
Fixed
@
1992
T+2.5
20Y
22Y
1991
Trans. Oct.
15,
T+5
= +13 dB
T-4 Nov.22
Feb. 9,
until
Apr. 30?
Nov. 5 calibration done

upstream of preamp.
in field
APPENDIX B
TESTS FOR D E DEPENDENCE
An
assumption that needs to be
made
between n pairs of points from two
random samples from a normally

coefficient such as Pearson's r
in
order to interpret the value of a correlation
sets of data,
and
Y,
is
that the data are both
distributed population. In these cases a linear correlation
[Mendenhall and Scheaffer, 1973]
is
commonly used and
given by
is
U-X){Yi-Y)
r
(B-l)
Pearson's r coefficient takes on values from +1 for a positive correlation where
on a
lie
straight line with positive slope
of negative slope.
An
r value of
and -1 where
0 implies no
of data. For determining correlations between
all
correlation exists
De and
all
points
the points lie along a line
between the two
sets
source longitudes or occurrence
probability a problem arises in that the data are not normally distributed so interpretation
of Pearson's r coefficient becomes unclear.
Another measure of correlation which does not depend on the distribution of
is
Spearman's rank correlation coefficient. In
each Xi
is
replaced by
its
also replaced by their ranks.
is
the rank of the
z'th
this
method
corresponding rank among the

If
/?,
for
N pairs
values.
represents the rank of the
zth
and
of measurements
The Y values
element of
and
are
S,
element of Y then the Spearman's rank correlation coefficient, rs
109
110
[Press et
al.,
1992]
is
GZ(Ri-Si)
Spearman's rs takes on the same range of values as Pearson's r correlation
sample computation of Spearman's

non-Io-A
DE
rank of De,
Oc.
at
20
Prob.
rs
is
coefficient.
given below.
MHz
N
rank of Oc.
Prob.,
(R
= 27
(R
S)
5)
3.06
24
0.0248
12
12
!44
2.38
22
0.0391
18
16
1.01
17
0.0702
20
-3
-0.84
12
0.0187
11
-2.25
0.0149
-1
-3.09
0.0029
-3.07
0.0022
-2.18
0.0068
-0.47
13
0.0473
19
-6
36
36
18
0.0951
24
-6
2.84
23
0.1308
26
-3
3.09
25
0.0799
22
2.22
20
0.0969
25
-5
25
0.66
16
0.0339
16
-1.10
11
0.0299
13
-2
-2.47
1.44
0.0166
10
-3
-3.19
0.0052
-2
-3.05
0.0062
-1
-0.32
14
0.0755
21
-7
49
1.94
19
0.0823
23
-4
16
3.40
26
0.1574
27
-1
3.60
27
0.0334
15
12
144
2.34
21
0.0327
14
49
0.45
15
0.0350
17
-2
-1.57
10
0.0048
49
-3.00
0.0075
-2
-3.70
0.0054
>
-4
H(R
4
16
-
S)
650
(B-3)
The Pearson's
r coefficient for the
same
set of data is 0.812. In this instance relaxing the
restriction of normality in the use of Pearson's r
The goal of
this
study
is
to
does not introduce a significant
error.
determine whether the two populations from which the
measurements were taken are also correlated.
we
In other words, can
reject the null
hypothesis of an rs value of 0 for the population in favor of the alternative hypothesis

that \r s
>
0 for the population given the correlation measured for the sample. In order
we must know how
to test the significance of the rs value
rankings of
against an arbitrary ranking of
Y.
rs
is
distributed for
This will be a two-tailed
Student's
[Kendall
distribution can be used with the
and Gibbons,
1990].
The
For
exists in the population.
assume a positive or negative correlation
approximation can be used for the frequency distribution of rs
N>
a Student's
possible
we do
not
35 a normal
For 19 <
number of degrees of freedom,
test statistic for
all
test if
distribution
< 30
the
df=N-2
is
(B-4)
For the value of rs computed above and 25 degrees of freedom the

6.704.
table of critical values of the Student's
probability
is less
drawn from
1%
in
Chapter 3
to classify the
sources was based on these significance

greater than
95%
candidate for a
for correlation
shows
De
confidence of a
tests.
Any
test
dependence. Data
was deemed a possible candidate
for
in
in
sets of
De
dependence.
DE
The
dependence of each of the
which showed a confidence
No
is
measurements were
level
likely
which the confidence was between 80 and
which there was
t,
that for a two-tailed test the
two
between the two populations was deemed a very
could be reached for those data

correlation.
that the population that these
uncorrelated given the rank correlation coefficient determined above.
is
headings used
than
test statistic,
95%
conclusion on
De dependence
80%
probability for
less than an
REFERENCES
Acuna, M.
H., K.
magnetosphere,
Chapter
W. Behannon, and
in
Acuna, M.
and N.
H.,
81, 2917-2922,
E. P. Connerney, Jupiter's magnetic field and
Cambridge University
pp. 1-50,
L,
J.
Physics of the Jovian Magnetosphere, edited by A.
Ness,
F.
Press,
New
The main magnetic
26, 1193-1199,
field
of Jupiter,
101,
J.
Geophys. Res.,
activity, Planet.
from Jupiter observed
at
effects
the
in
decameter-wave
the earth and from Voyager, Astron. Astrophys.,
142-149, 1981.
Barrow, C. H., and T. D. Carr, Radio observations of the Jovian magnetic
Space
Space
1978.
Barrow, C. H., Latitudinal beaming and local time

radiation
Dessler,
1976.
Barrow, C. H., Jupiter's decametric radio emission and solar

Sci.,
J.
York, 1983.
Res., 12, (8)155-{8)164,
Barrow, C. H., M. D. Desch, and
field,
Adv.
1992.
F.
Genova, Solar wind control of
Jupiter's decametric
radio emission, Astron. Astrophys., 165, 244-250, 1986.
Bigg, E. K., Influence of the
satellite Io
on
Jupiter's decametric radiation, Nature, 203,
1008-1010, 1964.
Bishop, R.
L.,
Handbook 1996, Royal Astronomical
Observer's
Society of Canada,
Toronto, Canada, 1995.
Bolton,
S., S.
between
Gulkis,
M.
J.
Klein,
I.
de Pater, and T.
J.
Thompson, Correlation studies
wind parameters and the decimetric radio emission from
solar
Jupiter, J.
Geophys. Res., 94, 121-128, 1989.

Boudjada, M. Y., and Y. Leblanc, The variability of Jovian decametric radiation from
1978
to 1988, Adv.
Bozyan,
F. A.,
radiation,
Burke, B.
J.
and
Space
J.
Res., 12, (8)95-(8)98,
1992.
N. Douglas, Directivity and stimulation
in
Jovian decametric
Geophys. Res., 81, 3387-3392, 1976.
F.,
and K. L. Franklin, Observations of a variable radio source associated
with the planet Jupiter,
J.
Geophys. Res., 60, 213-217, 1955.
112
113
Carr, T. D.,
emissions,
M. D. Desch, and
in
pp. 226-284,
K. Alexander, Phenomenology of magnetospheric radio
Physics of the Jovian Magnetosphere, edited by A.
Cambridge University
Carr, T. D., and S. Gulkis,

8,
J.
Press,
New
The magnetosphere
Dessler, Chapter 7,
J.
York, 1983.
of Jupiter, Ame. Rev. of Astro. Astrophys.,
577-618, 1969.
Carr, T. D., A. G. Smith, H. Bollhagen,
J.
N.
Six,
F.
and N. E. Chatterton, Recent
decameter- wave-length observations of Jupiter, Saturn and Venus, Astrophys.
J.,
134,
105-125, 1961.
Carr, T. D., A. G. Smith, F. F. Donivan, and H.
I.
Register,
of the decametric radiation of Jupiter, Radio Science,
1957, Astrophys.
Connerney,
Emission HI, edited by H. O.

Austrian
Academy
Connerney,
Rucker,
Jupiter's magnetic field,

S.
J.
Bauer, and
M.
in
L.
Planetary Radio
Kaiser, pp.
13-33,
of Sciences Press, Vienna, 1992.
E. P.,
J.
periodicities
127, 274-283, 1958.
Doing more with
E. P.,
J.
J.,
The twelve-year
495-503, 1970.
and C. H. Barrow, 18-megacycle observations of
Carr, T. D., A. G. Smith, R. Peeple,

Jupiter in
5,
Magnetic
fields
of the outer planets,
J.
Geophys. Res., 98, 18659-
18679, 1993.
Connerney,
J.
E. P., R. Baron, T. Satoh, and T.
Owen, Images of excited
at the foot
of the Io flux tube in Jupiter's atmosphere, Science, 262, 1035-1038, 1993.
Desch,
M.
D.,
Groundbased and Spacecraft
hectometer wavelengths.
PhD
studies of Jupiter at
Decameter and
dissertation, Univ. of Florida, Gainesville, Florida, 1976.
Desch, M. D., Io-phase motion and Jovian decametre source locations, Nature. 272,
339-340, 1978.
Desch, M. D., Radio emission signature of Saturn immersions

J.
in Jupiter's
magnetic
tail,
Geophys. Res., 88, 6904-6910, 1983.
Desch, M. D., T. D. Carr, and

large array, Icarus, 25,
Douglas,
J.
N.,
and H.
sources, Nature, 199,
J.
J.
Levy, Observations of Jupiter
at
26.3
MHz
using a
12-17, 1975.
Smith, Change
in rotation
period of Jupiter's decameter radio
1080-1081, 1963.
Dulk, G. A., Apparent changes
in the rotation rate
of Jupiter, Icarus,
7,
173-182, 1967.
114
Flagg, R.
W.
S.,
B. Greenman, F. Reyes, and T. D. Carr,
Jovian Decametric Radio Noise Bursts Spectra, Volume
1,
Catalog of High Resolution

Dept. of Astronomy, Univ.
of Florida, Gainesville, 1991.
Franklin, K. L.,
64, 37-39,
An
account of the discovery of Jupiter as a radio source, Astron.
Franklin, K. L., and B. F. Burke, Radio observations of Jupiter, Astron.
The
Gallet, R. M.,
MC
in
J.,
1959.
J.,
61, 177, 1956.
emissions on 18 and 20
results of the observations of Jupiter's radio
1956 and 1957, IRE Transactions on Antennas and Propagation, AP5, 327-328,
1957.
Gallet,
R. M., Radio observations of Jupiter, in Planets
G. P.
Kuiper, and B.
and
edited by
Satellites,
Middlehurst, pp. 500-533, University of Chicago Press,
M.
Chicago, 1961.
Gardner, F.
F.,
and C. A. Shain, Further observations of radio emissions from the planet
Jupiter, Austr. J. Phys., 11,
Genova,
55-69, 1958.
and Y. Leblanc, Interplanetary
F.,
scintillation
DAM
and Jovian
emission,
Astron. Astrophys., 98, 133-139, 1981.
Gulkis,
and T. D. Carr, Radio
S.,
Heath, D.
F.,
rotation period of Jupiter, Science, 154, 257, 1966.
and B. M. Schlesinger, The
in solar ultraviolet irradiance, J.
Higgins. C. A.,
Mg
280-nm doublet
as a monitor of changes
Geophys. Res., 91, 8672-8682, 1986.
A new determination
PhD
of Jupiter's radio rotation period.
dissertation,
Univ. of Florida, Gainesville, Florida, 1996.
Imai, K., L.
Wang, and
T. D. Carr, Origin of Jupiter's decametric modulation lanes, in
Planetary Radio Emissions

pp. 69-90, Austrian
III,
Academy
edited by H. 0.
International Astronomical Union,
1AU Symposium
H.
J.
Smith,
Rucker, S.
J.
Bauer, and M. L. Kaiser,
Commision 40
(radio astronomy), in Proceedings
No. 12 URSI Symposium No. I, edited by B. F.
and
W.
J.
Burke, A. G.
Warwick, Information Bulletin No.
8,
London,
Smith,
1962.
International Astronomical Union.
Kaiser,
M.
L.,
and L. N. Garcia, Jupiter's low-frequency radio spectrum:
the gaps, in Planetary
M.
L.
Radio Emission IV, edited by H. O.
Kaiser, Austrian
Academy
Rucker,
S.
J.
Filling in
Bauer, and
115
Kendall, M., and
D. Gibbons, Rank Correlation Methods,
J.
5th edition,
1990.
Kennedy, D.
J.,
Edward Arnold, London,
Polarization of the decametric radiation from Jupiter.
PhD
dissertation,
Univ. of Florida, Gainesville, Florida, 1969.
Kennedy,
J.
Underlying Processes of the Jovian Decametric Radiation.
R.,
dissertation, Univ. of Florida, Gainesville, Florida,
Leacock, R.
L.,
PhD
1976.
and A. G. Smith, Fine structure of Jupiter's decametric source B,
Nature Physical Science, 134, 60-61, 1973.
Lecacheux, A., Periodic variations of the position of Jovian decameter sources

longitude (System
Maeda,
III)
and phase of
K., and T. D. Carr,
Beam
Io,
structure of Jupiter's decametric radiation, Nature,
1984.
308, 166,
Mendenhall, W., and R. L. Scheaffer, Mathematical
Duxbury
in
Astron. Astrophys., 37, 301-304, 1974.
Press. North Scituate, Massachusetts,
W. H., S. A.
FORTRAN, The Art
Press,
Teukolsky,
W.
Vetterling,
Statistics
with Applications,
1973.
and B. Flannery, Numerical Recipes
in
of Scientific Computing, Cambridge University Press, Cambridge,
1992.
Richardson,
J.
D.,
J.
W.
Belcher, A.
properties of the solar
Statistical
J.
Lazarus, K.
wind (submitted),
I.
Paularena, and P. R. Gazis,
in
International Solar
Wind 8
Conference, Dana Point, California, 1995.
Riihimaa,
J.
J.,
Modulation lanes
in the
dynamic spectra of Jovian L
bursts, Astron.
Astrophys., 4, 180-188, 1970.
Roberts,
J.
A., and
M. M. Komesaroff, Observations of
Jupiter's radio spectrum
and
polarization in the range from 6 to 100 cm, Icarus, 4, 127-156, 1965.
Schwenn,
medium, in Physics of the Inner

Marsch, pp. 99-171, Springer- Verlag,
R., Large-scale structure of the interplanetary
Heliospahere
I.,
edited by R.
Schwenn, and E.
Heidelberg, Germany, 1990.
Shain, C. A., 18.3 Mc/s radiation from Jupiter, Austr.
J.
Phys., 9, 61-73, 1956.
Sloanaker, R. M., Apparent temperature of Jupiter at a wavelength of 10 cm, Astron.

J.,
64, 346,
1959.
116
Smith, A. G., T. D. Carr,
observations of Jupiter, in
Sci.
and N.
F.
Six,
Results of recent decameter-wavelength
La Physique des Planetes,
pp. 543-550, Soc.
Smith, A. G., G. R. Lebo, N.
F. Six, T.
D. Carr, H. Bollhagen,
J.
May, and
Decameter-wavelength observations of Jupiter: The apparitions of 1961

Astrophys.
St.
Royale des
de Liege, Liege, France, 1962.
C,
Jupiter's
term periodicities.
Terasawa,
PhD
decameter and kilometer emissions:
Higher resolution studies of Jupiter's decametric radio emissions.
R.,
R.,
J.
and long
131, 1978.
dissertation, Univ. of Florida, Gainesville, Florida,
Thieman,
satellite effects
K. Maezawa, and S. Machida, Solar wind effect on Jupiter's non-Io
T.,
J.
1962,
dissertation, Univ. of Florida, Gainesville, Florida, 1985.
related radio emission. Nature, 273,
Thieman,
Levy,
141, 457-477, 1965.
J.,
Cyr, O.
J.
and
PhD
1977.
and A. G. Smith, Frequency and time dependence of the Jovian
decametric radio emissions:
nineteen-year high-resolution study,
J.
Geophys. Res.,
83, 3303-3309, 1978.
Thieman,
J.
R.,
A. G. Smith, and
J.
May, Motion of
Jupiter's decametric sources in Io
phase, Astrophysical Letters, 16, 83-86, 1975.
Van Allen,
J.
A., G. H.
intensity radiation
1958.
by
Ludwig, E. C. Ray, and C. E. Mcllwain, Observations of high
satellites
1958 alpha and gamma. Jet Propulsion, 28, 588-592,
BIOGRAPHICAL SKETCH
Leonard N. Garcia was born on August 21, 1967,
Marilyn Garcia. Leonard has two younger
sisters,
in
Miami, Florida
Rosie and Jenny.
watching Star Trek and Star Wars and looking back finds
He
attended Southwest
Miami High School and
from Florida International University.

graduate studies
Catholic and
known
to
is
at the
University of Florida.
a Knight of
On
how
to
begin his
Easter of 1994 Leonard
became a
to Gainesville in
1990
Leonard does enjoy good jokes and has been
Columbus.
some science
fiction
from Isaac Asimov and Arthur C. Clarke.
enjoys building models and studying his family
tree.
His favorite dish
is
steak with chopped onions and a bit of parsley on top and a side of platanos
moros. Finish
in the
any musical
to play
received his B.S. in physics
watch Saturday morning cartoons. Leonard has read the Lord of the Rings, the
Chronicles of Narnia, and
He
He moved
have always
that his interests
been about the same. Leonard plays no sports nor does he know
instruments.
Leonardo and
to
Leonard grew up
it
off with a flan
and
cafe',
a palomilla
maduros and
;Que Rico! Leonard found Miss Sarah Matestic
mountains of Kentucky (or did she find him?) and plans to share his
117
life
with her.
certify that
have read
this
study and that
standards of scholarly presentation and

dissertation for the degree of
is
in
fully
my
opinion
it
conforms
to acceptable
adequate, in scope and quality, as a
Doctor of Philosophy.
Thomas D.
Carr,
Chairman
Professor of Astronomy and Physics
certify that
have read
this
study and that

is
in
fully
my
opinion
it
conforms
to acceptable
Humberto Campins
Associate Professor of Astronomy
certify that
have read
this
study and that

is
in
my
opinion
fully adequate,
in
it
conforms
to acceptable
scope and quality, as a
George R. Lebo
Associate Professor of Astronomy
certify that
have read
this
study and that

is
in
fully
my
opinion
it
conforms
to acceptable
Alex G. Smith
Distinguished Service Professor of
Astronomy and Physics
certify that
have read
this
study and that

is
in
fully
my
opinion
adequate,
in
it
conforms
to acceptable
scope and quality, as a
Charles F. Hooper
Professor of Physics
This dissertation was submitted to the Graduate Faculty of the Department of

Astronomy in the College of Liberal Arts and Sciences, and to the Graduate School
and was accepted as partial fulfillment of the requirements for the degree of Doctor of
Philosophy.
December 1996
Dean, Graduate School

Long-Term Periodicities in The Jovian Decametric Emission

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Long-Term Periodicities in The Jovian Decametric Emission

Caricato da

Copyright:

Formati disponibili

LONG-TERM PERIODICITIES

THE JOVIAN DECAMETRIC EMISSION

LEONARD NORMAN GARCIA

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL

Carr, for his patience with

work with someone who was so instrumental

Jupiter's radio emissions.

and respected influence on our understanding of

thank the other members of

has benefited greatly from their helpful suggestions and recommendations.

thanks to Francisco Reyes,

planetary radio group.

the observatory, building antennas or

spent in his office

miss the times spent out

the lawn even the

could not understand a paper or

thanks for the hours

needed his opinion on

make some sense of

possible solar effects on the Jovian magnetosphere.

advice and encouragement are greatly appreciated.

had the opportunity

thank Kazumasa Imai for his helpful suggestions

during the time he spent with us in Florida.

Wes Greenman and

very thankful for the people

Space Flight Center.

Science Data Center for the opportunity to consult with them on

that all of those

has been enriched by

can not be certain that

education without them.

around ideas, when

Graduate Student Researcher's

of the graduate students

being there for

was stumped on a problem or when

by the work of the astronomy department

through a never ending bureaucratic maze.

who have always been

a tremendous source of support to

had the prayers of many people from around the country.

for the blessings

to withstand the strain of a graduate

would have been able

must thank Chuck Higgins and Liyun Wang,

great officemates and friends.

had the opportunity

Hunter and Glenda Smith

They have been

wish to acknowledge the National Solar Observatory/Sacramento

Peak for the monthly average sunspot numbers used

to thank the Office of University Programs

for their financial support through the

research and for

of course solely responsible for the conclusions presented

was uncertain of myself.

have received. Foremost among those whose prayers have touched me

She has shared

She has been there

and suffered through

Frequency Dependence of Jovian Radio Emissions