Content-Type: multipart/alternative; boundary="------------EB96E91586A575846B1C3D57"
X-Sun-Content-Length: 14932


--------------EB96E91586A575846B1C3D57
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Dear Jolene,

Yet again, apologies for the delay in replying about the correlator
data. You are correct in saying that the correlator plots I sent were
only for the
fixed energy ACF (usually set at the lowest correlator energy
band)  for S/C 4 . I've attached Correlator data for the stepped
energy ACFs for S/C 3. The first 2 panels of the figure 1 show fixed
energy correlator data (45 eV) as before , panels 3-6  all correspond to

stepped energy data (panel 3 = zero-lag amplitude, panel 4=
corresponding
energy, panel 5= non-zero lags, panel 6= FFT spectrogram of non-zero
lags).

Due to only one summed stepped energy  ACF being outputted
per spin , one way to view panel 3 is as a diagonal sampling slice
(running from high to low
energies) through the HEEA energy/time/flux spectrogram plot you sent
(the actual eV
energies being plotted in figure 1, panel 4). Thus the high PEACE flux
at 100-200 eV between 19:23:00
and 19:24:20 in your plot is reflected in the zero-lag  correlator
measurements (panel  3) showing a peak
at about 200 eV (at about 19:23:30), the same goes for the high flux
seen in the HEEA spectrogram
at about 80 eV (between 19:25:20 19:25:50) corresponding to the 80 eV
peak in the zero-lag panel 3 at
about 19:25:30. Panel 5 shows the non-zero lags (with the 31 lag values
running up the vertical axis between
0.012 and 0.37 milliseconds) and the corresponding FFT of those values
(panel 6, vertical
axis running from 1.4 to 42 kHz) for these sampling slices. The actual
count rate for electrons (deduced from the correlator
data) in the 100-200 eV band is about 30000 counts/second. Unfortunatley
this electron
sampling rate is too low to unambiguously show time structure at
frequencies between  1.4 to 42 kHz
without doing averaging of the correlator data from subsequent
correllator energy sweeps
where it is known that the bi-polar pulses are present, to see if a
repeated frequency becomes
appearent.

In figure 1 the non-zero lags have been normalised by the mean of the 31
lag values so that
values oscillate about the value 1 and thus  the amplitude scale  shows
time structuring independent
of count rate (i.e. independent of where in the energy sweep you are).
Figure 2 (panel 2) shows
non-zero lag data which has not been normalised and thus
allows the isdat igr spectrogram filter to pick out features as a
function of
count rate (i.e. high flux is shown is all red, low flux is all black
and the intermediate energies
between 100 - 300 eV have a color scale).  Figure 3 shows the non-zero
lags (panel 2) and
corresponding FFT (panel 3) for a 30 second interval encompassing the
WBD waveform
plots you sent. The figure 3, panel 3 FFT plots between 19:24:20 and
19:24:28 indicate possible electron
coherency at energies 100-200 eV on timescales of  12-15 kHz (although
we would have to compare
a lot more times of bi-polar events and non bi-polar events to say
anything conclusive).

As I  mentioned in my last email what might be possible is to compare
ratios of counts measured by
PEACE with ratios of counts measured by the zero-lags in the correlator
which are biased towards giving high
<counts>**2 amplitudes when the electrons are NOT evenly distributed in
time , this ratio being
larger when the electron counts are bunched closer towards the basis
series of electron count samples used
by the correlator (of duration 0.37 milliseconds). This is only an idea,
which I'll test out once we have
the PEACE-ISDAT software working (hopefully next week when our engineer
returns, ... I've requested  last
week some PEACE IDFS data from Andrew Fazakerly for the 18/2/01 event).
Presumably the electrons
are bunched in the potential well of the 1 millisecond bi-polar pulses ,
or at least have structuring approximately on this
time scale. If this idea works it should then (at least from the
electron dynamics point of view) be
possible to use the correlator data to survey when this ratio goes high
as an indicator of short,
millisecond time scale electron structuring (and hence association with
the bi-polar pulses you measure
in the Wideband electric field waveforms). I'll email back to you when
we've got the software working
that can plot this ratio. Incidently, you are very welcome to have our
current ISDAT-DWP module for displaying
correlator data (although I will update and improve this module as time
goes by), you should be able to
down load it using CVS from the Uppsala server, but it might be easier
if I send you the tar file (you then
just have to unpack the tar file in isdat/server/Wec/Dwp and run
"make" in the highest level isdat directory
to compile a new dbh executable, the package also comes with
documentation).

If you have managed to do any survey of intervals when you do and do not
see
the bi-polar pulses we can perhaps  use this to compare with the
corelator data in a more detailed fashion.
The plot that Mel Goldstein generated of the velocity distribution
showing the counter streaming
beams looks  compelling (I notice  the plot is constructed from 0.17
seconds ? , approx
1/24 th spin, presumably it must use simultaneous HEEA and LEEA data to
obtain both look directions).
Incidentally, the correlator uses the PEACE field aligned zones (as
determined from the previous spin) to
supply counts to the correlator (the zone is switched each half spin to
the it's opposite number to maximise
the counts on the assumption that the electrons are predominantly field
aligned, by running some software
on the PEACE CDROM data we can find out exactly which PEACE polar zones
were used to supply the correlator).

Once again, apologies in not being able to reply sooner.

Best Regards,

Andy

ps. Yes, I will be going to the Cluster Workshop in October, as will
several other people from Sussex, will
you be attending ?

--
********************************************************************
 Dr. Andrew  Buckley, Research Fellow,
 Space Science Centre,
 School of Engineering and Information Technology,
 The University of Sussex, Falmer
 Brighton, E. Sussex, BN1 9QT, UK.
 Email:     A.M.Buckley@sussex.ac.uk
 Telephone: 01273 678599
 Fax:       01273 678399
********************************************************************



--------------EB96E91586A575846B1C3D57
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit

<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
<body text="#000000" bgcolor="#FFFFFF" link="#0000FF" vlink="#FF0000" alink="#000088">
Dear Jolene,
<p>Yet again, apologies for the delay in replying about the correlator
<br>data. You are correct in saying that the correlator plots I sent were
only for the
<br>fixed energy ACF (usually set at the lowest correlator energy
<br>band)&nbsp; for S/C 4 . I've attached Correlator data for the stepped
<br>energy ACFs for S/C 3. The first 2 panels of the figure 1 show fixed
<br>energy correlator data (45 eV) as before , panels 3-6&nbsp; all correspond
to
<br>stepped energy data (panel 3 = zero-lag amplitude, panel 4= corresponding
<br>energy, panel 5= non-zero lags, panel 6= FFT spectrogram of non-zero
lags).
<p>Due to only one summed stepped energy&nbsp; ACF being outputted
<br>per spin , one way to view panel 3 is as a diagonal sampling slice
(running from high to low
<br>energies) through the HEEA energy/time/flux spectrogram plot you sent
(the actual eV
<br>energies being plotted in figure 1, panel 4). Thus the high PEACE flux
at 100-200 eV between 19:23:00
<br>and 19:24:20 in your plot is reflected in the zero-lag&nbsp; correlator
measurements (panel&nbsp; 3) showing a peak
<br>at about 200 eV (at about 19:23:30), the same goes for the high flux
seen in the HEEA&nbsp;spectrogram
<br>at about 80 eV (between 19:25:20 19:25:50) corresponding to the 80
eV peak in the zero-lag panel 3 at
<br>about 19:25:30. Panel 5 shows the non-zero lags (with the 31 lag values
running up the vertical axis between
<br>0.012 and 0.37 milliseconds) and the corresponding FFT of those values
(panel 6, vertical
<br>axis running from 1.4 to 42 kHz) for these sampling slices. The actual
count rate for electrons (deduced from the correlator
<br>data) in the 100-200 eV band is about 30000 counts/second. Unfortunatley
this electron
<br>sampling rate is too low to unambiguously show time structure at frequencies
between&nbsp; 1.4 to 42 kHz
<br>without doing averaging of the correlator data from subsequent correllator
energy sweeps
<br>where it is known that the bi-polar pulses are present, to see if a
repeated frequency becomes
<br>appearent.
<p>In figure 1 the non-zero lags have been normalised by the mean of the
31 lag values so that
<br>values oscillate about the value 1 and thus&nbsp; the amplitude scale&nbsp;
shows time structuring independent
<br>of count rate (i.e. independent of where in the energy sweep you are).
Figure 2 (panel 2) shows
<br>non-zero lag data which has not been normalised and thus
<br>allows the isdat igr spectrogram filter to pick out features as a function
of
<br>count rate (i.e. high flux is shown is all red, low flux is all black
and the intermediate energies
<br>between 100 - 300 eV have a color scale).&nbsp; Figure 3 shows the
non-zero lags (panel 2) and
<br>corresponding FFT (panel 3) for a 30 second interval encompassing the
WBD waveform
<br>plots you sent. The figure 3, panel 3 FFT plots between 19:24:20 and
19:24:28 indicate possible electron
<br>coherency at energies 100-200 eV on timescales of&nbsp; 12-15 kHz (although
we would have to compare
<br>a lot more times of bi-polar events and non bi-polar events to say
anything conclusive).
<p>As I&nbsp; mentioned in my last email what might be possible is to compare
ratios of counts measured by
<br>PEACE with ratios of counts measured by the zero-lags in the correlator
which are biased towards giving high
<br>&lt;counts>**2 amplitudes when the electrons are NOT evenly distributed
in time , this ratio being
<br>larger when the electron counts are bunched closer towards the basis
series of electron count samples used
<br>by the correlator (of duration 0.37 milliseconds). This is only an
idea, which I'll test out once we have
<br>the PEACE-ISDAT software working (hopefully next week when our engineer
returns, ... I've requested&nbsp; last
<br>week some PEACE&nbsp;IDFS&nbsp;data from Andrew Fazakerly for the 18/2/01
event). Presumably the electrons
<br>are bunched in the potential well of the 1 millisecond bi-polar pulses
, or at least have structuring approximately on this
<br>time scale. If this idea works it should then (at least from the electron
dynamics point of view) be
<br>possible to use the correlator data to survey when this ratio goes
high as an indicator of short,
<br>millisecond time scale electron structuring (and hence association
with the bi-polar pulses you measure
<br>in the Wideband electric field waveforms). I'll email back to you when
we've got the software working
<br>that can plot this ratio. Incidently, you are very welcome to have
our current ISDAT-DWP module for displaying
<br>correlator data (although I&nbsp;will update and improve this module
as time goes by), you should be able to
<br>down load it using CVS from the Uppsala server, but it might be easier
if I send you the tar file (you then
<br>just have to unpack the tar file in isdat/server/Wec/Dwp and run "make"&nbsp;in
the highest level isdat directory
<br>to compile a new dbh executable, the package also comes with documentation).
<p>If you have managed to do any survey of intervals when you do and do
not see
<br>the bi-polar pulses we can perhaps&nbsp; use this to compare with the
corelator data in a more detailed fashion.
<br>The plot that Mel Goldstein generated of the velocity distribution
showing the counter streaming
<br>beams looks&nbsp; compelling (I notice&nbsp; the plot is constructed
from 0.17 seconds ? , approx
<br>1/24 th spin, presumably it must use simultaneous HEEA&nbsp;and LEEA
data to obtain both look directions).
<br>Incidentally, the correlator uses the PEACE field aligned zones (as
determined from the previous spin) to
<br>supply counts to the correlator (the zone is switched each half spin
to the it's opposite number to maximise
<br>the counts on the assumption that the electrons are predominantly field
aligned, by running some software
<br>on the PEACE CDROM data we can find out exactly which PEACE&nbsp;polar
zones were used to supply the correlator).
<p>Once again, apologies in not being able to reply sooner.
<p>Best Regards,
<p>Andy
<p>ps. Yes, I will be going to the Cluster Workshop in October, as will
several other people from Sussex, will
<br>you be attending ?
<pre>--&nbsp;
********************************************************************
&nbsp;Dr. Andrew&nbsp; Buckley, Research Fellow,&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
&nbsp;Space Science Centre,&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
&nbsp;School of Engineering and Information Technology,&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
&nbsp;The University of Sussex, Falmer&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
&nbsp;Brighton, E. Sussex, BN1 9QT, UK.&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
&nbsp;Email:&nbsp;&nbsp;&nbsp;&nbsp; A.M.Buckley@sussex.ac.uk&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
&nbsp;Telephone: 01273 678599&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
&nbsp;Fax:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 01273 678399&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
********************************************************************</pre>
&nbsp;
</body>
</html>

--------------EB96E91586A575846B1C3D57--