Re: DRA Project

From: Bart De Pontieu <BDP_at_MPE.MPE-GARCHING.MPG.DE>
Date: Tue, 9 May 1995 14:04:00 -0400

>> I may be interpreting too much here (but I know Walter will stop me if I
>> do that :-), but wouldn't the following indices (indicated with 'BDP')
>> be closer to reality?
>> Time WN BDP
>> 0.00 0 0
>> 20.36 1 1
>> 68.45 2 3
>> 114.67 3 5
>> 164.85 4 7
>But I fear, however, that you have exposed the possibility of
>a massive complication for an observer. Even with a completely clear sky,
>a tape running and close attention to recording detail upon it during the
>observing session, there is the inevitable likelihood that some events
>will be lost and omitted from the report.

I agree. However, if some of these events really *are* not recorded, that
is not necessarily a problem. Specifically, for DRA observations
(new subscribers may well want to read up on the Determination of Rotation
Axis Project before continuing - send a message with
Subject: archive get latest/290
to SeeSat-L-request_at_iris01.plasma.mpe-garching.mpg.de )
I will only use timings of which I think they agree with the assumptions
inherent to the software. This means : only flashes of the sides, no
secondary flashes, no timings for satellites that have two periods between
primary maxima, etc... So, before an observation is used in the DRA analysis,
it will be tested for these assumptions. It is thus very useful if the
observer tries to indicate which flashes were secondary, where he missed
some out, etc...
 
>I'm thinking here especially of
>wildly flashing objects such as 10820 = 78-42A = DMSP F3, 18958 = 88-20A =
>C* 1933, 19210 = 88-50A = C* 1953, and 23099 = 94-27A = SROSS-C2.

These objects are not part of the DRA program, partly because of the problem
you indicate in the above, but mostly because they are payloads so that I
have no clue where the flashes come from (i.e. which part of the satellite
causes them) and whether it is in a simple rotational state.
For our regular flash period observations, such detailed reporting is not
requested. Inherent in those observations is the demand that the observer
find some regularity in the timings and report that periodicity. If he fails
to observe regularity, he should just report 'irregular'. I or anyone
else in the Belgian Working Group Satellites (as far as I know) do not
have plans (at this time) to make a detailed study of payload lightcurves.
Hence our modest (i.e. just flash period, not light curve) requests for those
satellites. I know *you*, Walter, are studying a whole
class of payloads, and would probably like to receive more details of the
lightcurve, and I invite you to start a project aimed at studying those
satellites. SeeSat-L (or Flash) is the appropriate forum to ask for support
for such an undertaking. BTW, it may well be that some of our Benelux observers
(most of which are not on SeeSat) would join such a project.
It is obvious that the PPAS format is not suited for such a pursuit.
 
>If the observer is expected to make inferences such as you have made, he
>may well never complete the process of constructing the report! Is it not
>better to construct such "theories" from the observer's raw data?

I would like the observer to report as much as possible. He/she was there
and knows best what was seen. There is no obligation to 'construct' such a
detailed report, but it would save me a lot of time. Esp. since I know that
most of the observers in the DRA project are very experienced and can analyse
their data just as well as I can. Naturally this does not exclude less
experienced observers from contributing.

>You may recall I'd like to be able
>to report times for PPAS in hundredths of seconds for the same reason,
>namely, that it is more accurate to simply copy data out of my log than to
>process that same data into a specific format.
>Carried one step further,
>I'd like to be able to report raw times and the watch correction. How far
>does this go)?
 
Ignoring for a moment the enormous difficulties of changing an existing
format, all of your proposals shift the work-load from the observer to the
collector and that's why I don't like them. If we can't trust the observer
to correctly change '5.68' to '5.7', how then can we trust his observation
at all?

>Or are you suggesting that the report format needs to be fleshed out to
>support reporting of the facts which would support such inferences, such
>as brightness of each flash, duration of each flash, sharpness of each
>flash?

This is a very valuable suggestion. The report format I proposed will be the
one that ends up being fed to the software, but this should not mean that
you can't report magnitudes, duration of flash, etc... In fact, some
observers have done that, and it surely helps me with the analysis. So, please
do report as much as possible.

>Another aspect is the importance of avoiding any undue pressure on
>the observer to conform what he sees to what he can report, or to report
>only behaviour perceived as "formattable". I'm not sure this is possible.

A distinction has to be made between PPAS (regular flash period) observations
and DRA observations. Minimally, the DRA observations should contain flash
timings, nothing else. This means no pressure whatsoever is put on the
observer to 'format' the observation. He is invited to comment his observation
as much as possible, but that still doesn't mean he is forced to format his
observations.

>The PPAS format certainly does pressure the observer to report periods,
>but this accomplishes the scientifically useful task of inducing him to
>extract whatever regularity he can from whatever irregularity he observes.

I agree fully.

>It also accomplishes the scientifically destructive task of tossing out
>any useful record of the irregularity (I'm not referring here to the
>remarks he may append).

Agreed, there is room for improvement. As you indicate, the PPAS format
does allow for comments, even comments longer than 25 characters are
included into our archive (in the PPAS.REM file). A large group of objects
*is* regular and can be reported on by using the PPAS. But even so, the PPAS
format is not suitable for a certain group of objects or passes where
regularity is hard to discern. See my comments above about such 'irregular'
objects.

Sorry for going on so long. I hope the subject still interests some of you.
If you think it shouldn't be on SeeSat (which I could understand), let me
know, and we'll reply through private mail. SeeSat is still in a forming
phase. Its contents are to be determined by all of the 100 subscribers, not
just by me, Walter or any of the other seesat-old-timers. If you think SeeSat
should change, don't just unsubscribe, I invite you to propose changes.

Cheers,
      Bart
Received on Tue May 09 1995 - 14:05:20 UTC

This archive was generated by hypermail 2.2.0 : Fri Mar 07 2014 - 00:14:29 UTC