[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Significance of TASS Data




Date: Thu, 14 Sep 2006 23:01:35 -0500
From: Thomas F. Droege <droege@fastmail.fm>
To: tass@tass-survey.org
Subject: Significance of TASS Data

Hello Martin,

You write:

"I do have to say that I am surprised - on reflection disappointed is
perhaps a better word - that
stars with a range in the downloadable data of around 0.7 magnitudes
cannot be said to be of any
particular significance."

Just what do you mean by this statement?  Do you mean that you have
studied the entire tass data set and come to the conclusion that it's
accuracy is 0.7 mag?  I would dispute that. For some information on the
relative accuracy of the tass data see the curves in the S&T Note at:
http://www.tass-survey.org/tass/showtell/st0010.html

Since you have made a statement about the quality of the data, I would
like to see you defend your conclusion. I suggest that you read some of
the over 100 tech notes where we have investigated the data.  As you
will see over time we have made many improvements.

This is a very large data set.  By the normal rules of statistics, some
data points will will take advantage of all those measurements and will
be far from the expected value.

There are also funny things that happen when you don't look at the raw
images.  We have most of the raw images and are looking for someplace
for them to live.  (They may actually have a home.)

As an example, some time back I got very excited when I discovered
several stars that brightend a magnitude or so.  After frantic study, it
turnd out that an asteroid had wandered by and just happened to be
sitting on the subject star.  This has happened several times now.
Satellites also can have a similar effect, so one has to investigate
funny data points.

Still, taking into account the several hundred million measurements, the
data is much better than the 0.7 magnitude that you state. This is a
data set with well explained characteristics.  We don't do your work for
you.

Tom Droege

-- 
  Thomas F. Droege
  droege@fastmail.fm