The Problem was SETI, now STI

[Tom Droege <tdroege2@earthlink.net>]

Let me try to outline the problem of the moment.

There are 5 Mark IV systems that can presently run and take data.  One of 
them, TOM1 has taken most of the data available, about 18,000 images stored 
on 350 CDs.  TOM2 and TOM3 are near operation, as is ROB.  MICHAEL, ARNE, 
and MIKE could take or are taking data.

MICHAEL, ARNE, and MIKE are in the hands of people who know how to process 
data if they take it.  They also understand what they are doing and be 
relied on to not make processing calibration errors.  The TOMs are in the 
hands of someone (me) that will take some time (a year or so) before he is 
up to speed in processing the data.  I am sure I could push data through 
Michael's pipeline sooner than a year.  The problem is in understanding 
what I am doing.  That will take longer.  It is dangerous to process a lot 
of data without understanding all the steps.

There are a dozen or so people in the group with the required skills to run 
a pipeline, but who do not have data.  There are a smaller number who have 
both the computer skills and the experience in astronomical calibrations to 
generate a reliable result.  Many of these have Mark IVs.  They have them 
because I hope they will run them and start working on the calibration 
problems.

Eventually I (and others) will be able to process data as we take it.  This 
is not true today.  If the present data were reduced to star lists, then we 
could start looking for interesting things.  Doug has proposed one thing to 
study.

The longer the time history is, the more valuable this data becomes.  That 
is why I am taking data even though I can not yet process it.

My seti proposal was an attempt to get started on the processing.  This is 
not yet just pushing CDs through a pipeline.  It has to be done with 
intelligence.  There is much to learn about error.

Many of you on this list have the required computer skills to start 
processing this data.  I suspect that most of you do not (yet) have the 
skills in astronomy that are required to generate a result that has the 
best possible calibration.  Those of you with Unix skills are one step 
ahead of me.  I have to learn Unix skills, then I can start processing data 
where I will learn calibration skills.

Why not have a few of you start learning the calibration skills?  There are 
a number of people on the list that can help with the real problems.  As I 
have mentioned, this is the level of work that you do as a graduate student 
to get a PhD these days.  One takes course work, then joins some experiment 
and then sits in front of a terminal grinding data for a couple of years 
under the supervision of experienced experimenters.  We have experienced 
experimenters, we have data, what is needed is more experienced programmers 
sitting in front of terminals grinding on the data.  I promise you that you 
will learn a lot.  The advantage of this data is that it does not require 
very much in the way of course work to get to the point of having an 
understanding of what you are doing.  One needs to learn something about 
statistics, and error, but most of you here have educations that at least 
touched on these topics.  You can learn what is needed.  (Experts might 
disagree and might want to comment.)

OK, I will comment.  It is easy to generate a lot of bad data from the 
images that have been taken.  Generating a reliable result will take 
special skill.  I think there are enough of you on this list that one or 
two will have what it takes.  The way to find out is to have a go at 
it.  There are lots of examples in the tech notes of how experts go about 
this type of analysis.  As I keep saying, you will learn a lot if you 
try.  These days generating a piece of experimental equipment, running it, 
and analyzing the results are all separate skills.  Not many do all three 
well.  These days there are teams of hundreds (at Fermilab) doing 
each.  Somewhat smaller groups at observatories.  With remote operation, I 
suppose you can get a PhD in astronomy these days without ever having 
visited a mountain or looked through a telescope.  I suspect most advisors 
will push you to do it once.

OK, seti has been set up so that the processors can do it mindlessly.  That 
is not the problem here, so I was wrong to refer to seti.  This is a 
"Search for Terrestrial Intelligence."  STI.  I am looking for a few good 
data analyzers to start working on this data.

Tom Droege