Subject: DIS system thruput history

From: elt@astro.Princeton.EDU

Submitted: Fri, 24 May 1996 17:30:57 -0400 (EDT)

Message number: 17 (previous: 16, next: 18 up: Index)

In the context of reported low and decreasing throughputs of DIS, it has
several times been suggested (most recently by Michael Strauss) that the
lens monitoring group use its repeated observations of the lens fields 
which have been obtained regularly starting in December 1994 to look for
such a decline in sensitivity.

We have now done such an exercise and the results are displayed in the
4 gzip compressed postscript file plots available in the elt/:disthruput
subdirectory of the anon ftp area on astro.princeton.edu .  These plots
show raw count rates for each of the 9 secondary photometric comparison
stars in our 0957+561A,B field, separately for the g and r bands.  In each
case the count rates have been divided by the average value for the star
in question over all of the frames displayed (a convenience for display
purposes since the actual rates vary over about 2 orders of magnitude).
In the g.ps and r.ps plots, the data for all the stars is superimposed,
while for the gpara.ps and rpara.ps the data for each star is offfset by
an additive constant to make the nearly identical behavior of each more
apparent.

These data are not so suitable as one might first think for this purpose 
because our programs concerns itself only with *relative* photometry of
the lens images with respect to the secondary standards in the field.  This
allows us to take data in (sometimes grossly) non-photometric conditions,
to ignore air mass corrections and to use aperature photometry despite
variations in seeing.  All these factors, which cancel out for purposes of
relative photometry, do affect the count rates plotted.  Thus, for any
given star, we see a large spread in count rates depending on observing
conditions; these variations cannot be interpreted as DIS throughput
variations on a point-by-point basis.

Nevertheless, there are suffciently many observations that long term
changes are still apparent.  Basically, the upper envelope of the
distribution of count rates for a given star represents observations
made in relatively good (photometric and seeing) conditions.  Points
taken with air mass > 1.2 have also been excluded from the plots,
although this makes no qualitative difference, it does decrease the
scatter somewhat.

The data show a definite decrease in the DIS throughput (really the
total efficiency of the whole system including telescope, etc.) in
the r band between the 1995 and 1996 0957+561 seasons.  The amplitude
of the decrease is at least 30% or so.  Moreover, the decline
appears to be continuing and perhaps accelerating in the 1996 data.
In the g band, any difference between 1995 and 1996 is less apparent,
but there is an apparent decline during 1996.  The history of the
counting rate for different stars is very similar which gives one
some confidence that the measurement is meaningful.

While these data do support the contention that the DIS system throughput
(in imaging mode, at least) has dropped substantially since the instrument
went into regular use, the g vs. r difference is puzzling.  Typically,
reflectivities degrade more rapidly in the blue than the red.

Tomislav Kundic
Ed Turner
APO APO APO APO APO  Apache Point Observatory 3.5m  APO APO APO
APO
APO  This is message 17 in the apo35-dis archive. You can find
APO  the archive on http://astro.princeton.edu:82/apo35-dis/INDEX.html
APO  To join/leave the list, send mail to apo35-request@astro.princeton.edu
APO  To post a message, mail it to apo35-dis@astro.princeton.edu
APO
APO APO APO APO APO APO APO APO APO APO APO APO APO APO APO APO