Subject: dust shutdowns


Submitted: Thu, 2 Oct 1997 17:20:58 -0600 (MDT)

Message number: 181 (previous: 180, next: 182 up: Index)

Dust shutdowns

I have looked at Gordon Richards plots, and would like to point out a
number of issues. 

1. While Gordon, for the period 05/22-07/24 would infer an average
shutdown percentage of about 10, the actual night time shutdown
percentage to dust only (not including other weather) for that period,
according to Gloria's graphs is at least 15%. Since weather and dust
may go together, I suspect that the shutdown criterion is actually
reached many more times during the nights than 15% of the time. Either
we have more dust (or humidity ....) at night, or we shut down below
1600 a significant number of the time. I would expect, with winds
dying down in general after sunset that dust content would be less at
night, not more.

2. The lack of correlation between 1 and 0.3 micron dust implies that
our shut downs appear to be completely ineffective at avoiding small
dust particles. However, I do not advocate shutting down based on
counts for 0.3 micron particles, not knowing which particles settle
down most and are most damaging. But this does illustrate that we may
not know what we are doing.

3. I would rephrase Gordon's conclusion regarding the effectiveness of
the shutdowns. In spite of shutting down 10 to 15% of the time for
dust only, we appear to avoid no more than 20% to 30% of the possible
total dust accumulation. Even if we assume that 1 micron dust
accumulation is the primary cause for a drop in throughput of, say,
15% per year, this implies that by shutting down 10 to 15% of good
observing time, we stretch realuminization from every 2 yrs (if we
want to realuminize when we reach 70% or so of optimum throughput) to
maybe every 2.5 to 3.0 years or so. If realuminization takes 1 month,
then we save 1 month every 5 to 6 years, but we shut down the
telescope for dust at least one month per year, so we have lost 6
months over that 6 year period due to dust shutdowns. Even if I am off
by factors of two, we appear not to be doing very well under the
current scenario. Even if mirror degradation happens faster than in
the example above, we would be losing out under the current situation.

What would be worth doing is repeat Gordon's analysis for the night
time data, and add up exactly for the dust shutdowns how much dust we
avoided. One reason for doing this is that my analysis ignores dust
accumulation avoided for shutdowns other than dust (so weather,
instruments, etc.) What we really need is accumulation during time
that we actually observed, and compare that with the fraction of dust
avoided by shutdowns only for dust conditions.

I agree with his final conclusion: In my mind we have not much reason
to believe that we are doing anything better than a regular flash
light test at this point, and we are likely shutting down too
much. There are many unknowns in this whole issue. Given this, I see
no reason why we should use a different criterion for dust shut downs
than used at other observatories, until we actually understand the
dust monitoring figures. I am in favor of continuing the dust
monitoring to learn more what we may be measuring. But I suggest that
we readdress our shutdown policy based on dust.

					Rene Walterbos
APO APO APO APO APO  Apache Point Observatory 3.5m  APO APO APO
APO  This is message 181 in the apo35-general archive. You can find
APO  the archive on
APO  To join/leave the list, send mail to
APO  To post a message, mail it to