Weather at the PROMPT Telescopes

How does the scheduling algorithm decide whether or not to close the telescopes? I gather there must be only rudimentary ability to discriminate based on partial cloud cover or haze since there seem to be a lot of images exposed when conditions are cloudy. What are the long term plans? Is it possible to tie the PROMPT telescopes to whether or not some CTIO telescope is in use?

On a (maybe) related issue — have you thought about mounting some kind of inexpensive nightsky domecam that would show the sky, perhaps even placing an occasional image on the web? (or maybe even replace the PROMPTcam with a PROMPT nightskycam now that construction is done)? I've found that the cheap sky camera that we have at TCO is so good that I rely on it more than I rely on my own eyes to pick up incoming clouds (only cost a couple hundred $$). Hey I'd even chip in for one……

Question about Scheduling

Aaron, Dan, et al — thank you for the impressive software upgrade.  Huge help to an already-impressive system.  Two questions regarding the algorithm:

 1.  If a sequence of, say, 4 cycles of observations are scheduled and Skynet only finishes 3 of them on a single night (for whatever reason):  will Skynet return to complete the sequence on following nights?

 2.  How do successive cycles of observations within a single job deal with the airmass limit?  For example, if the first two blocks of observations are under the airmass limit but the third block of observations exceeds it what happens?  (Will it take only the first two?  Will it kill all 3? THrow up?  What?) 

Solar System Objects

It would be nice to be able to observe solar system objects.

Currently if we type in a planet, we tried "jupiter", it can't find it. I think this is because the RA and Dec are time dependent.

In the future, it would be nice to use PROMPT to study comets, asteroids, and other such stuff — not just letting students observe the major planets for class.

Thanks so much! — Jonathan

Telescope and software announcements

P3 and P5 have had their cameras return, so they are back online. We are finetuning some of the software, so they may not be perfect right away, but feel free to schedule jobs for them.

We have deployed new versions of the software that control the PROMPT scopes, so there might be a bug or two that pop up. Please report any oddities that you notice.

There was an update to the web interface as well. In addition to some small cosmetic changes, we now have an early version of scheduled observations. You can now specify a minimum start time for your observations (so tell it not to start until after 1:00 AM for instance) in addition to the solar angle and airmass constraints. You can also make the job periodically repeating (take these same 3 V band exposures every hour for a total of four sets, for instance.) You should be aware that due to the nature of que based scheduling, there is no guarantee that your observation will happen at exactly 1:00 AM or that it will repeat in exactly 1 hour (in the above examples) since there may be a higher priority job in your way. But this should provide you with greater flexibility for certain types of science.

–Aaron

Suggestion

I don't imagine any of the Prompt programmers are (exactly) looking for things to do, but here is one item for the list that you might want to think about. Someday. It would be nice to be able to queue up an object for 2 or 3 separate images at different times on a given night, or be able to take images on successive nights but not separated by, always, the same time interval (which leads to aliasing in the kind of work that I am doing, and possibly Dan Caton). I do that scheduling now in a crude way by submitting jobs with different airmass limits, and that sorta works. But the minimum airmass of 1.5 is significantly too large. Even if it were 1.25 or 1.3 it would help me enormously as I might be able to tweak the system to give me exposures within an hour or so of the meridian — though I'm not sure I understand how the existing algorithm schedules. I'm also wondering if the 80 second maximum exposure might be raised sometime — especially since it seems to me that tracking on P4 is better these days.

FTP Site

Is the SKYNET ftp site up, and capable of providing data downloads?  Would someone post its address here?  Thanks.

Prompt 2

Does anyone know if prompt2 will be back online tonight? Is there some good way to find out the plans for a situation like this (so I can know whether I should bother to submit jobs for it)? Perhaps some kind of regular posting that gives us the general outline of when a telescope will go down, or return?

Why Jobs Remain in the Queue

Is there some way to see what caused a night’s observations to terminate?  I’m watching a block of my jobs from Feb 22 that should reach the telescope early in the morning each night, but don’t (as of Feb 26).  Could there be some way to enter into the log if clouds terminate the use of some telescope?  Or do jobs that have been waiting for some nights get notched to a lower priority?  Would I be better off to resubmit a job that doesn’t run after a couple of nights?

Job Listing in Skynet “Observations”

Instead of renumbering the blocks of 10 recent jobs starting with “1″ why not reverse the order?  This way paging back through jobs to find an old job would be easier since you could look in, say, block 91 to find job 913.