|
|
|
____________________
|
|
|
|
FAQ
- Why do I have to do this?
- A variety of reasons.
- The OMP project is providing a software system, of which the
JCMT-OT is a necessary component, to manage flexible scheduling at
JCMT. Its purpose is to increase overall observing efficiency and
facilitate communication with the PIs, and at the same time reduce the
workload on our overworked queue managers and TSSs. It will also allow
you, the PI, to control your observing strategy in a fashion similar
to that of classical scheduling. Note that the MSBs that you specify
directly run the telescope and the data-acquisition during the
observations.
- The JCMT-OT and UKIRT-OT present a unified user interface to all
JAC observing
- What is the deadline for submissions?
- There isn't one. However if you don't submit MSBs for your
sources
by the time they are observable you could be throwing away
opportunities of having your sources observed. Nevertheless, note
that there is no requirement that you submit your entire project
at once. Therefore you can submit some MSBs, wait for them to
get done so you can analyze the results and then submit other
MSBs to follow up on those results. Your MSBs will keep on being
scheduled until you have exceeded the total time allocated to your
project.
- What happens to my MSBs
after I submit them?
- During flexible observing, the observer/TSS will run a tool
called JCMT-QT (Query Tool) that allows them to browse our database
for suitable MSBs. The QT by default returns as yet unobserved MSBs
ranked by TAG priority containing targets that are actually
observable, whose site quality requirements match the current
conditions, and whose projects still have allocated time left. However,
the observer/TSS is able to use the QT to retrieve any submitted MSB
at their discretion.
When an MSB is selected for actual execution, the Query Tool can be
used to retrieve it from our database. The observer/TSS is then
presented with a list of the observations in that MSB and given the
opportunity to remove redundant calibrations from the MSB or to defer
them until a suitable calibrator becomes available. When they are
satisfied that the MSB is suitable for execution they send it off to
the instrument-specific control system.
The MSB may be send to a translator which converts it into a complete
set of instructions for the telescope, receiver(s), and backend.
The instructions are then placed in a queue that can be processed. As
the queue works its way down the list it takes any sensible course of
action, such as prompting the observer/TSS when intervention is
needed.
After the observations have been carried out and deemed acceptable by
the TSS, the MSB is marked as done and will not be rescheduled unless
it is resubmitted from the JCMT-OT or marked Undone via the project
feedback system. If repeats of the MSB were requested, the repeat counter
is decremented by 1.
- What makes a good MSB?
- First the usual things, accurate co-ordinates, suitable
calibrations, good chop positions etc.
The aspect of the MSB you most want to think about is its
length. Let's say you have been allocated eight hours for a
single source. Now, you could submit a single 8-hour MSB if you
wanted. But that would be a very silly thing to do. For one thing, if
your source is not up for more than eight hours, it would never be
scheduled as our software would detect that it would set before the
MSB could be completed. But even if your source were observable for 8
hours solid, the observer won't chose to undertake it except near the
very beginning of the night, which vastly decreases your scheduling
opportunities.
At the other end of the scale, using up your 8 hours as 48 10-minute
MSBs is a grossly inefficient use of telescope time. If the 48 MSBs
happen to be scheduled on different nights the time spent on overheads
will be huge and that time will come out of your project
allocation. On the other hand, if they all happen to be scheduled one
after the other they would have to be sent to the queue one by one,
thus driving the TSS and observer insane. (But perhaps for astronomers, that
last statement is redundant.)
The golden mean? A 60-to-90 minute MSB, with 30-to-120 minutes still
quite reasonable. If you believe your program cannot be sensibly
constructed in MSBs that fall in that duration range please consult us
for ideas.
- My MSBs only last half an hour so it is
a waste to do a calibration for each one!
- Don't worry about this. The person doing the observing will have
the choice of not doing calibrations that are highly redundant.
- I want to do something that is not in the
library
- You certainly can. However I suggest you let us know for two
reasons. First, for your benefit, we'll be able to look your
MSBs over to ensure that they will actually do what you intend.
Secondly, for our benefit, we would like the chance to
create a library MSB that matches your observing mode in case
that will be useful to other users.
- What does this warning mean ? :
"Some of the observations in the MSB contained
default settings for the target information"
- You probably forgot to enter your co-ordinates! If you really
meant to observe 0:0:0, 0:0:0 just ignore it.
- I don't agree with the time estimates shown
in the OT
- We are fine-tuning these, but they will always be just that -
estimates. However it doesn't really matter if they are a bit off. The
estimates are used for three things. First, they are used by the QT to
check whether the observation can be carried out before the source
sets. Second, they allow the observer/TSS to plan the observing. For
example, if they believe that the tau is going to change in the next
hour, they can select an MSB that is estimated to last less than an
hour.
Third, time estimates are used to provisionally debit remaining time
during the night to keep a running tab on projects' use of their
allocation. However the final time accounting is done the next day
based on the actual time used by a project as derived from the data
files and logs. If you find you still have time left after that you
can submit an MSB to use it up.
Please notify us if you believe the time estimate to be grossly in error.
|
|
