Duties for Hectospec Observers -
Responsibilities and Observing Protocols
Feb 18, 2005
Revised Nov 14, 2005
I. Responsibilities
We believe that the queue observing
mode for Hectospec and Hectochelle
has been a scientific and operational success. These
queues are run somewhat
differently than those elsewhere; here are the main concepts:
(1) each project is allocated queue time equal to the number of
assigned nights, where a "night" is equal to the average length
of all the Hecto
nights.
(2) As much as possible, during the first 3-4 nights, some data
will be obtained for each of the projects.
(3) After that period, we will try to maintain an equal fraction
of completion (time_spent/time_allocated) for all projects. In
this manner,
bad weather is evened out among the projects. Some variation from
equality
will naturally
occur during the brighter moon phases and marginal conditions, times
which we tend
to observe targets that can withstand such battering.
(4) Each configuration is charged
20 minutes overhead; each exposure is charged 1.5
minutes. Time spent on a target
where it is determined the configuration was
faulty (e.g., bad guide stars) is charged to the project.
Time spent during marginal conditions on ultimately worthless spectra
is not charged (so one
needs to look at the data soon after it is taken).
(5) Flats, comparison exposures etc are taken during the
afternoons. Flux & velocity standards are not routine, and
will be charged to the project.
Up until 2005, the Hecto team and FLWO
staff have supported the operation of the queue in four ways: (1) SAO
scientists and engineers have maintained andserviced the instruments as
necessary, (2) Nelson Caldwell has scheduled the queue observations,
(3) Perry Berlind and Mike Calkins have operated the robots, and (4) a
Hecto team member has typically been present to provide nightly
scientific supervision. For now,
we intend to keep the first three elements in place, but hope to turn
thenightly scientific supervision over to trained observers from the
pool of those with assigned Hectospec and Hectochelle time. We
have requested that observers come to the telescope during their
assigned time in the past, but would now like to clarify what is
expected of observers and have them come to the telescope in blocks of
time that make better sense for queue operation.
We would plan to divide each Hecto run into blocks of ~3 nights which
would be managed by one or two observers. The nights will not
necessarily correspond to the assigned nights on the telescope
schedule. We
have the freedom to shift the times around for the convenience of
observers and the queue. During the assigned nights, the astronomer
would attend to the following items, which center around insuring that
good quality data are obtained.
Observer Responsibilities
- Run the spectrograph/CCD acquisition control software
- Annotate the data logs (now under automation), with comments
on conditions, data quality, problems encountered, etc.
- Check the operation of the spectrograph/CCD at the beginning of the
night, and monitor readout noise, spectrograph focus, thermal
flexure, etc. Normally, the actual focussing would be done
by the robot
operators, Perry and Mike, who would also fill the CCD
dewar.
-
Be knowledgeable about the fiber assignment code "xfitfibs", in
particular with regard to the retrictions on rotator position
and
guide star selection, to the extent of being able to run the
program
at the telescope should the need arise.
- Be knowledgeable about the normal sequence of operating the
positioner and acquiring fields, so that when problems with
acquiring a field occur, the robot operators can be advised as to
how to proceed (e.g., moving on to another field because of poor
guide stars). This would not include actually operating the
positioner; that task would remain in the capable hands of
trained
personnel.
- Do quick look reductions of data as it appears, checking for overall
quality, and in particular insuring that the spectra fullfill
program goals. E.g., are objects detected at all (coords ok?),
are
objects underexposed or overexposed, etc.
- Help make decisions regarding the queue during times of marginal
weather or seeing, choosing targets from a the nightly list.
To
aid the on-site astronomers, each group with Hectospec time will
be expected to supply a brief summary of their data and calibration
requirements.
In order to facilitate scheduling of observing blocks, we will ask
people
with allocations for their availability to go to the
telescope. Any team
member is eligible.
II. Observing Protocols
In order to facilitate the TDC
Hectospec reduction pipeline script
as well as to produce
consistent, high quality data, we require
the following observing
protocols be followed when
observing with Hectospec.
1. Take ten biases at
the beginning of the evening.
2. Take darks every
night, at least five at 15 minutes each.
These are usually taken at the end of the night.
3. Take ten domeflats
at the beginning of the evening. The
exposure time for domeflats is 2 seconds.
4. Take comps
(henears) right after domeflats if possible.
There should be five 300 second exposures, or, if that is
not
possible, a minimum of three 300 second exposures.
The
comp exposures must be contiguous as they will be
combined to make one file. It is important to take comps
and
domeflats close in time because the domeflat extraction
profile, without any shift, is used for the comp extraction
and
there can be a small drift in the fiber positions over
the
night.
5. Take sky flats
every night when possible. These can be started
right after sunset and/or up to sunrise.
We
recommend a series of up to ten
sky
flats in increasing exposure times after sunset or
in
decreasing exposure times before sunrise. So, for
example, you might have a series of sky flats with
exposure times of: 2, 5, 5, 8, 12, 20, 25, 30. The
increase or decrease in exposure time is made because
of
the rapidly darkening or brightening sky. Sky flats
are
used for the fiber throughput correction in the
reduction pipeline.
6. At least three
contiguous observations should be made of
your
science fields. We need a minimum of three to do a
good
job of removing cosmic rays in the combine
program.
7. A minimum of thirty
fibers must be used for skys, more
if
you can do so without losing a significant amount of
data/observing time. Set the parameter for the number
of
skys in the fitfibs program to 30 or higher when
you
are preparing your observing fields. Skys are
subtracted separately for each chip so a total of
thirty or more skys increases the liklihood of having
an
adequate number on each chip.
8. If you are
observing stars, don't mix faint and bright
stars in the same field . The range of magnitudes should
not
be greater than four magnitudes. You do not want to
saturate the bright stars to get the faint ones. Make more
than
one configuration if needed.
nc, st & dgf