Previous Article
Next Article
Table of Contents

Queue Scheduling: A Progress Report (1Sep93)
(from KPNO, NOAO Newsletter No. 35, 1 September 1993)

Two new queue scheduling programs are underway at Kitt Peak National
Observatory. The first is our Summer Queue Pilot Program, an experiment
designed to give us some experience with all aspects of running a queue
scheduled observing program.  Regular readers may recall that we
accepted e-mail applications for this program up until 30 April.
Proposals were accepted for three telescope/ instrument combinations:
the 2.1-m with GoldCam, the 0.9-m with T2KA direct, and the Coud Feed
with Camera 5/Grating A/T2KB. Observing programs on each of these
telescopes were to be carried out for a period of approximately two
weeks between 15 July and 15 September.

We received 49 proposals for this program. Nine of these were for the
Coude Feed, thirteen were for the 2.1-m, and the remaining 26 were for
the 0.9-m. In our announcement of this opportunity, we tried to
distinguish this program from service observing.  While it is true that
staff observers will be making the observations in place of the
proposers, the order that they will be made depends on priority
according to judgment of the scientific justification. The advantage of
this approach is that the more highly ranked proposals have a better
chance of obtaining their observations successfully than if they were
just given some specific nights on the telescope. In addition,
calibration exposures and setup time can be shared among programs,
leading to greater overall absolute efficiency. The distribution of
sizes of program proposed is shown in the figure below. To compute
these sizes we have added appropriate overhead for setup on each new
object and readout time for each exposure. It can be seen that the
range extends from very small programs which would be difficult to
schedule in the regular mode to the more usual programs which represent
several nights of observing. These requests represent oversubscription
ratios of 1.0, 2.1, and 1.5 for the Coud Feed, 0.9-m, and 2.1-m
respectively. These ratios assume that all the time will be clear and
usable. If, as past experience would indicate, only one-third of the
time is usable, these ratios should be multiplied by 3.

The second program represents an attempt to begin queue scheduling some
of the proposals given time by the usual route. Five proposals for the
0.9-m telescope and CCD imaging which were highly ranked by the TAC
were selected for the fall queue observations. The total number of
nights recommended by the TAC for these programs was increased by 30%.
This number of nights has been scheduled on the 0.9-m during October
and November for the queue execution of these observing projects. The
PIs involved, whom we thank for allowing us to schedule in this way,
are Paul Hodge, Dan Maoz, Eugene de Geus, Annette Ferguson, and Paul
Francis. We look forward to getting feedback from these proposers and
from those who get data from the summer experiment so that we can
understand whether queue scheduling helps us better meet the needs of
the community.

                        [figure not included}

Todd Boroson, David De Young,
Taft Armandroff, Caty Pilachowski

Previous Article
Next Article
Table of Contents