Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.eso.org/~bleibund/papers/NMO95.ps Äàòà èçìåíåíèÿ: Sun Jan 7 21:58:59 2001 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 01:55:34 2012 Êîäèðîâêà:

The Astronomer -- Observer Interface
Bruno Leibundgut
European Southern Observatory, Karl­Schwarzschild­Strasse 2,
D--85748 Garching, Germany
Abstract. The minimal information exchange between the astronomers,
who are not present while their observations are obtained at the telescope,
and the observer performing said observations is laid out and detailed.
This will define the main interaction of astronomers and the observatory
in service mode and will be one of the critical areas for acceptance by the
astronomical community.
1. Introduction
Once the decision to transgress the current conventional modes of operating
ground­based optical and infrared telescopes to a new observing regime has
been made concepts to transform the ideas into workable schemes have to be
developed. This paper describes some of the concepts for a simple and clear
design of the information flow between astronomers and observers at ESO. As­
tronomers in the context of this paper are the persons interested in the scientific
content of the data while observers are the acting figures at the telescope.
Currently the responsibility of data procurement rests completely with the
astronomer who performs the observations autonomously, sometimes assisted
by technical experts. Removing the astronomer from the process of data ac­
quisition implies that all information which has been available to the expert
scientist in judging the scientific quality of the observations is delegated to an
expert observer with technical and operational knowledge of the astronomical
instrument. This splitting of the process demands profound and complete infor­
mation exchange and thus should lead to as clean and transparent an interface
as possible. A successful implementation fosters an increase in the observatory's
performance by making optimal use of all available expertise.
There are, of course, two interfaces to be considered: the input information
for the successful procurement of the scientific data and the delivery of the data
to the astronomer. The former constitutes a new layer of communication while
the latter is solved by providing much of the needed information in the data
file headers and associated logs which are available to the astronomer. Here,
we will concentrate on the information passage which defines the observational
procedures.
At ESO a dedicated group of astronomers has drafted a proposal for the
implementation of service observing as it was outlined in the science operations
plan for the VLT (Baade 1995, see also Baade these proceedings). The group
consisted of D. Baade, R. Fosbury, P. GrosbÜl, J. Spyromilio, J. Wampler, and
1

myself. The ideas outlined here present some of the results from the discussions
in this group (GrosbÜl & Leibundgut 1995). The technical implementation in
the framework of the general data flow of the VLT is described by GrosbÜl &
Peron (1995 and these proceedings).
The VLT is not the first observatory to embark on the experiment of service
observing. Successful examples are the operations of the Hubble Space Telescope
and many radio telescopes, as well as collaborative efforts among scientists in
the community at large and observatory astronomers. While all these examples
are very useful in providing guidelines, the special requirement of rapid reaction
to particular meteorological conditions adds a new quality to scheduling and
operations of telescopes like the VLT or Gemini. It was thus deemed necessary
to develop a new design for the service operations of the VLT.
2. Observation Blocks
It is our believe that the definition of all scientific observations should be created
by the applying astronomers. This implies that we can design procedures which
are understandable and manageable by people not exposed to them on a frequent
basis. At the same time the process has to be transparent in order to empower
the astronomers to make best use of it. As such it is important to provide the
users with a concept as familiar to them as possible.
We have defined several distinct ``qualities'' which are important for observa­
tions. They are the identification of the celestial target, the technical description
of the instrument and telescope operations during the observation, a description
of conditions and requirements which apply to a given observation (e.g. meteo,
time, accuracy), calibrations needed to achieve the proposed scientific result,
and general comments. Each of these ``qualities'' is described separately for ev­
ery observation and combined to form what was christened the ''Observation
Block''.
A restriction imposed on the observation blocks was to create a suitable unit
for scheduling observations at short notice (flexible scheduling). To have units
small enough for rapid reactions observation blocks define only observations of
a single celestial object (Chavan 1995). Another limitation we introduced was
to confine an observation block to a single instrument or instrument mode.
An important feature of observation blocks is their modularity. This offers
the possibility of self­referencing and multiple uses of its parts. We envisage the
possibility to call another observation block to, e.g., guarantee a sequence of ob­
servations. Individual parts of observation blocks can be reused for observations
of a different object or the same object with a different setup. Calibrations can
be specified for each individual or a set of observations and can be combined by
the scheduler to be run as often as necessary.
For illustration purposes we give an example of a possible implementation
of observation blocks in the following. The case of an observation of a field in
the Large Magellanic Cloud with the Infrared Spectrograph and Array Cam­
era (ISAAC) to search for heavily obscured Asymptotic Giant Branch stars is
presented.
2

Observation block of the LMC AGB search field 1:
target info: AGB­LMC1.loc
instrument operation template: ISAAC­IOT01, AGB­LMC1.iot01
calibration: ISAAC­ICT01, AGB­LMC1.iot01
conditions, requirements: AGB­LMC1.cond
comments: Field 1 of the search campaign
of AGB stars in the LMC
The observation block is simply a file with pointers to other files each of
which contains specific information required for the observation. The target file
holds the celestial position of the object as well as one of a suitable guide star
possibly with additional information. Instrument operation templates will be de­
veloped by ESO to ease the operation of regularly used instrumental modes (see
below). Calibration needs can be specified separately. The file AGB­LMC1.cond
contains a description of the demands on the observation while the comment
section conveys supplementary information. Additional categories not included
in the example could be specifications of object acquisition (possibly another
template) and pointers to other observation blocks which are associated with
this one (e.g. observation of standard stars). Since we are allowing for several
exposures within a single observation block (e.g. imaging in several filters for a
single target) several templates (or sequences) could be called.
Collecting all astronomical information in a central unit has the advantage
to access the relevant information quickly without foregoing modularity. For
the observatory the observation blocks are the natural place where to tie on
additional information, like the rating of the proposal, the proposal abstract,
and book keeping utilities.
ESO is planning to provide complete ``standard'' observation blocks for reg­
ular instrument calibrations, focus exposures, skyflats, and also certain standard
star observations.
We have tried to develop a well­structured format of the observation blocks
and their subunits in order to prevent misunderstandings or communication
problems. It is essential to have a clearly structured entity while also offering a
free format subunit (like the comments). The hope is to design a robust concept
which can be understood by the astronomer, the observer, the scheduler, and the
technical environment (telescope and instruments). The challenge is to create a
structure appealing to human apprehension and suited for machine processing.
3. Instrument Operation Templates
As a special example of one of the subunits of observation blocks we describe here
instrument operation templates. All instruments of the VLT will be operated
through sequences (Allaert 1995, these proceedings). To relieve the astronomer
from designing their own sequences ESO will offer some standard sequences for
which only a few parameters have to be specified. The work of the astronomer
consists in determining which instrumental parameters are needed once the ap­
propriate template has been chosen. A more detailed description of templates
can be found in the ISAAC manual (Spyromilio 1995).
3

Templates (and sequences) will allow the astronomer to completely control
the instrumental setup even though not present during the observation. The
templates represent predefined operations making the handling of observations
much more user friendly. By using well­defined templates the observations can
also be linked properly into online reductions.
In the example the templates are specified by their name (ISAAC­IOT01) and
a parameter file (AGB­LMC1.iot01) which defines all particulars of the ISAAC
setup and the mode of telescope operation during the observation (e.g. chop­
ping).
4. Summary
It has to be emphasized that the actual implementation of service observing
as a major operational mode of future telescopes can only succeed if sufficient
acceptance by the astronomical community can be gained. Astronomers should
retain control over their observations as much as possible even though they will
not assist the data acquisition in real time. By providing them with a structure
like the observation blocks and its subunits they can define observations with a
high degree of specification.
The fact that similar concepts have been developed independently by UKIRT
(Bridger, these proceedings) and Gemini (Walker, these proceedings) is encour­
aging. It might even offer the possibility of exchanging or translating observation
definitions into the ones of another observatory. The Hubble Space Telescope
has led the way for many of the concepts and can be used for guidance in some
aspects. To be able to swiftly react to changing meteorological conditions, how­
ever, asks for a more detailed description of the requirements of observations.
Nevertheless, the lessons learnt by STScI are very relevant for the development
of ground­based scheduling systems (Johnston & Miller 1994). Astronomers will
accept the radical changes much easier when they can recognize some of their
traditional ways in the system and feel in control.
Acknowledgments. I would like to thank all participants in the working
group for invigorating discussions which led to the concepts outlined here.
References
Allaert, E. 1995, Sequencer Manual, VLT­MAN­ESO­17220­0737
Baade, D. 1995, VLT Science Operations Plan, VLT­PLA­ESO­10000­0441
Chavan, A. M. 1995, Scheduling Tool, OSDH­SPE­ESO­00000­0005
GrosbÜl, P. & Leibundgut, B. 1995, Problem Statement of the VLT Online Data
Flow, VLT­SPE­ESO­10100­0749
GrosbÜl, P. & Peron, M. 1995, Analysis Document of the VLT Online Data
Flow, VLT­SPE­ESO­10100­0790
Johnston, M. D. & Miller, G. E. 1994, Intelligent Scheduling, eds. M. Zweben
& M. S. Fox (San Francisco: Morgan Kaufmann), 391
Spyromilio, J. 1995, ISAAC Manual, VLT­MAN­ESO­14100­0841
4