Figure 1. The ISO orbit, showing the observing constraints imposed by the Sun, the Earth, the Moon and Jupiter (not to scale)
ISO will be operated in a highly-elliptical orbit with a perigee height of 1000 km, an apogee height of 70 600 km and a period of just under 24 hours. The lower parts of this orbit lie within the Earth's Van Allen belts, regions of trapped electrons and protons. When the satellite is inside these belts, the majority of ISO's detectors are scientifically unusable due to effects caused by radiation impacts. ISO will spend roughly 16 hours per day outside the radiation belts and during this time all its detectors may be operated at maximum sensitivity.
There is no data storage available on board ISO and only very limited storage for telecommands. Therefore, for scientific use, ISO needs two ground stations to provide continuous contact. One of these stations, co-located with the ISO Control Centre at Villafranca, Spain, is provided by ESA. The second one, at Goldstone, USA, acts as a relay between ISO and its Control Centre for telemetry and telecommands for part of each day.
For thermal and power reasons for the spacecraft and also to prevent celestial stray light reaching the instruments and degrading their performance, there are a number of very stringent constraints on the allowed pointing directions for ISO. These constraints are shown in Figure 1. Note that Jupiter is normally kept more than 7 degrees away from the optical axis but can be pointed at, if it is the target of the observation. These constraints mean that, at any moment, ISO only has a limited area of sky available for observations (averaging approximately 12% and ranging from about 8% to 20%).
Figure 1. The ISO orbit, showing the observing constraints imposed by the Sun, the Earth, the Moon and Jupiter (not to scale)
The 'available' part of the sky changes with time as ISO's orbit precesses and as the Earth plus ISO moves around the Sun. However, during the 18-month in-orbit lifetime, approximately 12-15% of the sky (depending on the exact launch date) will not emerge from behind the various constraints, particularly that due to the Earth. Thus, this area will never be observable by ISO. The precise location of this hole depends on the exact date and time of launch. For other reasons, the launch date is constrained to two different periods of the year. For a launch in the spring period, the hole will be in the region of the galactic centre and for an autumn launch, it will be in the Orion region.
These sky accessibility constraints and the need to keep open the possibility of launching in either season meant that, during the satellite development, it has effectively been necessary to plan two different observing programmes in parallel. This has been achieved by having two different lists of astronomical targets for each research programme.
In keeping with ISO's role as an observatory, the majority of its observing time will be distributed to the general astronomical community in ESA Member States, the USA and Japan via two competitive 'Calls for Observing Proposals', one pre-launch and the other post-launch. Based on pre- launch estimates of efficiency, a total of about 3600 hours of 'Open Time' will be available during the mission. Since it is believed that ISO will make new and unexpected discoveries and, in many cases, will be the only facility capable of follow-on investigations of these discoveries for many years, an additional 250 hours of observing time, called 'Discretionary Time', will be made available for 'observations that could not have been foreseen at the time of proposal selection'. This mechanism introduces a suitable degree of flexibility into the ISO observing programme.
The remaining time, amounting to about 2800 hours of 'Guaranteed Time', is reserved for those scientists involved in building and operating the facility. They are: the four Principal Investigator teams that built ISO's scientific instruments; the group of five Mission Scientists who advise ESA on ISO; ISAS and NASA who, via collaborative agreements with ESA, provide the second ground station and its associated resources; and the astronomers in the ISO Science Operations Team who are responsible for the scientific operations of ISO and who support the community in using ISO.
Following ISO's launch, there will be a period of about three weeks during which the satellite will be commissioned. Then, there will be a performance verification phase of nearly eight weeks during which the core instrument calibrations will be carried out and the in-orbit performance of ISO and its instruments established. Thereafter, ISO will enter into its routine operations phase, which will last at least 16 months. Figure 2 shows the average daily distribution of observing time between all parties in this routine phase.
Figure 2. The average daily distribution (in minutes) of ISO time during the routine phase. ISO spends about 16 out of each 24 hours outside the radiation belts; this time is available for science observations
Observing programmes for both the Open and Guaranteed Time were prepared in two phases. During Phase 1, potential observers worked at their own institutes to establish the scientific rationale for their programmes and to specify the desired observations, including instrument and mode, filter or wavelength, and total time required. During Phase 2, the successful proposers visited one of two support centres to enter full details of the observations into the Science Operations Centre's Mission Database.
Preparation of the Guaranteed Time programme started more than five years ago and has involved at least 150 astronomers. In order to take maximimum scientific advantage of ISO's limited lifetime, the holders of the guaranteed time have worked together to achieve the maximum possible degree of coherence among the 139 individual research projects. As part of the coordination process, two major workshops, each attended by over 100 scientists, were also organised. The ISO Science Team oversaw the final preparations and worked continuously to ensure the required coordination. The complete programme was prioritised into three bands and presented to the Observing Time Allocation Committee (OTAC), composed of external scientists, for their endorsement in February 1994. The whole guaranteed time programme was then published, not only to inform the community of which observations were reserved for the guaranteed-time holders but also to serve as a set of 'worked examples' thus permitting the general astronomy community to submit complementary and feasible observing proposals for the Open Time available.
The Open Time programme was then defined. The pre-launch Call for Observing Proposals was issued in April 1994, and was distributed to close to 2500 astronomers worldwide. In addition to the guaranteed-time summary, the 10-volume Call contained Observer's Manuals for the instruments and the satellite. The Science Operations Centre (SOC) also made 'hot-off-the-press' information available via electronic computer networks (ftp). Proposers had to use a special software tool, the ISO Remote Proposal Submission System, on their own computers to prepare observing proposals. They then submitted their pro-posals, via the network, to the SOC by August 1994 for processing and review.
A total of 1000 Phase-1 proposals was received. Approximately one third of these proposals were for stellar/circumstellar topics, another third for extragalactic studies and one quarter addressed the interstellar medium. The rest were split roughly equally between solar-system and cosmological subjects. In total, about four times more time was requested than will be available. The SOC assessed the technical aspects of each proposal, often through interaction with the proposer, and sent the assessments to the OTAC.
The OTAC was organised into seven panels, each one addressing a different scientific area. The OTAC ranked the proposals by scientific merit, often considering the details of individual observations. It was recommended to enter observations totalling 3000 hours of time (leaving some to be allocated by the post-launch 'Supplemental Call') into the Mission Database in two priority classes. Additionally, another lower priority 3000 hours of observations were identified for entry in order to help the Mission Planning system to maximise scientific usage of ISO.
In Phase 2 of the programming process, Guaranteed Time observers and the successful Open Time proposers were requested to enter their data into the Mission Database. They were invited to visit either the Proposal Data Entry Centre (PDEC) at ESTEC in Noordwijk, The Netherlands (mainly European and Japanese observers), or the equivalent centre at IPAC in Pasadena, USA (mainly US observers). Both centres provided visitors with a range of software support tools (astronomical and technical), and ISO experts were continuously on hand to provide assistance in optimising the observing programmes. Details of the observations, such as instrument and sub-instrument, source coordinates, filters, wavelength ranges, expected fluxes, and desired signal-to-noise ratio, were entered into the Mission Database using the Proposal Generation Aids (PGA) software. PGA checked the input parameters for validity at the time of entry and then calculated how much time would be needed for each observation. Another system, Proposal Handling, made further checks to ensure that only valid and feasible observations were passed on to the Mission Planning software. The interaction with observers proved to be extremely useful in helping the SOC staff to gain a better understanding of the various programmes.
PDEC and IPAC were open from December 1994 until the end of July 1995. During that time, full details of over 32 000 observations arising from about 800 proposals were entered. At PDEC, there was a total of 540 visitors and for part of the time, the centre was operating at its maximum capacity of 30 visitors in parallel. Figure 3, Figure 4, and Figure 5 give some statistical information from the Database, when most of the observations had been entered.
Figure 3. ISO instrument usage by number of observations (on left) and by time (on right)
Figure 4. Distribution of number of observations against time (note the logarithmic scale)
Figure 5. Expansion of Figure 4 showing the distribution of number of observations against time for observations less than one hour. (Note that the scale is now linear)
After the end of the in-orbit performance verification phase, it may well be necessary to modify the observing programme in light of the achieved performance. Depending on the volume and complexity of the changes, either these will be done by the SOC staff (in close contact with the observer) or the PDEC will be re-opened so that observers may update their own programmes directly. Another possibility under consideration is remote log-in to PDEC by observers.
It is planned to issue a Supplemental Call for Observing Proposals some time in the second quarter of 1996; a procedure similar to that used for the pre-launch Call will be followed.
The standard observing mode for the spacecraft will be three-axis-stabilised pointing at a selected target for a period of up to 10 hours with an accuracy of a few seconds of arc. In addition to this single pointing, the satellite has an on-board capability to point sequentially at a user-specified rectangular grid of positions on the sky. This is known as raster pointing and enables mapping of a larger area of the sky.
There are several ways of scheduling observations using the scientific mission planning system, MPP1 (Mission Planning Phase 1). The system treats each observation as a separate, schedulable element. If there is a scientific reason why two or more observations should be carried out contiguously in time, the observer may 'concatenate' them and force MPP1 to treat then as one observation. This facility has been used quite extensively, with about 50% of the observations being in concatenated chains.
It is also possible to request that an observation is carried out at a specific 'fixed' time. This is to permit, for example, observations of variable or periodic phenomena at specific times or co-ordinated observations with other facilities. However, it may not always be possible to carry out such observations due to the sky-coverage constraints. Also, such observations restrict the efficiency of the scheduling process; thus, the use of this facility will be restricted to absolutely essential cases. Currently, about 2000 observations are flagged as fixed time.
Another method of scheduling is using linked observations. A linked observation is composed of an initial exploratory or test observation that determines whether, and in what form, a more in-depth science observation (main) should subsequently be performed. This facility exists for cases where the source properties are so poorly known that there is a significant risk of wasting ISO time with the wrong instrument settings. Because of the extra demands it places on ground segment resources, a strong scientific justification is necessary for all linked observations.
Each of the four instruments on the spacecraft has many possible operating modes. To simplify the user interface, so-called Astronomical Observation Templates (AOT) have been defined for the astronomically-useful modes of the instrument. Each AOT is designed to carry out a specific type of astronomical observation. The LWS instrument has four such observing modes, SWS four, ISOCAM four, and ISOPHOT eleven. The AOT acts as an interface between the instrument and the observer. It allows users to specify their observation in terms familiar to them; the software (the AOT logic or AOTL) then generates directly and automatically the necessary spacecraft and instrument commands to be sent to the satellite.
The duration of an observation is calculated by determining exactly how the instrument will execute that set of measurements. To achieve this, the AOTL software needs not only the parameters filled in by the observer but also various tables of instrument and calibration parameters and logic on how to command the instrument. The use of tables simplifies the process of updating the system as knowledge of the calibration of an instrument, for example, changes. The Proposal Generation Aids software, used by the observer, calculates observing times using the AOTL.
The limited lifetime of ISO, the severe sky-coverage constraints, the complexity of the scientific instruments, together with the necessity to make many short observations, dictate that a pre-scheduling operation must be carried out in order to maximise the time spent acquiring useful astronomical data. The staff at the SOC will schedule the different observations within a proposal so as to optimise the overall usage of the facility.
Once in orbit, ISO will be operated from ESA's Villafranca Satellite Tracking Station, which is also the European home of the IUE observatory. Two ISO teams are co-located there: one is responsible for the operations of the spacecraft and the other, the Science Operations Team, is responsible for all aspects of the scientific operations ranging from the issue of the 'Calls for Observing Proposals', through the scheduling and use of the scientific instruments, to the pipeline data processing in which the data products for distribution to the observers are generated. A simplified overview of the SOC's activities is given in Figure 6.
Figure 6. Simplified overview of the SOC's activities
The main purpose of the scientific mission planning system MPP1 is to construct, for each ISO orbit, the sequence of astronomical observations that are to be carried out. It has to make a trade-off between maximising the overall observational efficiency and performing those observations with the highest scientific priority. The main inputs to MPP1 are information on sky visibility and the observations, including priorities, contained in the Mission Database. The efficiency of the schedules is increased when MPP1 works from a database containing more observations than can be made in the available time. Its main outputs are (a) a file, named the Planned Observations File (POF), that contains the time-ordered list of scheduled observations, with an exact time at which every group of commands must be sent to the instruments and/or the spacecraft, and (b) a second file, named the Instrument Command Sequence File (ICSF), that gives the details of the groups of instrument commands called up in the POF. These two output files are passed on to the Spacecraft Control Centre at least three days in advance for further processing, leading to the production of a Central Command Schedule containing all necessary spacecraft and instrument commands.
The observations will be carried out under the supervision of SOC staff in a 'service observing' mode. Observers are not expected to be present for their observations. The commands in the Central Command Schedule will be sent automatically to the satellite at the specified times. Overall safety of the spacecraft and instruments is the responsibility of the Spacecraft Control Centre. The SOC will monitor some aspects of the instrument performance and scientific data in real time.
The calibration and processing of ISO data will be a major task. ISO is a complex instrument with over 20 operating modes and a data volume of about 1 Gb per day. Infrared detectors with such a high sensitivity are intrinsically not well behaved, showing various non-linearities and memory effects, and having time-dependent calibrations. In addition, operating conditions in space are very different to those in the laboratory, for example, the effects of impacts from cosmic rays must be considered. The knowledge of how to calibrate and reduce the data will increase during the mission and the associated software will have to evolve strongly from the launch system. Changes in calibration impact all areas of science operations from the planning of the observations through the real-time data assessment to the off-line pipeline processing. Thus, each of these systems has been designed to allow changes to be made as easily as possible.
The off-line data processing can be thought of as a 'pipeline processing' system plus an 'interactive analysis' facility. However, the two are not separate systems and have many common features. The interactive analysis system will be mainly used for characterisation and calibration analysis of the instruments; and for working out how to upgrade the pipeline processing. The pipeline processing will produce a standard set of products from each observation. The observers are responsible for the scientific analysis and interpretation of their data. Thus, the goal of the pipeline processing is to provide observers with data in a form with which they can work without an unnecessarily deep knowledge of the specifics of the ISO instruments. Per observation, all observers will receive:
These data products will be supplied on CD-ROM in FITS format, and it is planned to dispatch them to the observers within one to two weeks of the observation.
All products distributed will also be stored in the ISO archive. The Principal Investigator for an observation has exclusive rights to the data for one year, starting from the time at which an adequate calibration is available. At the end of this proprietary period, the data enters the public domain and can be used for archive research by the scientific community.
Since it is expected that the understanding of the behaviour and performance of the instruments will improve as the mission progresses and the associated data reduction and processing will evolve greatly, the data archive will be somewhat inhomogeneous by the end of ISO's in-orbit operations. During a post-operations phase of several years, the entire data set will have to be systematically reprocessed using the best available knowledge of instrument calibrations and data reduction techniques.
ISO will be operated in a service observing mode with each orbit's observations being planned in detail and finalised at least three days in advance. SOC staff will supervise both the conduct of the observations and the quality of the data products, which will be sent to the observer on CD-ROM in FITS format within one to two weeks of the observation being carried out. After a one-year proprietary period, the data enter the public domain for archive observations. After the end of the in-orbit operations, the entire data set of ISO will be re-processed in a homogeneous manner. The observing programmes of ISO, in both the Guaranteed and Open Time, have been established in a two-phase process. Full details of over 32 000 observations have been entered into ISO's Mission Database, have been care-fully checked for validity and feasibility, and are ready for execution.
The world-wide user community now awaits eagerly ISO's launch and the first data from its unique instrument complement.
Entering observation data into the Mission Database at the PDEC
ESA Bulletin Nr. 84.