As a project that opened up a new era in space
astrophysics, IUE has of course had many 'firsts' during its
eighteen years of orbital operation. However, there have been
some truly unique accomplishments in the context of the IUE
project that must be recognised as real innovations in space
science and space engineering. The latest occurred shortly after
the decision to terminate the project in September this year.
The failure of IUE's no. 5 gyroscope on 6 March this year and
the subsequent reconfiguration of the spacecraft control mode
mean that it is the first three-axis-stabilised spacecraft to be
operated with only one functional gyroscope.
Other project highlights include the fact that:
After 18 years of annual extensions of its activities, the International Ultraviolet Explorer (IUE) Observatory has become a familiar hands-on facility to the worldwide astronomical community. The IUE project, run by NASA, ESA and PPARC (UK), has been the subject of an unprecedented response from the scientific community, in terms of both direct data acquisition and the use made of its data archive.
The worldwide scientific community was therefore very much surprised in December 1994 by the recommendation from a NASA Senior Review Panel, evaluating astrophysics projects on a 'science per US $' basis, that IUE orbital operations should be discontinued. Since European astronomers felt that such an ad-hoc termination would be damaging to ultraviolet astrophysics, the ESA IUE Project proposed a solution via a complete revision of the operations and science support procedures, which by then had been running successfully for eighteen years.
The resulting 'hybrid' operating mode has been designed to reduce overall project costs to a minimum in order to perform a 'Final IUE Science Programme'. Due to the timing of all this and the different funding cycles, negotiations, technical developments, user information distribution, funding needs evaluation etc., it became a race against the clock. Various pre- studies to evaluate improvements in efficiency had already been made at the ESA IUE Observatory in previous years. In January 1995, ESA's Astronomy Working Group expressed strong support for the suggested termination of orbital operations in an orderly manner at the end of 1997. However, the situation changed significantly as a result of the financial reductions imposed on ESA's Science Programme at the Ministerial Council Meeting in Toulouse in October 1995. The recommendation by the Space Science Advisory Committee (SSAC) to terminate IUE orbital operations not later than September 1996 was endorsed by the Science Programme Committee (SPC) in February 1996.
This is therefore an appropriate moment to highlight the most significant changes made in the context of the 'hybrid' IUE operations which had fully rejuvenated the long-running project and could serve as 'lessons learned' for future scientific mission operations.
Aspects of the operations that were redefined and streamlined in a significant way included:
These changes meant that all procedures had to be reviewed and activities revised in order to construct an economically and scientifically effective compromise. It is not feasible here to cover all the changes that have taken place, but we will focus on those in the areas of spacecraft control (TT&C), observatory functionality, as well as data processing and distribution (in parallel with the reprocessing of the more than 100 000 images already taken by IUE). The overall result is a modern up-to-date space observatory based on state-of-the-art hardware, software and communications capabilities.
The original IUE system comprised the spacecraft in geosynchronous orbit and two self-contained ground stations and 'observatories' equipped with subsystems for telecommand uplinking to and telemetry downlinking from the spacecraft. Ranging or tracking was only possible from the Wallops ground station as NASA was responsible for maintaining the spacecraft's orbit. Under this configuration, NASA exploited the satellite for 16 h/day and VILSPA for 8 h/day and the two observatories operated nearly independently. The Operations Control Center (OCC) located at Goddard Space Flight Center monitored the spacecraft on a 24-hour basis and served as a back-up to VILSPA for potential single- point failures (main computer system and VHF command antenna). The only communications link was a voice line for coordination between the two OCCs. In the 'hybrid' mode, the OCC at GSFC would no longer be manned on a 24 h basis and consequently the hot- backup support would no longer exist. Also, science operations and spacecraft control had to be conducted from VILSPA when the spacecraft was below VILSPA's horizon.
To cope with this new situation, it was decided to link the two observatories with two 64 kbit/s digital lines, so that the GSFC OCC, Observatory and TT&C interfaces to the Wallops Station could be operated remotely from VILSPA, and vice versa. This new configuration provides a full back-up ground station and observatory on both sides of the Atlantic, making local duplication superfluous. This new modus operandi is termed 'bent- pipe operations'. In coordination with ESOC, VILSPA designed the new configuration (Fig. 1) and built and installed the necessary equipment both at VILSPA and GSFC. ESOC also made the necessary provisions in Darmstadt and configured the VILSPA and GSFC communications nodes so that the IUE 'hybrid' operations could be started on 1 October 1995.
Figure 1. This diagram illustrates the 'hybrid' TT&C
configuration, showing the extended communications configuration
put in place to support the 'hybrid' IUE science operations. The
blue arrows show the original situation, and the blue boxes show
the VILSPA-developed hardware which allows complete mutual hot
back-up between the two control centres (VILSPA and GSFC)
The IUE communications in 'bent-pipe' mode are based on:
The core units for these 'bent-pipe' operations are:
The ISLU unit allows either side to remotely operate the Sigma 9/5 or T-85 computers at GSFC or VILSPA, respectively, so that TT&C operations can be performed using the ground station located on the other side of the Atlantic using only a 9.6 kbit/s bandwidth channel.
The TOCTM unit allows VILSPA to remotely operate the GSFC Telescope Operator Console and perform science operations remotely, i.e. to prepare and read IUE's on-board spectrograph and relay the images to Villafranca where they can be processed. Also, this unit can be used to bring raw telemetry to the station when there is a failure on the VILSPA front-end system or when the spacecraft controllers at VILSPA need to monitor manoeuvres performed by the space-craft when operating via Wallops.
With this infrastructure installed, up to 16 different modes of operation can be selected. The three most important are:
The installed system is fully bi-directional for the GSFC OCC and the VILSPA OCC Science Operations Centre. The system design is very flexible and efficient, as it has it has been based on the existing equipment in both OCCs, thereby minimising the number of new items to be manufactured or purchased off-the- shelf. The engineering was done by VILSPA's maintenance and operations engineers, which also kept the total financial outlay low.
In order to implement the 'hybrid' operations approach, the observatory organisation and working schemes (i.e. user interfacing, science observation and data delivery) were completely revised, from proposal submission, through data processing, to product delivery. The new schemes (Figs. 1-3) contain innovations in many aspects, the most far- reaching ones being the suppression of the time allocation in 8- hour shifts to a single Principal Investigator (PI) and his/her presence during the observations, and the electronic delivery of data. All of these changes were necessary to optimise the scientific output of the instrument and to minimise costs, without requiring an unreasonable effort from the users to accommodate the new scheme.
The application of new technologies in various areas has allowed a more automated and direct flow of information between the users and the observatory, both for preparing the observations and for retrieving the data (Fig. 2). To replace the standard mailing of large proposal packages, a system of electronic proposal submission and processing was implemented. Once processed, the proposals were forwarded to the members of the Joint IUE Allocation Committee (J-IUEAC) for scientific evaluation, and reviewed by the scientific staff of the observatory for technical feasibility. This technical review was also communicated to the J-IUEAC, so ensuring timely availability of all of the information needed for proposal evaluation.
Figure 2. Overall information flow from the proposal cycle to
data delivery in the 'hybrid' IUE configuration, relying on
modern electronic means wherever possible to streamline the data
flow
As soon as the approved observations were defined by the J- IUEAC, a master observation file was generated, including target coordinates, instrument setups and other scientific or coordination requirements which could affect the scheduling. This master observation file was then transferred to a Pre-observation Preparation System, where an automatic query to the HST Guide Star Catalogue and the Digitised Sky Survey provided a choice of guide/offset stars and finder fields for all targets. Contact Resident Astronomers, assigned to every programme, informed the PIs of any possible problems affecting their observations and the full details were clarified, taking into account the actual time allocation by the J-IUEAC. A final file containing all the targets to be observed, together with the information necessary to perform the observations (e.g. specific spacecraft command sequences), was loaded into the real-time operations software. This same information was also introduced into the scheduling system, which determined a first-order optimisation of the observation date according to: (a) the spacecraft constraints (mainly related to power availability and subsystem temperature restrictions); and (b) the scheduling requirements of the PIs, such as specific dates and/or intervals between observations, due either to the nature of the target (e.g. a binary system), or to the need for coordination with other spacecraft and/or ground observations (necessary for 36% of the programmes).
During the observation preparation, PIs were also requested to specify their preferences for data distribution. An associated data-distribution master file was then pre-pared for inclusion in the Data Processing and Distribution System, which ensured delivery of the output data in accordance with the PI's request (address, media, frequency ...).
When the observations are performed, the raw data are transferred electronically from the real-time computers to the image processing systems, where they are processed and archived. The processed data are automatically loaded onto magnetic disks for self-retrieval by the PI via electronic networks (they can also be sent by post, on request). The design of these systems has made all the information needed for the observations available electronically at the console of the Telescope Operations Computer; there, the sequence of commands to be uplinked to the spacecraft can be retrieved from a database, but it can also be modified or even interrupted in real time. This maintains the necessary flexibility in the science operations, so that decisions can be taken quickly based on prevailing circumstances. This scheme of working means that target-of- opportunity observations can still be easily accommodated, and any unexpected astronomical event can be observed within a few hours.
From the beginning of the IUE project, IUE data handling at VILSPA was driven by:
The processing software, known as IUESIPS (IUE Spectral Image Processing System), had been upgraded many times and at first shared the same hardware with the IUE real-time computer system (Sigma 9 and later T-85 at VILSPA). In the late '80s, IUESIPS was implemented under VAX/VMS within the ESO-MIDAS image processing system. Data transfer between the real-time system and the processing system - which required manual intervention for processing - was accomplished by copying the data from machine to machine on 9-track tapes.
The requirements for the 'hybrid' operations were completely different:
The service observing mode requires integrated scheduling. The observations are scheduled solely to optimise satellite operations, science requirements and observing efficiency, rather than by programme and observing shift. It was therefore necessary to re-design all data-handling procedures, implementing a highly streamlined processing pipeline all the way into the quality- control procedures and data-distribution mechanisms (Fig. 3).
Figure 3. Internal data reduction and delivery within the IUE
project, showing the major changes implemented to support the
'hybrid' operations. The schematic above the dashed yellow line
represents the situation as it was in the past (data flow
indicated by yellow arrows), and that below the line the
currently implemented system. The red flashes indicate automatic
data and information flow, without human intervention
The boundary conditions for this were:
This was quite a challenge, because the new systems had to be implemented while the old operations continued. In fact, the transition from the old 'classical' to the new 'hybrid' operations took place on 30 September 1995 at 23 h UT without interrupting any of the services that the ESA IUE Observatory provides to its users.
The data-handling system for the 'hybrid' operations consists of three main subsystems:
A special effort has been made to ensure that spectra delivered in these ways, from both the NEWSIPS and the IUESIPS software, can be handled by major astronomical image-processing systems. For example, the IUE Data Analysis Center (IUEDAC) at GSFC developed a new version under IDL to deal with high- and low-dispersion spectra, while the latest version of ESO/MIDAS (95NOV) includes an 'IUE context' developed at VILSPA.
This re-engineering of the data-handling subsystem started during the spring of 1995, and in late October we were able to deliver IUE data to the scientists of the 19th Episode and the astronomical user community as a whole, in Europe and the USA. In the five months since the start of 'hybrid' operations on 1 October 1995, more than 700 spectral files (19th Episode data) have been retrieved from the distribution node by more than 50 different scientists at external institutes.
During almost 20 years of operations, the IUE project has evolved in ways which could not have been anticipated at the time of the satellite's launch in 1978. As a consequence, it supplies support arrangements comparable to many other space projects of more recent vintage. In the 'hybrid' operations phase, IUE has become the only satellite capable of delivering processed science observations to the user community within 48 h, without restriction.
ESA Bulletin Nr. 87.