No. 7 - September 14, 2001 Edited by Thierry Montmerle & Marc Türler

<< Previous issue Next issue >>

Editorial

Concerning INTEGRAL, maybe the most awaited news of this summer was the release of the open time program after the first announcement of opportunity (AO-1). Since August, the whole list of accepted proposals is available on the INTEGRAL site at ESTEC together with the 3rd ISOC Newsletter, which gives some statistics on the accepted program.

This summer, in July, the INTEGRAL spacecraft was moved from Alenia, Italy to ESTEC, the Netherlands and you will find the complete story of this journey on ESA's science WWW site. Since then, the ISGRI detector layer of the IBIS imager was also delivered and its ``first light'' is shown below.

This newsletter presents you also an important part of the INTEGRAL mission, which is not well known: The INTEGRAL Mass Model. The idea, here, is to understand and model all possible background contaminations due to the physical mass, the shape and the composition of the INTEGRAL spacecraft.

Finally, in the ISDC News section we start a series of articles describing different aspects of the data processing system we are developing at the ISDC. This time, we concentrate on the data format and the archive organization.

The official launch date of INTEGRAL is now fixed to October 17, 2002, just less than 400 days from now...

Data structures

Different types of FITS file extensions are called ``data structures''. In total, for all INTEGRAL instruments and all levels of the data processing, the ISDC defined about thousand different data structures each identified by a specific EXTNAME (``extension name'') keyword.

The grouping concept

For each science window, a few hundred data structures are grouped together in a compound object called a ``science window group'' containing all the data belonging to this science window. It consists typically of about 30 MByte of raw data, 50 MByte of prepared data, 20 MByte of corrected data and 20 MByte of high-level products. All these data are easily accessible, although, physically, they are located in different extensions of many different FITS files.

Since an observation consists of many pointings, an ``observation group'' is defined as being a collection of many science window groups. An observation group is also identified by a distinct keyword and contains all the data belonging to a given observation. This amounts to about 20 GByte for a typical observation of 300 ksec.

The archive structure

• scw contains the products of the first steps of the data processing on a per-science window basis.
• obs contains the higher-level results of the standard analysis and the quick-look analysis.
• aux contains auxiliary information that is not part of the INTEGRAL telemetry (e.g. observation plan).
• cat contains astronomical catalogues used for the data analysis.
• ic contains instrument characteristic files (e.g. coded-mask patterns, response matrices, calibration data, etc.).
• idx contains ISDC ``index files'' used for fast searching of selected data products.

The INTEGRAL archive will be on-line and directly accessible through FTP or through a WWW browse interface. The archive browsing tool, based on HEASARC's Browse facility, will be described in a forthcoming issue of this newsletter.

The ISGRI gamma camera Delivery Review took place on the CEA premises in Saclay on July 6, 2001. This device will be mounted on the IBIS gamma-ray space telescope. With its 2600 cm² sensitive area, ISGRI is the first gamma-ray camera in the world to be equipped with semiconductor detectors operating at ambient temperature. The camera comprizes more than 16 000 independent cadmium telluride (CdTe) detectors. ISGRI was delivered to the IBIS consortium on July 12 to be integrated on the IBIS telescope, itself to be delivered to ESA in November 2001.

The ``first light''

The figure on the left shows the gamma-ray shadow of the discobol statuette. In a very emphatic way, the discobol symbolizes the intense effort before launch and the light announces the dawn of a prolific era for gamma-ray astronomers. One may also see an allusion to the accretion discs powering compact sources.

Because it is thinner, the disc is less absorbing than the athlete's arm. The source is located 60 cm above the camera centre, causing the count rate to decrease significantly (from light orange to dark orange) from the centre to the edges farthest from the source. The dark points in the image correspond to about 100 faulty pixels, i.e., less than 0.6 % of the total number. This small fraction illustrates the quality of the manufacturing and justifies, a posteriori, the confidence in these new detectors (the specification was less than 5 % faulty pixels).

In previous gamma-ray missions the background expected in orbit has been estimated by scaling the results from earlier missions using mainly empirical techniques. There has always been a will to try to calculate the expected background scientifically using known physics. In the past this could not be realised, however recently with the advent of far more powerful computers and advanced particle physics software this has become a realistic possibility. The INTEGRAL Mass Model (TIMM) represents the first major attempt at this.

The figure on the left shows an exploded view of the SPI instrument. The complexity of TIMM arises due to the large number of different elements.

TIMM is effectively composed of three essential parts: the spacecraft geometry and materials, the input particle spectra and the physics simulator. This produces raw event by event output data which can then be analysed (filtered, blurred, etc.) to produce truly representational spacecraft data. The spacecraft geometry and material composition is supplied to the mass modelling team by the INTEGRAL instrument teams. The input particle spectra are taken from the literature. The physics simulator used, based on GEANT 3.2.1, and known as GGOD has been extended to allow for realistic simulation of the induced radioactive decays. GGOD allows all the energy depositions within the instruments and the spacecraft structure to be recorded accurately, and this information can then be used to produce data which will be a facsimile of the real spacecraft data.

The figure on the right shows the GGOD software flow diagram. GGOD uses the input radiation environment and the detailed geometry model to create a "virtual instrument".

Mass models have many uses, from instrument design and optimisation, to estimation of the in-orbit background (and of course whether this has features within it) and finally through to optimisation of the analysis software and the production of artefact free survey data. In the immediate future TIMM will be used to support the calibration of INTEGRAL. The mass model has also been used to investigate many interesting issues as they have arisen during the project, for example the effect of putting the cold pipe through the SPI side shield, Solar flare effects and the effect of off-axis sources to name but a few.

Two Examples

The figure on the left shows a 1 keV binned spectrum of the SPI background. The emission-line contamination due to the radioactive decays is clearly visible. The exact identification of the lines is in progress, based on TGRS data.

The figure on the right shows a projection of the background across the PICsIT detector plane. The axes shown are in the Mass Model Co-ordinate system, where SPI is at negative X. There is evidence of a gradual increase in background towards SPI, but there is also clear evidence of an increase in background at the detector module boundaries, caused by internal structural elements.

Ultimately the mass model also has a major use in data analysis. The mass model allows us to "Flat Field" the data, this in effect reduces the systematic errors and allows us to approach the statistical limits of the instrument sensitivity. The ability to closely reproduce the background structure (temporally, spectrally and spatially) improves the subtraction of artefacts from the real data and therefore improves the sensitivity obtainable. At Southampton we are currently involved in producing an all-sky survey using archival BATSE data and the mass modelling technique. Within GEANT we have constructed a reasonably detailed model of the entire CGRO spacecraft. The BATSE detectors themselves have been modelled in much higher detail than the rest of CGRO because they have a larger effect on the results. The impact of the temporal and spectral Flat Fielding processes are illustrated in the following two figures.

The figure on the left clearly demonstrates the ability to flat field the BATSE data. The large temporal variations in the raw data are greatly reduced (by ~90 %) once the model has been applied.

The figure on the right shows a BATSE all sky survey image, limited to >0.5 Crab sources, derived from 50 days of data. The flat fielding technique has improved the sensitivity (to approaching the statistical limits) and the image quality, and exciting new results are expected soon.

Unlike BATSE, INTEGRAL has a long period orbit and temporal variations should therefore be less dramatic, however spatial variances will dominate in this imaging system and more classic detection plane flat fielding will be necessary to improve sensitivity and image quality. In the post-launch future we can expect TIMM to actively be used to subtract background from observations, thus reducing systematics, allowing us to reach the statistical sensitivity of the instruments. TIMM also allows for the change in instrument performance to be evaluated over time, which will be important for archiving INTEGRAL data.

Additional information can be found at TIMM's home page.

Abstracts of recently accepted papers

Other papers picked up for you

Upcoming conferences around the astrophysics of INTEGRAL

• Contributed announcements
  • Workshop on X-ray spectroscopy of active galactic nuclei with Chandra and XMM-Newton
    3-6 December 2001, MPE, Garching, Germany
    The high throughput and excellent spectral capabilities of XMM-Newton and Chandra have already revealed many new and unexpected observational results. However, some of these have led to controversial discussions and require a better understanding; e.g.
    • the nature of the soft X-ray excess: Line or continuum emission?
    • new spectral features and their interpretations, e.g. the nature of the unresolved transition array,
    • the presence of warm absorber edges,
    • the relationship between X-ray and ultraviolet absorption,
    • grating spectroscopy of the broad and narrow emission lines,
    • relativistic disk lines at cosmological redshifts,
    • theoretical interpretations and modeling issues
    The aim of the workshop is to provide a platform for intense discussions of these topics. Future X-ray observations with XEUS and Constellation-X, in light of the recent XMM-Newton and Chandra findings, will also be included.
    Registration and abstract submission deadline: May 1st, 2001
   

• Other conferences picked-up for you
  • Multi-Frequency Observations of Variable Stars and AGNs: First Italian-Jordanian Conference in Astrophysics
    17-20 September 2001, Mafraq, Jordan
  • The Chemical Composition of Stars
    19-22 September 2001, Seattle, WA, USA
  • European Summer School on Experimental Nuclear Astrophysics
    1-6 October 2001, Santa Tecla, Sicily, Italy
  • The evolution of galaxies. II. Basic Building Blocks (EuroConference)
    16-21 October 2001, St Denis, la Réunion, France
  • Gamma Ray Burst and Afterglow Astronomy
    5-9 November 2001, Woods Hole, Massachusetts, USA
  • New Visions of the X-ray Universe in the XMM-Newton and Chandra era
    26-30 November 2001, ESTEC, Noordwijk, The Netherlands
  • Blazar Astrophysics with BeppoSAX and other Observatories
    10-11 December 2001, Frascati, Italy
  • Black Holes: Theory Confronts Reality, Three Years Later
    25-28 February 2002, Institute for Theoretical Physics, University of California, Santa Barbara, USA
  • 5th MicroQuasar Workshop
    26 May - 1st June 2002, Cargese (Corsica), France
  • IAU Symposium No. 212: A Massive Star Odyssey, from Main Sequence to Supernova
    24-28 June 2002, Lanzarote, Canary Islands, Spain
  • International Advanced School Leonardo da Vinci: Summer Course 2002: Mission Concept and Payload Design in X and Gamma-Ray Astronomy
    1-19 July 2002, CNR Conference Center - Bologna, Italy
  • The Seventh International Symposium on Nuclei in the Cosmos (NIC7)
    8-12 July 2002, Fuji-Yoshida, Yamanashi, Japan
  • CNO in the Universe
    10-14 September 2002, Saint-Luc, Valais, Switzerland