Introduction

This Thesis describes the research work aimed to obtain a description of the spectral/angular response functions of the detectors of the Gamma-Ray Burst Monitor (GRBM) experiment on-board BeppoSAX, the Italian-Dutch X-ray astronomy satellite, and their application mainly to broad band (1.5-700 keV) spectral analysis of cosmic Gamma Ray Bursts (GRBs).
In this introduction we synthesize the astrophysical context, motivation and the steps of the work.

Gamma-Ray Bursts: the challenge of high energy astrophysics

The cosmic gamma-ray bursts (GRB) are the most mysterious objects in high energy astrophysics. They are intense and short flashes of gamma radiation arriving from any direction in the sky at unpredictable times. Since their discovery in 1967 (announced in 1973) by USA military satellites, many dedicated experiments have flown on scientific space missions, increasing the number of detections (more than 2500 nowadays) but also the complexity of the scientific case.
GRBs are typically detected in the hard X / soft gamma band (i.e. from 20 to 2000 keV), but they have been observed also down to few keV and up to tenths of GeV. Their typical duration is of few tenths of seconds, but events as short as one tenth of second and as long as hundreds of seconds have been measured. GRBs are very strong sources, with peak intensity that can exceed of one order of magnitude that of the (huge) particle and electromagnetic background. The bursts morphology is very complex and makes it difficult to classify them on the shape of their light curves, which varies from single pulse events to multiple pulses through a wide variety of shapes and time scales. Also their spectral properties vary a lot from event to event, but with the general trend of spectrum hardening during each pulse rise and softening during the pulse decay.
Due to the complexity of their phenomenology and to the very poor localizing capabilities of hard X-ray and gamma-ray experiments, the nature and origin of GRBs has remained very obscure till about 1990, with no possibility to choose between galactic and cosmological models, with the former preferred because of the difficulty in explaining the huge luminosity ($\sim$10⁵¹ erg/s for a distance corresponding to redshift 1) required to produce the observed fluxes in the latter. Even local models, in which GRBs are produced in the Oort cloud around the solar system, could not be excluded.
A first significative step forward GRBs origin has been made in the nineties with the BATSE experiment on-board the NASA Compton Gamma-Ray Observatory. This experiment has detected more than 2000 events in the 25-2000 keV energy range, estimating for each one duration, peak flux, fluence and, most important, a position in the sky with an accuracy going from about one to about 10 degrees. The BATSE localizing capabilities are still very poor in comparison with the capabilities of telescopes at lower wavelengths, but have permitted to demonstrate two fundamental properties of the GRBs distribution in the sky: the isotropy of the arrival directions and the paucity of weak events with respect to an homogeneous spatial distribution. These results put in serious difficulty classic galactic models and are naturally explained by cosmological models. Nevertheless, they don't exclude that GRBs may origin from an extended galactic halo with a minimum radius of $\sim$125 kPc.

An unprecedented break-through in GRBs science has been produced in 1997 by the Italian/Dutch satellite for X-ray astronomy BeppoSAX results. This mission is not specifically designed for GRBs studies, but carries the right instrumentation for a new approach to GRBs observations of these sources. At the time of writing, BeppoSAX has been able to detect simultaneously with a gamma-ray (Gamma-Ray Burst Monitor, 40-700 keV) and an X-ray (Wide Field Cameras, 1.5-26 keV) detector 15 GRBs, localizing them with an unprecedented (few arcmin) accuracy (by exploiting the localizing capabilities of the X-ray detector) and perform fast (6-8 hours after the detection) follow-up with high sensitivity X-ray telescopes (0.2-10 keV, 1 arcmin accuracy) of 12 of these events (plus 2 localized by RXTE/ASM). For all the localized events, BeppoSAX has provided X and gamma-ray light curves and spectra, extending down to 1.5 keV the energy band on which the models for the emission mechanisms can be tested. For 13 events ,the first of which in February, 28 1997 (GRB970228) and the last in December, 26 1998(GRB981226), an X-ray afterglow, i.e. a power-law decay emission in the 0.2-10 keV energy band following the burst, has been detected. The precision and rapidity of these detections, and the wide dissemination of information has permitted other observatories to point the events, leading to the discovery for some of them of optical and radio counterparts, and the measurement of their distance through the optical redshift estimation. The most important consequence of these results is that the redshifts estimated from some of the optical counterparts range between 0.8 (GRB970508) and 3.4 (GRB971214). Thus, the cosmological origin of GRBs seems to be confirmed, consistently also with BATSE results on bursts distribution in the sky. In addition, the new phenomenology discovered, i.e. X-ray light curve and spectra of the burst, X, optical and radio afterglows, have enriched in an unprecedented way the observational picture, allowing the challenging of existing theoretical models and the growth of new ones.

GRB science with the BeppoSAX Gamma-Ray Burst Monitor

The BeppoSAX Gamma-Ray Burst Monitor (GRBM) consists of four cesium iodide (sodium activated) crystal scintillators, 1 cm thick, 1136 cm² geometric area each, acting primarily as lateral anticoincidence shields of the Phoswich Detection System (PDS), equipped with dedicated electronics and data storage/transmission for detection (trigger) and high resolution (down to $\sim$0.5 ms) timing of GRBs in the 40-700 keV energy range. In addition, the measured counts/s in the 40-700 and >100 keV bands and energy spectra in the 40-700 keV band accumulated over 128 s time intervals are continuously stored and transmitted to ground among the PDS housekeeping.

The BeppoSAX payload design is such that two GRBM detectors are co-aligned with the two Wide Field Cameras (WFC, two coded mask proportional counters, 1.5-26 keV energy band) on-board the same satellite. This particular configuration allows for detecting GRBs in the WFC and the GRBM simultaneously, giving the BeppoSAX mission unprecedented capabilities in the study of these objects. In particular, the BeppoSAX broad band GRB spectral study capabilities, can give us a lot of almost new information on the physical mechanisms underlying GRB emission, on the environment in which the event takes place and on the connection between the GRB proper and its afterglow. Indeed, despite the improvements due to the afterglows discovery, the origin and physics of these mysterious cosmic events is still unclear and debated, and the extension of spectral study to low energies (i.e. few keV) can put several constraints on theoretical models. Almost all GRB detectors on-board past and still active missions operate above about 20 keV, as BATSE. Very few spectral data down to 2 keV of GRB were available before BeppoSAX detections and most of them are from the GRB detector on-board the Japanese mission Ginga, whose low-energy data suffered the lack of the knowledge of the event direction. To exploit the GRBM + WFC spectral capabilities, a satisfactory knowledge of the response functions of the GRBM detectors in the field of view of the WFC is required. The location of the PDS at the center of the BeppoSAX payload is such that other scientific instruments, associated electronics, satellite structures, etc., partially obscure the fields of view of each shield. Thus, the reconstruction of the GRBM detectors response as a function of incident photon energy and direction is complex, and many efforts have been made do address it by means of extensive on-ground calibrations and the development of a model of the whole satellite for Monte Carlo simulations. The analysis of calibration data is the core of the work described in this Thesis, and, in addition to the reconstruction of GRBM capabilities of estimating source flux and spectral properties, it is the basis for event positioning and Monte Carlo model verification. The first simultaneous detection by GRBM and WFC of a GRB (GRB960720, about three months after the satellite launch) soon evidenziated the urgent need for an accurate knowledge of the GRBM response function, especially in the part of the field of view overlapped with the WFC. This work was then given the highest priority. The work on GRB localization is also improving, based partially on the work described here.

As expected from the payload configuration, calibrations show that the fields of view of the two detectors co-aligned with the WFC are rather clean and uniform, and that their response functions can be reconstructed by means of analytical models based on ground calibrations, with use of preliminary Monte Carlo simulations results as a tool to discriminate passive contributions and to extrapolate detectors efficiency at energies not covered by calibrations.

Thesis structure

This Thesis reports the processes that starting from on-ground calibrations data analysis have led to the reconstruction of the response functions of the GRBM detectors co-aligned with the WFC, the development of specific software and data analysis techniques needed to extrapolate source flux and spectral evolution from the 1 s ratemeters and average spectra from the 128 s energy spectra, in-flight calibration against Crab nebula measurements and BATSE results, and the application of response matrices, data reduction and analysis techniques and software to the study of spectral properties of GRBs and soft gamma repeaters (SGR).

In chapter 1 we give an overall description of the BeppoSAX scientific payload which is functional to the understanding of the mass distribution around each GRBM detector, and of the data reduction and analysis softwares specifically developed for this mission. A detailed description of the PDS/GRBM experiment is then given, with particular stress on the scientific data available from the GRBM.

The next two chapters describe the steps followed in the construction of the response functions for the two detectors co-aligned with the WFC (chapter 3), starting from a detailed description of on-ground calibrations and their results (chapter 2), going through the estimation of on-axis effective area as a function of energy, the interpolation with analytical functions of the effective area dependency on source direction, and finally the modeling of calibration line spectra and the estimation of spurious contributions. In chapter 3, we give also a detailed description of calibrations data formats and of the software that was utilized for data reduction and analysis, which is partly work of the PDS/GRBM hardware group (data reduction and quick analysis) and partly original work for this Thesis (specific on-ground calibration data analysis software, modeling and fitting of detectors response, response matrices handling and conversion in standard formats). A brief description of the Monte Carlo model is also given in this chapter, together with comparison between calibration and simulation results.

Chapter 4 describes the GRBM in-flight performances and data analysis techniques and software, with particular emphasis on in-orbit background, data reduction, background subtraction and spectral analysis techniques for the 1 s ratemeters in the GRBM and AC bands and the 240 channels / 128 s spectra. In addition, here are reported the methods developed to obtain measurements of Crab Nebula flux and spectrum in order to verify GRBM response functions, the cross-checks with BATSE results on GRBs simultaneously detected by the two experiments and the cross-calibration of WFC and GRBM using the Crab nebula spectrum.

The fifth chapter contains the description and results of the application of GRBM response function to the spectral analysis of GRBs simultaneously detected by one of the WFC. For most of these events, depending also on the statistical quality of the data, broad band average spectrum and spectral evolution analysis have been performed. As discussed above, these results are of high interest in GRB science and many of them have been published in scientific papers (some of them are included in the appendix).
The work and results on the GRBM detection of the soft gamma repeater SGR1900+14 is also reported, being a very interesting scientific case and an example of application of the GRBM response matrix outside the WFC field of view.

In conclusion, chapter 6 resumes the work done and briefly outlines the work in progress and the future perspectives.