Co-Investigator - The Gamma-Ray Astronomy Team - NASA

(MPE) produced BGO detectors for the Integralmission and developed the Compton Telescope(COMPTEL) instrument on the Compton Gamma-Ray Observatory (CGRO).An important aspect of our proposal is the majorhardware contribution by MPE, allowing us to meetthe stringent cost cap. All of the detectors and thelow-voltage and high-voltage power supplies will beprovided by MPE at no cost to NASA. MSFC willprovide the DPU, software development, managementoversight, engineering support, and testing.Ground operations are simple and undemanding.Commanding will consist only of occasional instrumentadjustments. Spectra, time histories, andrefined locations will be generated for all triggeredbursts. All data products are delivered to the ScienceOperations Center (SOC) within 6 weeks of receipt,and a web-based burst catalog is updated at leastweekly.Educational and societal opportunitiesGamma-ray bursts, the most powerful explosions inthe universe, have great appeal to the public andtherefore provide an excellent tool for public outreach.We will collaborate with the LAT team toproduce an exciting and informative integrated EPOprogram on gamma-ray bursts.In summary, our Burst Monitor is ideally suited tothe purposes of GLAST for the following reasons:• Our instrument will greatly enhance thescientific capabilities of the mission forGRB’s.• We meet the nominal sensitivity performancewith a 5σ detection threshold of 0.35 photonscm –2 s –1 and an on-board trigger threshold of

1.0 Science Goals and Objectives1.1 Introduction1.1.1 Importance of Gamma-Ray BurstsCosmic GRB’s are intense flashes of gamma raysthat last from milliseconds to hundreds of secondsand have a great diversity of temporal morphologies.They come at random times, from randomdirections, briefly dominating the sky and thenfading without a trace in the gamma-ray energyband. The origin of these bursts is one ofastronomy’s greatest mysteries. With the discoverythat they arise from sources at cosmological distances,GRB’s are known to be the most powerfulexplosions in the universe, each releasing theequivalent energy of several solar masses, as gammarays, during their brief lifetime. Since the universeis transparent to gamma radiation, sources of thistype could potentially be very distant, and thus veryold, possibly describing conditions in the earlyuniverse. Currently, popular models for burst progenitorsinclude merging binary neutron stars,accretion-induced collapse of a single compactobject to a black hole, or magnetically poweredsupernovae (called ‘hypernovae’).One of the main missions of NASA’s Space ScienceStrategic Enterprise is to advance and communicatescientific knowledge and understanding of themysteries of universe. The GRB puzzle is one of afew challenges to astronomers that has also caughtthe public’s interest. This interest presents anopportunity for effective education and publicoutreach, an integral part of NASA’s research andmissions.1.1.2 Gamma-Ray Burst Observations1.1.2.1 Establishing the Distance ScaleGRB’s were discovered serendipitously and werequickly determined to be cosmic sources of extremelyenergetic radiation, unrelated to our localsolar neighborhood. The discovery of GRB’s camelate in the 1960’s, with the launch of several defense-relatedsatellites with gamma-ray detectioncapability, part of a program called Vela that monitoredRussian compliance with the recently signednuclear test ban treaty in space. At that time, theoreticalwork suggested that gamma rays should beobservable from supernovae. An archival search ofthe collected data turned up no correlation betweenoutbursts of gamma rays and observed supernovae;however, bursts of gamma rays, occurring at randomtimes, were observed. Analysis to determine timingdifferences between detections, from several separatesatellites in Earth orbit, quickly ruled out theSun as the source. Thus, a new cosmic phenomenonwas discovered. With time, the baseline betweenobserving satellites was extended to planetarydistances, around Venus and several missions toMars, by the Pioneer Venus Orbiter (PVO) and theRussian Venera probes. This first interplanetarynetwork (IPN) was able to firmly push back theclosest distance to the newly named gamma-raybursts to well beyond the solar neighborhood bydetermining the maximum allowed curvature,assuming a spherical wavefront, impinging on allspacecraft. More importantly, by triangulation, theIPN was able to narrow the possible locations forthe elusive sources of GRB’s to small, arcminsquare error boxes on the sky. When examined,sometimes years after the fact, no special type ofastrophysical object was revealed as common to alllocations.The lack of an observable counterpart was vexingfor several reasons. Of most importance was theintrinsic efficiency of gamma radiation emitted bystrongly magnetized galactic neutron stars, whichemerged as a popular model. This model madeseveral testable predictions. Although the detectedbursts appeared isotropic and homogenous, theproposed source population of neutron stars shouldmap out the galactic disk on the sky. Following thedevelopment of very sensitive instruments, it wasexpected that the newly detected dim bursts wouldbe anisotropic, indicating an absence of sourcesbeyond the galactic plane.This set the stage for BATSE on board CGRO.BATSE’s principal scientific goals were the localizationof large numbers of events on the sky andincreased sensitivity, by roughly a factor of 10, overany existing detector. Localization was done with aset of eight matched detectors, pointed in eight1

Figure 2.—Locations and fluences of 2,013 GRB’s observed with BATSE over 8.5 years.The locations are consistent with isotropy. In the context of a local galactic model, this can beexplained only if faint sources (violet and blue) dominate the intensity distribution. Instead, there is arelative deficiency of faint sources, implying that GRB’s are simultaneously isotropic andinhomogenous. The local distance scale is excluded and a cosmological distance scale is indicated.different directions, as defined by a regular octahedron.Each detector was constructed as a flat plateof scintillating material to maximize differences inresponse with different angles between the sourceand the detectors. In addition, the detectors werelarge in area to maximize their sensitivity to theweakest events. CGRO, launched in April 1991,introduced an entirely new era in GRB observations.After the first year of observations, twothings were very clear: 1) GRB’s were isotropic onthe sky, independent of their brightness, in starkcontrast with expectation, and 2) their relativenumbers decreased with intensity in a way thatindicated inhomogeneity, as was expected (seefigure 2). Both observations, taken together, did notcorrespond with known populations of galacticsources of any kind, strongly suggesting that thesources should be associated with distant galaxies,and that GRB’s were truly cosmological objectswith tremendous intrinsic brightness. Despiteseveral challenges, these observations are evenmore firmly established today, after 8 1/2 years ofin-orbit operations by BATSE.1.1.2.2 Temporal Properties of BurstsIn general, bursts consist of one or more episodes ofemission, each of which may be smooth or irregular,joined or distinct, all of differing intensities. That isto say, no two burst light curves are alike. Severalexamples of bursts, observed by BATSE, can beseen in figure 3. The time history of GRB 910522,in figure 3 begins with a small pulse followed byover 100 s of inactivity, before resuming with acomplex series of pulses that includes the peak ofemission. If the later complex were to have beenblocked from view by the Earth, the first pulsewould be quite acceptable as a separate, completeburst. Smooth, single-pulse bursts, such as GRB990206 in figure 3, may seem to be an easilyidentifiableseparate class until each is examinedclosely, revealing differences in duration, intensity,and residual fluctuations. Only the total duration has2

Figure 3.—Time histories of six GRB’s. The morphologies are very diverse.3

strong spectral evolution. While the pulses between45 and 50 s are present in all three energy bands, thepulse at 15 s is radically softer, being weak in the 21to 62 keV band and absent above 330 keV.Figure 4.—Hardness-ratio and durationcharacteristics of GRB’s. When GRB’s areplotted against hardness ratio and duration, twoclasses become evident: short/hard and long/soft.No other distinguishing characteristics of theclasses have been identified.been found to be useful in classifying bursts(Kouveliotou, et al. 1993). As can be seen in figure4, when burst duration is plotted against a measureof the burst hardness (the ratio of BATSE energychannels 3 and 2), the whole population separatesinto two distinct classes, short/hard and long/soft. Itshould be emphasized that the two classes are notdifferent in any other observable behavior, such astheir spatial or intensity distributions, which mightbe a clue to what underlying physical cause distinguishesthem.While some attempts have been made to deconvolvebursts into individual pulses, as yet no method hasbeen found that is totally model-independent(Norris, et al. 1996). Other time-domain techniqueshave been applied, with mixed results. Band (1997)has done cross-correlation of burst time historiesbetween several broad energy bands, demonstratingthat pulses at lower energies lag in time behind thesame pulse at higher energies. The width of pulsesgenerally diminishes with increasing energy, as canbe seen dramatically in figure 5. Not only do thepulses narrow at higher energies, implying spectralevolution during pulses, the spectral hardnessdiffers greatly from pulse to pulse, showing furtherBesides the general property of duration, bursts arecharacterized by several other quantities, presentedin the BATSE burst catalog series. Some of thesequantities, such as peak flux and fluence, are basedupon a nominal energy band, so the GLAST BurstMonitor (GBM) should have good sensitivity overthe same energy range of 50–300 keV as used forthe BATSE catalog. Of more interest to theoreticalwork is the extension of these quantities to bolometricmeasures, including energy bands below andabove the nominal 50–300 keV band to includeessentially all the burst energy. This would requirecontinuous energy coverage from a few keV up tothat of the LAT.1.1.2.3 Spectral Properties of BurstsAs soon as bursts were discovered, their energycharacteristics were analyzed by the best availablemethods. With small detectors, the results werespotty at best, deriving a rough temperature estimateof typically several hundred keV. With better instrumentation,it was soon realized that burst spectragenerally had a high-energy, non-thermal power lawcomponent. Recent analyses contributed to thegrowing consensus that most, if not all, burst spectracould be fit with a single four-parameter functionalform, determined empirically. The canonicalspectral form consists of two power law segmentsjoined smoothly at a characteristic break energy(E break). A representative broadband GRB spectrumis shown in figure 6, using the CGRO observationsof GRB 990123 (Briggs, et al. 1999a). Both panelsshow the deconvolved spectrum over three decadesin energy. The top panel shows the spectrum inphoton flux units, while the bottom panel shows thethe spectrum in F units. The F spectrum is E 2times the photon flux spectrum; a flat spectrum inF units has equal energy per decade. The advantageof the F presentation is that it depicts theenergetics: in the case of GRB 990123 it shows thatthe bulk of the energy is emitted between a fewhundred keV and a few MeV.4

Figure 5.—Time history of GRB 990510 in 3 energy ranges.The light curves are for 5.2 to 8.5 keV (top) as observed with the Wide Field Camera on BeppoSAX,and for 21 to 62 keV (middle) and >330 keV (bottom) as observed with BATSE Large Area Detectors.This illustrates how dramatically time profiles can differ over the energy range to be observed withthe GBM.5

Figure 6.—Spectrum of GRB 990123.The spectrum is for a 32 s interval for which data from all four CGRO instruments is available.Overall, the spectrum is well fit with a model in which a low-energy power-law smoothly transitions toa high-energy power-law. The lowest-energy point provides evidence of an x-ray excess. The υF υ plotshows that the few hundred keV to few MeV band dominates the energetics of this event and that thepeak energy of emission is about E break= 700 keV.This simple spectral shape is in sharp contrast withthe extremely varied temporal behavior of bursts,which remains unclassified, except for the durationbimodality. The temporal behavior of the spectrumof GRB 990123 is shown in figure 7. The top panelsshow light curves in six energy bands; the narrowingof pulses with increasing energy is apparent.The bottom panels show the evolution of the spectralparameters E breakand . The overall trend is forboth parameters to decrease, coexisting with thepattern of E breakincreasing for each pulse. Thecombination of a hard-to-soft trend and a hardnessintensitycorrelation is typical of GRB’s (Ford, et al.1995). In the case of GRB 990123, E breakreaches theunusually high value of 1,470 ± 110 keV at thepeak of the most intense pulse.6

Figure 7.—Time history of GRB 990123.Light curves in six selected energy ranges observed with the CGRO instruments and the evolution oftwo spectral parameters, the break energy E breakand the low-energy power-law slope α. E breakvariesover the range ~200 keV to ~1400 keV.7

fact that the α distribution peaks at –1, rather than at–2/3 constrains the more popular blast wave modelas well (Cen, 1999). Further tests of the blast wavemodel are quite sensitive to the relationship betweenthe two power-law indices (Preece, et al. 1999),requiring an accurate determination of , which isFigure 8.—Distribution of E break.The histogram shows the values measured for5,000 spectra from 156 GRB’s observed withBATSE. The distribution is an importantconstraint on the range of Lorentz factors ofGRB blastwaves because any intrinsiccharacteristic energy is Doppler shifted.Interestingly, E break, the single spectral parameterthat can theoretically indicate relative Dopplermotion between the observer and the source, ischaracterized by a log-normal distribution of surprisinglynarrow width, peaking at 250 keV (fig. 8).Likely causes of relative motion include cosmologicalredshift and bulk Lorentz motion of emittingparticles, as required by blast-wave models.Mallozzi, et al. (1995) have presented evidence forthe cosmological redshift, in that the average valuefor the E breakdistribution lies at progressively lowerenergies for bursts with lower peak intensities.The two remaining spectral form parameters, thelow-energy ( ) and high-energy ( ) power-lawindices, are broadly distributed around –1 and –2,respectively (Preece, et al. 1999) (fig. 9). While notsensitive to relative motion, there is considerablescience in each of these as well. The width of thedistribution severely constrains the applicability ofat least one popular burst emission model, basedupon synchrotron emission from shocked electrons(Tavani, 1996, and Rees and Meszaros, 1992). TheFigure 9.—Distributions of the low-energyspectral index and the high-energy index .Low values of pose difficulties for synchrotronmodels, while values of above –2 areunphysical unless a break occurs athigher energy.8

the most difficult to do with the current instrumentation.While the LAT will measure , only a broadbandburst monitor will be able to connect thisobservation with the other two spectral parameters.Those spectra with >–2 are especially interesting,in that the power per unit decade increases withincreasing energy, implying an unphysical infiniteenergy output. Where there were enough counts tobe statistically significant, the high-energy powerlaw component has been observed to extend throughthe bandpasses of all the CGRO instruments, includingEGRET (e.g. fig. 6). It is currently unknownat what typical energies this component cutsoff, yet arguments based on estimates of the efficiencyfor hard photon production, or theenergization of the particles that emit such photons,clearly predict some limiting energy for burstemission that the LAT should be able to determine.It is essential to know how this cut off, should it beobserved, behaves in relation to the evolution ofother spectral parameters, so that a unified pictureof the emission can be assembled.This picture should extend into the hard x-ray band,where a hint of extra spectral structure has beenobserved in 15 percent of all bursts. Figure 6 alsoshows an example of this, where the photon rate atthe lowest energy point is significantly differentthan the model rate. Here again, the correlationbetween the high and low regimes of energy shouldbe determined, with the LAT determining thepresence or absence of high-energy photons andthe GBM covering the lowest energies at a betterresolution and sensitivity than is possible withBATSE.1.1.2.4 Discovery of Burst Afterglowand CounterpartsAlthough bursts’ output typically peak in the50–300 keV energy range, localization usinggamma-ray data simply cannot produce the accuracyrequired for telescope-based observations ofthe source. In the end, it took roughly 30 yearsfrom the discovery of GRB’s to determine thedistance to a sufficient number of candidate GRBcounterparts, emitting at other wavelengths, for thecosmological distance scale for bursts to be provenand for counterparts to be discovered. The firstobservation of counterparts came quite late in theprocess, requiring pointed x-ray observations byDutch–Italian instruments on board the BeppoSAXorbiting x-ray observatory, launched in April 1996.Several bursts were observed to fade in the x-rayband accessible to the Wide Field Camera (WFC)on BeppoSAX at long enough time scales for theextremely sensitive Narrow Field Instruments (NFI)to be pointed at the source. Finally, a fading pointsource was found by optical telescopes with alocation consistent with that provided by the SAXinstruments. Imaged much later by very powerfultelescopes, such as Keck and Hubble, these firstsource identifications emerged as distant galaxies,after the point source had faded away. It should benoted that the time scales for fading in GRB aftergloware not the same as the light curves establishedfor typical type II supernovae, so any associationbetween these two would involve unusual examplesof both. In several cases, the underlying galaxieswere bright enough for optical line emission spectroscopy,which gave a typical red shift for bursthost galaxies of z=1.The fading remnants of bursts, when they areobserved, cool rapidly with the peak in emissionpassing through x-ray, optical, and down to radiowavelengths. This behavior is reminiscent of anadiabatically cooling shocked fireball, which isthe standard picture for the observed afterglows ofbursts. Shocked acceleration of electrons is alsoconsistent with the detection of very high-energyphotons at times much later than the beginning ofthe event, as the time scale for accelerating electronsgrows increasingly longer with energy. Thedetection of very high-energy photons emitted bythese electrons, is one of the science goals of theLAT. However, there should be a context into whichthese higher-energy observations are placed. Interestingly,SAX has been unable to find an afterglowto any burst from the short duration class, thus nocounterparts for these bursts has been discovered.EGRET also was unable to investigate the shortclass bursts, with a dead time per photon that islonger than many of the bursts’ durations, eventhough they are harder, on the average, than thelonger class. Since the high-energy behavior thatthe LAT will explore for these events is completelyunknown, the GBM is required for their proper9

classification in the classical burst energy regime of50–300 keV.1.1.3 Existing and Near-Future CapabilitiesGiven the difficult questions that still surroundGRB’s, such as gamma-ray production and energetics,it is important to look to the present andplanned missions that have burst detection capabilitiesfor answers. Currently, BATSE is very healthyand can last for a considerable number of years,given the present orbit of CGRO. BATSE willcontinue to help progress burst research by makingcontext observations coincident with other instruments,such as the WFC on BeppoSAX, and it maywell create the longest duration dataset of generalburst properties when it passes the more than14-year operational lifetime of the pioneer Venusorbiter (PVO), sometime after the launch ofGLAST. EGRET’s burst operations are severelyconstrained, due to its nearly exhausted supply ofgas. BeppoSAX can continue to observe bursts inthe WFC and track afterglow in the NFI well past itsdesign lifetime; however, the spacecraft has had anumber of setbacks due to failures of the stabilizinggyros, and is now operating on only one. Also, theability of BeppoSAX alone, to detect bursts in thetraditional gamma-ray band, is limited by thecharacteristics of the active shielding surroundingthe x-ray detectors—the data are limited to eitherpoor time resolution or poor energy resolution.Other instruments, such as the Burst Monitor onUlysses and a similar instrument on WIND, althoughvery small, are serving the important role ofmaintaining the IPN.To complement the currently operating suite ofdetectors in orbit, several new missions have beenplanned for various GRB studies. Each of these hasa particular emphasis; however, the focus of all willbe on counterpart identification. It is hoped thatafter several years of operation, several hundredGRB hosts will have been identified and theirdistances measured. First, the High-Energy TransientExplorer (HETE-II), a broadband instrumentprimarily intended to investigate the hard x-rayportion of the burst spectrum, 2–400 keV, with goodresolution, will be launched in early 2000. In addition,the soft x-ray camera will provide onboardlocalization to 10 arcmin accuracy for approximately25 bursts a year. HETE-II has the capabilityto transmit the location quickly to the ground, forindependent observations by other instruments andobservatories, and will be operational for up to2 years. Following HETE-II, SWIFT will providerapid localization of hundreds of GRB’s to within2.5 arcsec and broadband spectroscopy (10–150keV in the Burst Alert Telescope) of the afterglowof bursts. The spacecraft will slew to point coalignedx-ray and ultraviolet (UV) telescopes at thesource location in roughly 50 s. Neither HETE-IInor SWIFT will accomplish broadband gamma-rayspectroscopy of bursts, since neither will be able toobserve the high-energy power law portion of thetypical burst continuum or even the energy rangewhere the break occurs in many bursts. Without aBurst Monitor on the GLAST spacecraft, there willbe no U.S. mission after BATSE that will fill thegap above 400 keV up to 10 MeV, the lower thresholdof the LAT, yet this is the energy range thatmakes bursts uniquely gamma-ray events.1.1.4 Needs Unmet by Currently PlannedMissionsAlthough current and planned hard x-ray imagingmissions will measure the x-ray continuum for largenumbers of GRB’s, they will not perform broadbandgamma-ray spectroscopy of bursts, nor is it clearthat they will be operational in the GLAST timeframe. It will be important to combine the excitingdiscoveries that will come from future plannedmissions with what is already known from previousburst studies. The energy of the spectral break,E break, is the only fitted continuum parameter thatscales with the relative motion between the sourceand the observer, which must be the product of thecosmological red shift of the host and the bulkmotion of a relativistically expanding fireball.Moreover, E breakis the spectral component that hastraditionally been used to characterize a burst’shardness. The limited high-energy coverage ofmany of the next generation missions will preventthem from determining this important spectralparameter. Thus, they may not be able to place theirobservations of very well localized bursts in contextof the known hardness distribution, and will belimited in their ability to explore hardness correlationswith other burst properties.10

If a second spectral component at lower energies isdiscovered, as indications of an x-ray excess seemto imply, its characteristic energy, along with that ofthe spectral break observed at higher energies,should be investigated to determine their correlation.This is especially important for distinguishingbetween several theoretical models that provide amechanism for the observed x-ray excesses—themeasurement of the break energy will be essentialin order to do correlative studies. Occasionally, thespectral break energy is high enough that it ispoorly constrained by the BATSE data. A detectorcovering the 2–20 MeV range will be able to determinethe correct value. Even for cases where E breakiswell constrained, may not be, unless there existsthe capability to observe in a broad energy band,above roughly 1 MeV.Questions left unanswered by EGRET can finally beinvestigated by the LAT, with the GBM making theimportant connection with known burst behavior atlower energies. The whole question of actual energy-resolvedburst time history at EGRET energiesthat has been left open is: What changes can beobserved in the temporal behavior of bursts as theobserved energy band increases? For example, it isknown that pulses in bursts tend toward a narrowingin time at higher energies, as well as a shift in thepeak of pulse emission toward the beginning asshown in figure 7 (Fenimore, et al. 1995). With avery low deadtime for the LAT, compared withEGRET, burst pulses can be examined with goodcount statistics at much higher energies than waspossible with EGRET to see if the trend continues.Of course, this will not be possible without simultaneousobservation at lower energies by a dedicatedBurst Monitor. Long term behavior at the highestenergies is also in doubt. With EGRET observing aburst lasting longer than 90 minutes where nocoincident delayed photons were detected byBATSE, the issue of persistent emission at highenergies has been left open. Again, simultaneouslower energy context observations will determinethe independence or interdependence of higherenergybands. In some theories, involving shockacceleration of electrons, the expectation is that thehighest energy electrons will be those that havebeen accelerated the longest. If so, many bursts mayhave very little flux of high energies at the onset,and the high-energy flux will gradually build overtime.In the case where burst progenitors may be highlymagnetized neutron stars undergoing collision oraccretion-induced collapse into a black hole, themagnetic field plays an important role in the higherenergyemission. Baring and Harding (1997) haveshown that the spectrum above ~50 MeV shouldhave a break from photon-photon pair production.Detection of such a feature requires knowledge ofthe continuum spectrum at all energies, especiallyfor a broad region just below the onset of the feature,to establish a baseline spectral index forextrapolation.1.2 Large Area TelescopeBurst ObservationsThe primary mission of the GLAST LAT, withrespect to bursts, is exploration. The only instrumentthat it can be compared with is the EGRETon board CGRO, as this was the first high-energyinstrument that could observe bursts as they wereoccurring and with locations determined wellenough that the high-energy events could be classifiedas coming from the GRB. The results fromEGRET were limited by dead time of about 100 msper event, which is as long as entire pulses in somebursts. With better sensitivity than EGRET and amuch wider FOV, the LAT will probe the relativelyunknown aspects of GRB’s above 100 MeV, wherethe effects of high-energy particle acceleration,relativistic beaming, and intergalactic attenuationare most clearly observed. The LAT will detect farmore GRB’s than was possible with EGRET, perhapsas many as 200 per year, with an effective areaover six times greater, a response that does not fallrapidly above 500 MeV and a FOV that is morethan four times larger. The GLAST LAT will alsoprovide high-quality spectral and temporal measurements,and will be able to localize many GRBsources with a high precision. Roughly 100 burstsper year can be localized to better than 10 arcminand a few per year to better than 1 arcmin (Bonnellet al. 1997). Within several minutes of burst onset,the LAT may be able to relay burst locations to11

ground- and space-based observatories to search forafterglow emission.The spectral observations from EGRET just scratchthe surface of what may be a diverse phenomenon.With 45 photons greater than 30 MeV from fourbursts, the average spectral index obtained, 1.95 ±0.25, is consistent with the average of high-energyspectral indices obtained with BATSE, 2.2, for amuch larger number of bursts in the sample (156),and a correspondingly larger number of spectrafitted (5,500). This high-energy portion of thecontinuum can be determined with the extendedenergy range of the GLAST LAT, and a distributionof spectral indices can be built up, to be comparedwith the BATSE result and to increase the precisionof the GBM spectral fits. The temporal behavior ofthe high-energy continuum can be determined quitewell by the GLAST LAT for comparison with lowerenergies. For some of the spectra fitted to data fromBATSE events, the spectral indices were smallerthan 2, indicating that the peak energy of the powerdistribution per unit decade has not been reached inthe BATSE energy band, less than 2 MeV. In addition,the EGRET result is similarly low; at someenergy there must be a roll-off in the high-energyspectrum, where the power output of the burstpeaks. The energy of this break will help to determinecharacteristics of the source emission processand if it is due to absorption of the extragalacticbackground light, it may be an independent measurementof a large source redshift. The presence ofhigh-energy photons in bursts is a very importantindication of the energetics of the emitters. Ingeneral, the higher the energies that are observed,the higher the bulk Lorentz factors must be at thesource in order to avoid runaway pair-productioncascades that are inconsistent with the observedspectra. In at least one EGRET event (GRB 940217)high-energy emission continued for 5,000 s after theend of the event, as determined by BATSE. Thesingle photon, detected at roughly 18 GeV, resultedin an extremely accurate burst location, and theextended emission contained a considerable fractionof the total fluence of the burst. In addition, itprovided an upper limit to the distance at which theburst source may lie of z less than 2, owing to theopacity due to pair production on the infraredbackground of photons with energies exceding10 GeV. Observations of this kind can be done forvery long periods of time with GLAST, with itslarge FOV.1.3 Observations Needed to Supportthe Large Area TelescopeThere are important limitations to the effectivenessof the main GLAST instrument as a burst detectorin its current configuration. First, high-energymeasurements alone do not reveal how individualbursts fit into the full population. This problem ismost evident in terms of GRB energy spectra, wherethe most characteristic spectral feature, E break, occursaround a few hundred keV (Band, et al. 1993), wellbelow the currently envisioned GLAST maininstrument threshold. The spectrum above E breakistypically a power law with no indication of a break(Dingus, et al. 1997; Catelli, et al. 1997), yet thepower-law index, in many cases, appears to beflatter than –2 (Preece, et al. 1999) which, forphysical reasons, cannot continue to indefinitelyhigh energies without steepening. The high-energybreak may, in a few cases, be measured by the LATalone, but in many cases will be properly constrainedonly by jointly fitting wide-band spectra.For some subset of the bursts, this may be possiblewith instruments on other satellites, but optimumcoordination is obtained by including a BurstMonitor on GLAST that has appropriate sensitivityin the energy range from x rays to near the LATthreshold. An accurate comparison of the spectralindex measured in the GBM energy range withthe measurement in the LAT energy range is essentialto distinguish between intrinsic spectral breaksand those caused by intergalactic attenuation of thehigher-energy photons.A further concern is that, without an instrument thatcovers the conventional GRB energy range, the LATobservations cannot be placed in the context of thewhole GRB population. The database of GRBobservations, produced by BATSE, will remaindefinitive for the foreseeable future. It will bescientifically wasteful if the properties of GRB’s, asmeasured by the GLAST LAT, cannot be associatedwith the BATSE bursts. This is optimally done withan instrument that covers the same energy range asBATSE, especially if it has similar trigger characteristics.12

An interesting related question is whether thereexists a significant population of hard spectrumbursts that has been missed or poorly sampled byBATSE. Although several authors (Lloyd, et al.1999 and Piran, et al. 1996) have claimed evidencefor such a population, other studies have not confirmedthis (Harris, et al. 1997 and Brainerd, et al.1999). The GLAST LAT can definitively settle thisquestion, but only if sufficient simultaneous coverageis available in the BATSE energy range.A further significant concern for the GLAST LAT,as a burst detector, is the technical problems associatedwith triggering. Although the LAT can triggeron the total event rate, without knowledge of theirsky location, this is not optimal. The wide FOV ofthe LAT implies a background rate that cannot beneglected for burst triggering, and the best triggersensitivity will be obtained by binning the eventsaccording to their sky location. This can easily bedone during ground data processing, but is a significantconstraint for onboard triggering. A BurstMonitor with sufficient sensitivity can provideonboard triggers with a corresponding savings incomplexity of onboard LAT data processing, andhence a savings in cost.1.4 Burst Monitor Requirements1.4.1 Lower-Energy Context MeasurementsThe GBM is designed to provide near full-sky burstobservations, in an energy band that overlaps boththat of the LAT and the hard x-ray regime, unifyingthem for the first time. It provides an importantcontext in which the highest energy observations ofGRB’s, by the LAT, can be placed. Much of theburst science from GLAST will be entirely new, asthe expected number of bursts per year will be manytimes the total number of events observed byEGRET over its entire mission to date. Some of theexpected results cannot be determined withoutsimultaneous observation at wavelengths below thelower threshold of 10 MeV for the LAT, such as thenarrowing of pulses and the shift of their peak intime toward the beginning. The discovery of highenergyspectral breaks or rollovers, in the LATenergy range, for bursts with too much high-energypower (spectral index >–2) rests on the ability todetermine the spectral index with good accuracy.This may only be possible with an instrument thathas good spectral coverage below that of the LATwithout large energy gaps. Determining whichspectral hardness class an individual burst belongsto can only be possible with an instrument thatcovers the crucial energy range, 100–400 keV, thatbounds most of the fitted spectral breaks, as shownby BATSE data. Indeed, it has been shown, by ananalysis of Solar Maximum Mission (SMM) data,that there is no large class of high-break energybursts extending beyond the tail of the BATSEdistribution, so energy coverage below approximately500 keV is as crucial to determining thespectral break as higher coverage is to determiningthe high-energy spectral index. In addition, theenergy range 20–2,000 keV contains much of theenergy output of typical bursts, and it is in thisrange that burst fluxes and fluences are traditionallycalculated. Burst Monitor observations are neededto determine the near-bolometric peak fluxes andfluences in conjunction with the LAT and to allowcomparison with the general population of events.1.4.2 Provide Localization for Bursts Overa Wide Field of ViewFor roughly 100 bursts a year, the LAT will obtainburst locations better than 10 arcmin. There is noneed for a narrow-field detector to independentlyproduce good burst locations. Rather, what isrequired is a wide-field monitor that can localizebursts to a few degrees to permit repointing of thespacecraft for optimal observations by the LAT. Thebest design for a Burst Monitor will observe the halfof the sky surrounding the pointing direction of theLAT.1.4.3 Enhance Sensitivity of Main InstrumentFinally, there will be many bursts for which theLAT will register only a few high-energy photons.The lower-energy Burst Monitor will detect theseand can provide a source location that is accurate toseveral degrees. This location can be used to searchthe LAT data for coincident photons from thesource, greatly enhancing the sensitivity over havingno monitor at all. The reason for this is that theangular resolving power of the LAT is so great, overthe portion of the sky it observes, that there will bea tremendous number of possible angular resolutionelements to search through at all times for possible13

triggers. In the absence of any context instrument,it may be feasible to search for these coincidentphotons only in data archived on the ground, greatlyreducing the number of possible good source locationsfrom the LAT. In addition, the constellation ofother spacecraft, operating concurrently withGLAST, that could possibly provide timely GRBsource localization is completely unknown.1.5 Science Investigation PlanOur primary science investigation will be elucidatingthe relation between keV/MeV and GeV burstemission by time resolved spectroscopy using datafrom the Burst Monitor and the LAT. We willperform simultaneous spectral fits to the data of oneor more Burst Monitor NaI detectors, a BurstMonitor BGO detector, and the LAT. We willdetermine whether a single spectral model, such asthe Band GRB function (Band, et al. 1993), cansimultaneously explain the keV–GeV data. We willrelate the behavior in the classical GRB band, suchas the value of E peak, to the GeV observations of theLAT. We will correlate the evolution of the keV–MeV spectrum to LAT observations. The two basicproducts will be spectra for subintervals of the burstand light curves for energy bands. For the purposeof time-resolved wide-band spectroscopy we proposenonexclusive access to GRB data from burststhat trigger the Burst Monitor. We will produce aspectral catalog of time-resolved model fit parameters,plus fit residuals. Example analyses, simulatedresults and expected performance are detailedin section 2.5.1.We propose two other main science projects:1) Generation of GRB locations and, 2) publicationof GRB catalogs. These have been chosen for theirimportance and because development of the algorithmsand procedures requires the experience andknowledge of the instrument team.As a service, we will produce prompt GRB locationswhich will allow repointing of the spacecraftto initiate LAT observations, aid in detecting GRB’sin the LAT data, and make possible prompt groundbasedobservation and possibly observation by otherspacecraft. The locations will help identify GRBevents in the LAT data. Locations will be generatedusing the same technique used for the BATSEinstrument, by comparing rates of several NaIdetectors. A simplified algorithm will be implementedin the flight software to produce near realtimelocations, to facilitate space and ground-basedfollow-up observations. A more sophisticatedalgorithm will be implemented in ground softwareto calculate more accurate locations. Further implementationdetails and expected accuracy are givenin section 2.5.2.We will produce a catalog of bursts that will includeparameters such as fluence, peak flux, and duration.These parameters will be defined as closely aspossible to those in the BATSE catalog so thatbursts, observed by GLAST, can be related to thelarge sample of the BATSE catalogs. Scientists inHuntsville, using their experience of producing theBATSE catalogs, will implement most of the proceduresand software for producing values for thecatalog. To extend the detection threshold as lowas possible, scientists at MPE will develop anuntriggered burst search. This search will detectfainter events by using more sophisticated algorithmsthan will be implemented in the flight software.The ground-based burst search willincorporate a higher order or more physical background,which can be fit to a longer time period.Data from more than two NaI detectors and fromthe BGO detectors can be used to “trigger,” and fluxincreases can be searched for on many timescales,etc. Techniques like these have been successful infinding untriggered bursts in the BATSE data stream(Kommers, et al. 1997, and Stern, et al. 1999). Theextended detection threshold will yield additionalevents to search for in the LAT data (several scientistshave proposed a very hard burst class whichmight be faint in the few hundred keV band), extendthe fluence and flux range of detected bursts, andincrease the catalog size for burst population studies.We have proposed three investigations of GRB’s.Much of the language of the AO is oriented towardinvestigations based on a fixed number of discretesource observations. Our instrument is more suitablefor continuing investigations of transientGRB’s, so we are proposing key project multiyearefforts. These investigations have been chosen14

ecause of their importance and because theyrequire the detailed knowledge of the instrumentpossessed by the instrument team. While the developmentof the catalog and location projects will beessentially completed during the first year of phaseE, these projects clearly should be continued for theduration of the mission. The scope and difficulty ofthe time-resolved spectroscopy project make it bestsuited to a multiyear effort. We do not intend forthese projects to impede other researchers and planfull cooperation with the GLAST guest investigator(GI) program. Both the Huntsville and MPE teamshave good track records in supporting the CGRO GIprogram. We do not need exclusive access to thedata.1.6 Other Science CapabilitiesData from the Burst Monitor will permit otherimportant scientific investigations which will likelybe conducted by other scientists through the GIprogram. The extensive nature of the GBM dataincludes: 1) Continuous good temporal resolutiondata background time (BTIME), 2) high spectralresolution data background spectroscopy (BSPEC),and 3) trigger data at full temporal and spectralresolution of the detectors time triggered event(TTE). This GBM data will allow GI’s to proposeand conduct investigations that we haven’t conceived,extending the science return of the mission.GI analyses of GRB data acquired by the GBM andLAT will likely include temporal analyses such ascross-correlating profiles in different energy ranges,and temporal and spectral averaging of weakerevents to boost the signal-to-noise ratio. The promptlocations will enable ground based searches foremission in other wavelengths during the burstphase of GRB’s, as was successfully accomplishedfor GRB 990123 using a BATSE GCN location withσ=13˚ (Akerloff, et al. 1999). The locations mayalso be used by other spacecraft, to search forprompt emission and counterparts in other wavelengths(x-ray, ultraviolet, optical, infrared, etc.).As demonstrated by the BATSE experiment on theCGRO, numerous scientific investigations otherthan gamma-ray bursts can be undertaken with allskyhard x-ray and gamma-ray detectors such asthose of the GBM. Sources that are strong andvariable, relative to the background, are easilydistinguished as distant point sources. In particular,numerous solar flare investigations have made useof BATSE data because of its high efficiency forhard x-rays and near-continuous coverage. Studieshave included fast timing of hard x-ray flares andcorrelations with solar impulsive microwave emission.As with GRB’s, the location, intensity and spectralcharacteristics can be determined for short, intenseflares from soft gamma repeaters (SGR’s) andgalactic black hole systems. Pulsars are observedby folding their periodic signals. Many additionalsources can be observed by the Earth occultationmethod, which was pioneered by BATSE. Thismethod has proven to be extremely valuable forcontinuously monitoring over 70 known objects anddiscovering about a dozen new, bright (>100 mCrab)sources. While the smaller detector area of theGBM detectors will limit the GBM sensitivityto ~0.5 Crab for short-term flares or one-day occultationflux determinations, many investigations ofsources at this level can be made over the GLASTmission.Due to the limitation of data analysis resources inphase E, we do not propose to perform any of theabove data analysis under the GLAST GBM fundingprofile. All data will be properly archived inlow-level form and documented for analysis byothers. They can also be utilized in near real time,deposited into a repository, and/or given to approvedoutside investigators as specified by NASAHeadquarters or the GLAST Project Office.2.0 Science Implementation2.1 Instrument OverviewThe primary scientific goals of the Burst Monitorare to measure the spectrum of GRB’s below energiesaccessible to the LAT and to provide rapidapproximate burst locations over a wide FOV. Wehave selected a complement of detectors and a data15

system that meet or exceed these requirements, yetare simple and low risk.A top level block diagram of the Burst Monitor isshown in figure 10. There are 12 NaI scintillationdetectors and 2 BGO scintillation detectors. TheNaI detectors are 12.7-cm diameter by 1.27-cmthick, directly coupled to a PMT. These detectorsare oriented to provide good sky coverage andperform two functions: 1) Provide spectral coveragefrom about 5 keV–1 MeV, and 2) determine burstlocations using relative rates, similar to BATSE(Pendleton, et al. 1999). NaI is an ideal scintillationmaterial for this energy range combining low cost,Figure 10.—Top level Block Diagram.high efficiency, and adequate spectral resolution.The BGO detectors are 12.7-cm diameter by 12.7-cm thick. To provide better light collection, as wellas redundancy, the BGO crystals are directlycoupled to two PMT’s, on opposite sides, whoseoutputs are summed. The two BGO detectors areroughly omnidirectional and are positioned onopposite sides of the LAT providing full-sky coverage.They provide spectral coverage from 150 keV–30 MeV. The high density of BGO provides goodsensitivity over this difficult energy range. The HVof each PMT is separately controlled. In thebaseline design, a single high-voltage power supply(HVPS) box provides 16 separate outputs, whichTable 1.—A traceability matrix showing the GLAST Burst Monitor design characteristics asderived from the scientific goals and constraints.Goal or ConstraintLow-energy spectral measurementsField of view: >~3 steradiansBurst threshold: ~0.5 photons cm –2 s –1Mass: 50 kgPower: 50 W.Telemetry: 10 kbpsBurst MonitorSpectroscopic observations from ~5 keV to ~30 MeV8.6 steradians0.35 photons cm –2 s –1 for 5 σ0.57 photons cm –2 s –1 trigger threshold54.5 kg, with 20% contingency17.8 W. without contingency4 kbps normally, 9 kbps during bursts16

are routed to the individual detectors. Each detectorincorporates shaping circuitry and preamplificationof the PMT anode signal.The type, number, and size of the detectors werechosen to satisfy the scientific objectives:• The two detector types cover the entire energyrange from 5 keV–30 MeV, with good overlapbetween the NaI and BGO energy ranges, andbetween the BGO and LAT energy ranges.• The small physical size of all detectors allowsflexibility in placement.• The detector diameter is the same as the PMT toallow direct coupling, which results in goodlight collection, energy resolution, andsensitivity down to low-energy with minimalcomplexity and risk.• The thickness of the NaI detectors is optimumfor the energy range where bursts typically emitthe most energy and provides approximately acosine angular response, which is important fordetermining locations.• The number of NaI detectors provides acombination of good sky coverage, greatlyexceeding the nominal FOV in the AO, andgood sensitivity, closely matching the nominalthreshold.• The thickness of the BGO detectors and theirplacement provide approximately uniformresponse over the entire sky, despite blockage bythe LAT.A schematic diagram of the mounting of the detectorsto the spacecraft is shown in figure 11. Thedetectors will not block any part of the LAT FOVnor interfere with the solar panels. They easily fitbetween the LAT and the shroud envelope on twosides of the spacecraft. Figures 12 and 13 show topand side views. The mounting arrangement is quiteflexible and will be explored with the spacecraftcontractor during phase B. With our approach, FOVcan be traded for sensitivity simply by changing theorientation of the NaI detectors. For example, if it isdecided not to incorporate the capability for realtimerepointing of the spacecraft, we could matchour FOV to the LAT and improve our sensitivity bypositioning the NaI normals closer to the spacecraft+Z axis.Table 2.—Comparison of GBM and BATSE.BATSEGBMLarge Area DetectorsLow-Energy DetectorsMaterial NaI NaINumber 8 12Area 2,025 cm 2 126 cm 2Thickness 1.27 cm 1.27 cmEnergy range 25 keV to 1.8 MeV 5 keV to 1 MeV. Spectroscopy Detectors High-Energy DetectorsMaterial NaI BGONumber 8 2Area 126 cm 2 126 cm 2Thickness 7.62 cm 12.7 cmEnergy Range 30 keV to 10 MeV 150 keV to 30 MeVTotal Mass 850 kg ~50 kgTrigger Threshold ~0.2 photons cm –2 s –1

Figure 11.—Detector placement concept. Detector placement is flexible and will be coordinated withthe spacecraft contractor.The GBM instrumentation is similar to that ofBATSE on CGRO, since observation of gamma-raybursts is the objective of both instruments. There aresignificant differences, however, due primarily to:1) The GBM emphasis on spectral measurements tocomplement the LAT, 2) the blockage of the GBMFOV by the LAT, and 3) the mass and cost constraintson the GBM. BATSE used eight large areadetectors (LAD’s) and eight smaller spectroscopydetectors, analogous to the NaI and BGO detectorson GBM. A comparison of the two instruments isprovided in table 2. Key improvements compared toBATSE for the spectroscopy goals are lower energycoverage obtained by using a beryllium window onthe NaI detectors, better high-energy coverage byincluding BGO detectors, and better temporalresolution of spectra via a TTE datatype withsufficient memory to record very bright GRB’s. Thelarger number of NaI detectors viewing a smallerFOV will reduce the systematic errors for burstlocations and allow an improved triggering algorithm.18

Figure 12.—Detector placement concept,top view.2.2 Data FormatsTo achieve all of our scientific goals, the burstmonitor will have four datatypes, two continuouslyproduced datatypes and two datatypes produced inresponse to a trigger. The continuous datatypesBSPEC and BTIME provide good temporal andspectral resolution at all times, while the triggerdatatype TTE provides 5 µs data for triggers and thetrigger datatype TRIGDATA reports location andspectral estimates determined on board along withrates to allow the determination of improved locationson the ground in near real time.The goal of the trigger datatype TTE is to providethe maximal information about a trigger, with 5-µstime resolution and 128 channel spectral resolution.The limitations of TTE data will only be thoseinherent in the counting statistics and energy resolutionof the detectors. The TTE data will optimizeour ability to correlate GBM data with LAT data.For example, the trend of pulse width with energysuggests that GRB pulses might be very narrow atGeV energies, an idea which has not been testedwith EGRET because of EGRET’s long dead timeper event and low counting statistics. The TTEdatatype will allow binning of the GBM data accordingto time boundaries determined from pulsesobserved in the LAT data. In normal operation, TTEdata will be accumulated for the BGO and the twoNaI detectors with the best view of the source. ForFigure 13.—Detector placement concept,side view.very bright bursts or when the TTE memory stillcontains events from a previous burst, data will beaccumulated from only one NaI detector. Approximately50 s of pretrigger data will be provided inTTE to enable analysis of any precursor emission.The TTE temporal resolution of 5 µs is chosen tomatch the expected maximum deadtime per countand to match the probable performance of the LAT,for which table 1 of the NRA specifies a requirementof 10 µs and a goal of 2 µs. The spectralresolution of 128 channels was selected tooversample the data; the 5 to 1,000 keV NaI dataare spanned by ~30 resolution elements, while the150 keV to 30 MeV BGO data are spanned by ~85resolution elements. A large oversampling factor isunnecessary for the BGO data because of thelimited number of counts at the higher energies.We base the estimates of the memory requirementsfor TTE on the bright burst, GRB 940217, whichhad the largest number of counts of the burstssimulated for section 2.5.1. A similar burst observedwith the GBM would produce 650,000 counts in theNaI detector and 210,000 counts in the BGO detectorwith the best view of the source. Hence TTEmemory to contain 1 million events will suffice toaccumulate the data from one NaI and one BGOdetector. For fainter events, to improve statistics,data will be accumulated from two NaI detectors.The temporal resolution of 5 µs and the spectral19

esolution of 128 channels will require 24 bits perevent, so the memory requirement is three megabytes.The other datatype produced in response to a trigger,TRIGDATA, will contain several record types.One record type will report location and spectralestimates determined on board. This informationcan be used by the spacecraft or LAT computer torepoint the spacecraft, and for rapid ground-basedobservations. If the mission chooses to support aspecial real-time telemetry mode in response totriggers, the other record type will contain selecteddetector rates to enable more accurate near real-timecomputation of locations on the ground, using amore capable computer than the GBM DPU.The two background data types are designed for thegoals of providing background data for burst analysis,providing data for nontriggered events and forextremely long GRB’s, >500 s, and to permitdetection of bright sources via Earth occultation.The BSPEC data type accumulates 128 channelsof data from each detector with 8 s resolution, whilethe BTIME data type provides 0.256 s resolutionin four energy channels. One or two of the BTIMEenergy channels will correspond to the triggerenergy band so that trigger sensitivity can be preciselycalculated on the ground.Together, the two background datatypes (BSPECand BTIME) require 3,600 bits per s. Using 5,000bits per s of telemetry, the entire TTE memory willbe read out in 1 hr 20 min. At the expected triggerrate of 0.5 day –1 , collisions between events will berare. Three methods will be used to save some TTEmemory for a second trigger: 1) TTE accumulationwill end at ~500 s so that background data fromweak events will not consume the entire memory,2) for very bright bursts or if some of the memory isoccupied by a previous trigger, data will be accumulatedfrom one instead of the usual two NaI detectors,and 3) as readout occurs, portions of thememory will become available for another trigger.The telemetry usage of the Burst Monitor willtherefore range from about 4 kbits per s to about9 kbits per s, depending on whether triggered dataare being downloaded.2.3 Flight System Hardware2.3.1 DetectorsTo cover the energy range from 5 KeV to 30 MeVtwo scintillator materials are chosen: Sodium iodidefor the low energies, 5 KeV to 1 MeV, and bismuthgermanate for the high energies, 150 KeV to 30 MeV.Bismuth Germanate DetectorsBGO scintillation detector crystals are selected toprovide high efficiency and adequate resolution forthe higher-energy range of the GBM. This type ofdetector is being designed and fabricated by theMPE team members for the thick shield sectionsof the SPI instrument on the INTEGRAL spacecraft.The BGO crystals will be manufactured byCrismatec Corp. in France. For the GBM application,the BGO detectors will have a uniform lightcollection geometry and large photocathode area,resulting in superior resolution. While the lightoutput of BGO is ~20 percent of that of NaI (Tl),the high density and high average atomic numberof this material makes it preferable for detectors athigher energies.Two identical detectors, mounted on opposite sidesof the spacecraft, each have a single cylindricalBGO crystal that is 12.7-cm diameter by 12.7-cmhigh. The energy resolution will be ~14 percent at661 keV and ~4 percent at 10 MeV. Resolution as afunction of energy, is shown in figure 14. They aresufficiently thick for photons up to 40 MeV, asshown in the response curve in figure 15. Becauseof their large volume, there is significant photopeakefficiency, up to the energy range of the GLASTmain instrument.Table 3.—BGO detector characteristics.Number of detectors 2ThicknessDiameterEnergy rangeResolution at 100 keVResolution at 662 keVResolution at 10 MeVResolution at 20 MeV12.7 cm12.7 cm150 keV to 30 MeV35% FWHM14% FWHM4% FWHM3% FWHM20

Each of these BGO crystals are optically coupledthrough a fused silica window to two 12.7-cmdiameter PMT’s, attached on both ends of thecylinder. This design allows a homogenous lightcollection over the detector volume and also providesredundancy should one of the PMT’s fail ordegrade. The BGO detector would still be operational,but with lower resolution and gain if thisoccurs. The two detectors will be mounted onopposite sides of the spacecraft, providing nearly a4 steradian FOV. The sensitivity of the detectors,in the directions through the PMT’s, will be somewhatdiminished at energies below ~400 keV. Table3 provides the basic characteristics of each BGOdetector. Figure 16 shows the angular response; notethat the BGO crystal orientation is such that 90°corresponds to the crystal axis of symmetry.Sodium Iodide DetectorsTwelve identical NaI scintillation detectors will beused to determine GRB locations, as described insection 2.5.2. The spectral range for these detectorswill be 5 keV to 1 MeV. These are conventionaldetectors with a diameter of 12.7 cm and a thicknessof 1.27 cm. The thickness and the large diameter-tothicknessratio results in a response function similarto the LAD’s on the BATSE instrument. The angularresponse of the NaI detectors, as a function ofenergy, is shown in figure 16. This design simplifiesthe GRB location methodology, using provensoftware for this application.Figure 15.—Response of a GBM BGO detector.The total curve is for detecting a count of anyenergy from an incident photon, while the fullenergy peak is the response for capturing theentire energy of the photon. The double arrowshows the energy range of the BGO channels.Each crystal will be viewed by a single 12.7-cmPMT, providing excellent light collection andhaving a homogeneous response over the wholecrystal. The detector module will be evacuated andhermetically sealed using standard practices, usedon other scintillation detectors, designed for flight.The detector entrance window will be made from0.25-mm thick beryllium, electron-beam welded toan aluminum housing. This approach was successfullyused for the BATSE spectroscopy detectors.There will be a thin, low-z, highly reflective materialjust behind the entrance window, in contact withthe crystal, to provide good optical reflectivity andFigure 14.—Energy Resolution of the GBM NaIand BGO detectors.Figure 16.—Effective area of the GBM detectorsas a function of angle of incidence.21

There are at least two qualified vendors for 12.7-cmdiameter (5-inch) PMT’s for the Burst Monitor:Electron Tubes Ltd. and Hamamatsu Corp.The selection of the vendor for the PMT’s, thephotocathode type and dynode structure will bemade during the phase B studies.Figure 17.—Response of a GBM NaI detector.The total curve is for detecting a count of anyenergy from an incident photon, while the fullenergy peak is the response for capturing theentire energy of the photon. The double arrowshows the energy range of the NaI channels.ensure adequate sensitivity down to 5 keV. ThePMT will view the crystal through a fused silicaoptical window. The energy response of the NaIdetectors is given in figure 17. This calculation doesnot take into account absorption by a thin thermalblanket, which will be designed somewhat inconjuction with the spacecraft insulation. It isassumed that this can be done without significantlydegrading the low-energy performance of the NaIdetectors. Table 4 summarizes the characteristics ofthe NaI detectors.Photomultiplier DesignBoth types of detectors will use the same PMThousing design and preamplifier, with slight variationsin the bleeder strings and preamps, to accommodatelarger pulses from cosmic rays expected inthe BGO detectors. The NaI detectors will likelyoperate at higher voltages to achieve a greater gainthan the BGO detectors.Number of detectors 12ThicknessDiameterTable 4.—NaI detector characteristics.Energy rangeResolution at 6 keVResolution at 200 keVResolution at 300 keV1.27 cm12.7 cm5 keV to 1 MeV50% FWHM16% FWHM13% FWHMMechanical and Thermal InterfacesThe detectors will be mechanically mounted ontothe spacecraft with struts or brackets to provide therequired viewing orientation. Details of the mountingdesign must await the selection of the GLASTmain instrument and spacecraft. It is expected thatthe thermal requirements of the detectors can bepassively met using the same multilayer insulation(MLI) blankets that cover the spacecraft structure.The covering of the entrance window of the NaIdetectors may need to have fewer layers of MLI, toallow adequate x-ray transmission for the NaIdetectors. It is assumed that both the thermal blanketdesign and the mounting design will be providedby the spacecraft contractor, in consultation with theGBM team.CalibrationPreflight calibration of the GBM will be accomplishedusing a combination of Monte Carlo simulationsand calibrations. The GEANT simulationpackage (Brun, et al. 1993) will form the basis forsimulation of the photon interactions in the detectors,including secondary leptons and nuclearexcitations. Hadronic event responses (i.e., background)will also be simulated with GEANT linkedto FLUKA (Aarnio, et al. 1990). A detailed geometricaland chemical model of the detector unitsforms the basis for these simulations, supplementedby a coarse model of the GLAST main instrumentand spacecraft.Simulation codes can be uncertain in absolutemagnitude on the order of 30 percent, and the massmodel may be in error. Therefore calibration measurementsat specific energies and incidence anglesconstitute the second base for the detector responsesof the Burst Monitor. These measurements will beperformed at MPE prior to delivery of the detectorsto MSFC. Gamma-ray sources are available ascalibrated radioactivity standards, ±5 percent inintensity, in the energy range up to 4.4 MeV photon22

Table 5.—Gamma-ray sources for BGOcalibration measurements.Source Energy Angles241Am 59.5 keV 0˚, 60˚, 135˚57Co 122 keV 0˚137Cs 662 keV 0˚54Mn 835 keV 0˚22Na 1,275 keV, 511 keV 0˚, 60˚, 135˚88Y 1,840 keV 0˚24Na 2,754 keV 0˚241Am/ 9 Be 4,430 keV 0˚, 60˚, 135˚19F(p,a) 16 O 6,100 keV 03He(p,n) 4 He 19,800 keV 0energy. To calibrate the BGO detectors at higherenergies, up to 20 MeV, we will use nuclear gammarays from reactions triggered by a proton beam at aVan de Graaf accelerator. We have previouslyperformed such accelerator calibrations for theCOMPTEL instrument aboard the CGRO and thecalibrations of the SPI instrument aboard the ESAINTEGRAL Observatory will be completed by theend of 2000. Very detailed calibrations will beperformed in energy space, due to the requiredprecision of the differential response, for accuratespectral deconvolution. Additionally, at severalenergies, the directional change of the response willbe calibrated. Source alignment, with respect to thedetectors, through theodolite precalibrated setupswill be sufficient. Calibration data will be recordedcoincidentally with a standard 3-in. NaI detector, forwhich simulation codes have been extensivelycompared and checked. Analysis will use the toolsalso used for detector design and for science dataanalysis.Tables 5 and 6 list the calibration energies andangles for the BGO and NaI detectors, respectively.2.3.2 Power SuppliesThe low-voltage power supply (LVPS) providespower to the DPU and to the preamplifiers on eachPMT. There will be two LVPS’, cross-strapped forredundancy. There will be an HVPS for each PMTof the GBM, all housed in a single box. The HVPSwill be under the control of the DPU. Power supplydesigns will be conventional, space-qualified designs,derived from other flight programs such asROSAT and INTEGRAL, both of which were MPEprojects.2.3.3 Data Processing Unit2.3.3.1 IntroductionThe Burst Monitor instrument consists of 14 independent,remotely located detector modules. Eachmodule provides analog pulse height signals at ahigh rate. Rather than process the signals at eachdetector, the centrally located DPU collects, processesand packages data from all the individualdetectors. The main functions of the DPU are todigitize the analog pulse height signals, accumulatetime resolved, pulse height spectra with variabletemporal resolution, tag the spectra with time anddetector information, package the data for telemetry,and use the data to realize a burst trigger, andcompute burst locations. The other important DPUfunctions are to control the operation of the instrument,including power supply settings, and toaccumulate adjunct housekeeping and instrumentstatus data (temperatures, voltages, currents, etc.)for inclusion in the telemetry. The DPU acts as thesole electrical interface between the Burst Monitorinstrument and the GLAST spacecraft computer. Itis also the electrical interface for instrument integrationand instrument level testing.2.3.3.2 RequirementsThe DPU requirements are summarized in table 7.The origin of the requirements is explained insection 2.2. The most important demands on theDPU design are imposed by the expected data rates,in particular the nominal background rate, themaximum rate during GRB events, and the cumula-Table 6.—Gamma-ray sources for NaIcalibration measurements.Source Energy Angles55Fe 5.9 keV 0˚, 45˚, 80˚109Cd 22, 88 keV 0˚, 45˚, 80˚241Am 59.5 keV 0˚22Na 511, 1,275 keV 0˚, 45˚, 80˚137Cs 662 keV 0˚54Mn 835 keV 0˚88Y 1,840 keV 0˚23

Table 7. —Data processing unit requirements.Inputs14 detector/PMT analog signals, 0–5 VHousekeeping sensors (analog/digital)14 power supply voltage and currentTemp. sensors for each electronics board, each electronics box and each detectorDigital commands from spacecraft (S/C) interfaceClock Sync. from S/CDetector Inputs Count Rates 0.2 kcps per det, nominal100 kcps per det, peak; for ~100 sec; 1/wkADC resolution12 bitsDead time < 5 µsDynamic range 200:1Time-tagging5 µs resolutionFPGA’s Functions Data routing and buffering; spectralaccumulation; time tagging of events—allcontrollable via CPUTypeTBD, ~10 MHzCPU Functions Data transfer & formatting; control of FPGAand detector power supplies; burst trigger;burst location determinationTypeTBD, 5-10 MHzMemory4 MB data, 1 MB programSerial Command Interface Functions Transmit digital commands from CPU todetector power suppliesSpacecraft Interface Functions Commands, data and power; ~9 kbps datarateOther Hardware UTC clock Synchronized with S/C UTC clockTelemetry buffer~4 MBSurvival heaterstypical for space hardwarePhysical Size TBDWeight< 3 kgPower consumption< 5 wtive counts for a GRB. These data rates and theinstrument dead time requirements determine theappropriate processing speed, memory buffer depth,and processing architecture. The other major DPUrequirement is the need for rapid computation ofburst locations using data from all detectors.2.3.3.3 Hardware DesignThe block diagram shown in figure 18 provides aconceptual view of the baseline DPU design. Thereare two different board types: the data receiverelectronics board (DRE) and the processing electronicsboard (PRE). The general redundancystrategy for the DPU is that each board has onemain unit and one unpowered redundant spare.If the main unit fails, it can be deactivated and theredundant unit activated via spacecraft command.The DRE board is responsible for receiving theanalog detector data and converting it into useabledigital form. There are two identical analog dataprocessors, operating in parallel to accommodatethe required data rate. These consist of analogcircuitry to perform peak-hold and pulse shapingtasks on the input PMT pulse heights from the14 detectors. The conditioned pulse height signals24

Figure 18.—Conceptual view of the baseline DPU design.are then multiplexed into 12-bit ADC. This representsa significant oversampling of the detectorenergy resolution to allow for flexible, programmabledata binning. The digital signals are processedby an FPGA. This unit performs the tasksof event buffering, spectral accumulation, timetagging, and detector addressing under control ofthe CPU on the PRE board. In addition to the maindetector data path, the DRE also digitizes and routesadjunct housekeeping data from temperature,voltage, current, and scalar rate sensors dispersedthroughout the Burst Monitor instrument. Thesedata are multiplexed, digitized, and fed into thetelemetry stream at a low rate through the FPGA.The main functions of the PRE board are to act as acentral control unit, route the data to appropriatetelemetry buffers, and enable onboard burst triggeringand burst localization. The main data bus, forboth detector and housekeeping data, flows from theFPGA on the DRE board into a data storage bufferon the PRE board. The CPU then sorts, packages,and passes the data to a telemetry buffer before theyare sent to the spacecraft data bus through the DPUspacecraft interface. As the data are accumulated,the CPU passes selected detector rates to the bursttrigger program. In the event of a trigger, commandsare automatically issued to the DRE FPGA thatswitch from background to burst accumulationmode. Concurrently, selected detector rate data arepassed to the burst localization program. When thisprocess is complete, a priority burst location triggeralert message is inserted into the telemetry buffer,where it is passed to the spacecraft. The LAT andground telemetry will have access to the bursttrigger location messages via the spacecraft CPU.The PRE board contains a coordinated universaltime (UTC) counter that is synchronized with thespacecraft clock to facilitate time tagging of burst25

triggers and location alert messages. The PRE clocksignal is also routed to the DRE board to allow fasttime tagging of individual events and spectra. ThePRE board also includes serial command interfaceelectronics that allow control of the detector powersupplies, and therefore detector gain. The serialcommands can originate via an automatic gaincontrol (AGC) program run by the CPU, or bydirect command from the spacecraft commandinterface. The AGC program operates by monitoringthe peak channel of the 511 keV background lineand adjusting detector HV accordingly. The serialcommand interface is also used to deactivate detectorHV during passage through the South AtlanticAnomaly (SAA) via commands from the spacecraftcontroller.2.3.3.4 SoftwareThe DPU, FPGA, and CPU require several softwareelements to complete the tasks described above.These are described in detail in section 2.4.2.3.3.5 Resource EstimatesResource estimates for the DPU are shown intable 7.2.3.3.6 HeritageOur DPU may be provided as a modification of theCompact Environmental Anomaly Sensor (CEASE)manufactured by Amptek, Inc., Bedford, MA. Thispackage has been selected for several Department ofDefense (DoD) missions.2.4 Flight SoftwareThe GBM flight software will reside in the DPUand perform the following functions:• Receive commands, act on real time and storedcommands• Receive and package housekeeping data fortelemetry• Control transfer of background data accumulatedby FPGA’s to a spacecraft solid staterecorder for latter telemetry• Provide burst trigger• Control accumulation of TTE trigger dataprovided by FPGA• Transfer TTE data to a spacecraft solid staterecorder for latter telemetry• Provide AGC of the detectors using 511 keVbackground line• Buffer data to be used as background for triggeringand location calculation• Produce classification of triggers and estimationof spectrum• Calculate and output location of trigger• Produce and output TRIGDATA data forground-based location calculation.In normal operation, the flight software will supervisethe accumulation of BTIME and BSPECbackground data types, and buffer these data fortransfer to the spacecraft solid state memory. Thelast 200 s of background data will be kept by theflight software to serve as a background reference indetermining triggers and burst locations. The flightsoftware will convert sensor signals (thermistors)and hardware status signals (voltage levels) intohousekeeping data. It will receive commands andeither act upon them or store them for later use.Commanded functions include HV power on/offand levels, parameterized changes in the trigger,classification, and burst location algorithms, andselection of energy channel ranges for BTIME data.An important function of the flight software istriggering in response to a flux increase. TheBATSE approach is to require a 5.5 σ increase intwo detectors. With more detectors and detectortypes than BATSE, a different algorithm mightachieve better sensitivity while maintaining anegligible false trigger rate. Triggers will normallyactivate the TTE accumulation and production ofTRIGDATA data for real-time telemetry. Flightsoftware tracks the availability of TTE memory as itis filled by burst accumulation and emptied, bytransferal of the data, to the spacecraft solid staterecorder. The flight software will classify the causeof the trigger, so events that are probably not GRB’smight be processed differently, and the spacecraftmight not be repointed. Triggers caused by particleevents can be identified by detector rates inconsistentwith a point location at infinity, while solarflares can be identified by their location. A spectrumwill be estimated as a possible input in deciding26

whether to repoint the spacecraft. The flight softwarewill calculate a rough location within severalseconds. For long bursts, revised locations withreduced statistical errors will be calculated as theburst progresses. Special TRIGDATA data will betransferred to the spacecraft for immediate telemetryto enable more accurate locations to be calculatedon the ground in near real time.Triggering and onboard location algorithms will bemade easier by maintaining constant energy boundariesof the channels via gain control. Gain controlwill be performed by monitoring the 511 keVbackground line in each detector and adjusting thePMT HV’s or amplifier gains. BATSE experienceshows gain variations to occur with 12- and 24-hrperiods, depending on spacecraft altitude, in responseto temperature variations and SAA particledose. Because of good magnetic shielding, gainvariations on orbital timescales are very low. Sufficientcounts in the 511 keV line for accurate gaindetermination will accumulate once or twice per90 min orbit.The DPU vendor will provide some basic softwarenecessary for hardware testing. This will includereading the FPGA’s, controlling the TTE memory,accepting commands, outputting data, and outputtingHVPS control commands. We will try to reduceduplication of software by making use of vendorgenerated software where this is expeditious.We incorporate the capability to revise the flightsoftware by command, a feature that was used togreat advantage on BATSE on several occasions.For example, the BATSE data stream was altered tocompensate for the failure of the flight tape reordersearly in the CGRO mission. The GRM will alsoprovide memory dumps and memory check sums toenable detection and correction of single eventupsets, also as is done on BATSE.Development of the flight software comprises thefollowing tasks:• Study tradeoffs and algorithm for calculatinglocations on board• Product specification—requirements and design• Management plan• Assurance and test procedures• Software coding• Assurance and test reports• Software maintenance document (incorporatingversion description)• Software Users GuideDuring phase B we will study the best division oftasks between hardware and software. We will studyhow to implement the onboard location algorithm,assessing tradeoffs between location accuracy, timeto calculate locations, and CPU speed and memoryrequirements. A strategy for optimal use of the TTEmemory will be defined. These decisions and therequirements to perform the tasks listed above willbe incorporated into a Product Specification. Respondingto these requirements, the software designwill be described in the Product Specification,proceeding from a concept, to an architecturalspecification, to a detailed design. Using the ProductSpecification, a Management Plan and a documentof Assurance and Test Procedures will beproduced. The Management Plan will break into thesoftware design, coding, testing, and documentationstages, document the cost and schedule of thesestages, and describe methods of monitoringprogress and responding to difficulties. The Assuranceand Test Procedures document will describehow the software will be tested for correct functioningand satisfaction of the requirements. The Assuranceand Test Procedures will be used at stages ofsoftware coding, software acceptance testing, andduring integration. Results of these tests will bereported in the Assurance and Test Reports.Two principal documents, of continuing use, will beproduced. The User’s Guide will instruct operationsstaff and scientists on operation of the software,commands, and capabilities. The User’s Guide willalso contain descriptions of the data and data formatsthat will be used in the development of theoperations and analysis software. The MaintenanceManual will describe implementation details,modification aids, and how to adapt the code. It willassist programmers or scientists in implementingimprovements or identifying and solving errors.27

These plans and documents will be produced inaccordance with the NASA Software DocumentationStandard.2.5 System Performance2.5.1 Time-Resolved SpectroscopyPerformance2.5.1.1 Simulation and Analysis ApproachAs a demonstration of the performance of theGLAST mission with the GLAST Burst Monitor,we have created simulated spectra for the GBM NaIand BGO detectors, and for a conception of abaseline LAT. These simulated spectra are basedupon realistic models of the detectors and thebackgrounds and incorporate Poisson fluctuations.The simulated spectra are based upon GRB spectraobserved with instruments on the CGRO.For modeling the response to gamma-ray bursts, arepresentative direction was selected, 30˚ from theGLAST instrument axis and 30˚ azimuth. Thisplaces the hypothetical source at 30˚ from the axesof both the LAT and a GBM BGO detector, at 14.9˚from the axis of the best-illuminated NaI detector,and 29.0˚ from the axis of the second best-illuminatedNaI detector.The detector response models (DRM’s) were producedby Monte Carlo simulations of electromagneticcascades using a modified version of GEANT(Brun, et al. 1993). This program propagates photonsand electrons down to 1 keV and incorporates amodel of Compton scattering more accurate thanthe Klein-Nishina cross-section. The mass modelused in GEANT included the scintillation crystals,the detector housing, and the photomultipliers andtheir housings (see section 2.3.1). All of this materialwas illuminated with Monte Carlo photons sothat the response includes scattering of photonsfrom nearby materials into the scintillators. Theenergy range of the simulated photons exceeded theenergy range that will be recorded by the GBM (i.e.,5 keV to 1 MeV for the NaI, 150 keV to 30 MeVfor the BGO) to incorporate into the responsephotons outside the GBM channel range that produceevents in the GBM channel range via resolutionbroadening (fig. 14) and partial energyFigure 19.—Detector response as modeled withGEANT. The response as equivalent area isplotted vs the energies of the incident photons(y-axis) and the energies of the detected counts(x-axis). The prominent diagonal band is the fullenergyor photopeak response. The “wing”extending to the lower left from the photopeakresponse is due to the escape of fluorescent x-rays. The blue region in the upper left of theBGO diagram is the response to gamma raysscattered into the detector from nearby material.deposition. The energy range of input photons wasspanned by 128 pseudo-logarithmic bins. For eachbin, photons were simulated until 100,000 counts28

Figure 20—Background rates in the GBM BGOand NaI detectors. The estimates of the total rateand of the contribution of the diffuse x-ray areshown separately. The energy ranges of thechannels are depicted in bold.were detected. The response models are depicted infigures 15, 16, 17, and 19.Two terms contribute most of the background ratesin both NaI and BGO detectors: at lower energiesthe background from the diffuse sky flux dominates(Gehrels 1992), while at higher energies inducedradioactivity dominates. Induced radioactivity isproduced primarily by primary cosmic rays andhigh-energy (>100 MeV) proton irradiation inpasses through the SAA. Lower-energy protons loseenergy by ionization before they can producesignificant spallation.The background rate in the GBM NaI detectors wasestimated by scaling the background rate of aBATSE LAD and correcting for the differing transparenciesof the housings to the diffuse sky flux.The GBM NaI detectors have a full-width berylliumwindow, while the NaI of the BATSE LADs iscovered by an aluminum window and a plasticscintillator with aluminum covers. We thereforecalculated the diffuse sky flux contribution to thebackground of the BATSE LADs, subtracted it fromthe observed background, scaled the resultingbackground rates to the smaller size of the GBMNaI detectors, and then added the diffuse sky fluxbackground rate calculated for the GBM detectors.The background estimate is on firm foundationsbecause: 1) We have relied on actual space measurementsin a NaI detector of similar aspect ratio,view-angle characteristics and in a similiar orbit, 2)since the BATSE LAD’s and GBM NaI detectorshave the same thickness, the same factor simultaneouslyscales for area and volume dependentbackground effects, and 3) the diffuse sky flux iswell known. The contributions of the diffuse skyflux to the background rates were modeled withGEANT, using an isotropic flux rather than theplane waves used for the DRM’s. A simple model ofthe spacecraft was used to block photons from aportion of the sky. The resulting background modeland the diffuse sky flux contribution are shownfigure 20.The background in the GBM BGO detectors wasestimated by scaling the background rates ofBATSE Spectroscopy Detectors. Two Spectroscopydetectors operating at different gains were used tospan the energy range of the BGO detectors. Sincethe purpose of the BGO detectors is higher energycoverage, and because the induced radioactivitydominates at these energies, the scaling was basedupon the detector mass. In the energy band wherethe diffuse sky flux background dominates, thisprocedure estimates a background rate somewhatabove our estimate of the diffuse sky backgroundcomponent (figure 20).This procedure produces a good background estimatebecause both detectors are uncollimated withsimilar dimensions and similar view-angle characteristics.It is also known that for high-Z materialsirradiated by high-energy protons, the intensity,decay and spectral characteristics of the resultinginternal radiation, to first order, are dependent onlyupon the total mass of the material and not upon theparticular target nuclei. This results from the largenumber of radioactive isotopes produced and theaverage nuclear characteristics of the spallationproducts (Fishman, 1977, and Barbier, 1969). This29

assumes that the major fraction of the mass on thedetector is contained in the high-Z elements. This istrue for both NaI and BGO; thus we have assumedsimilar induced radioactivity per unit mass for bothtypes of detectors.The LAT is included in our modeling to show theperformance of the entire mission. Because we donot know the performance of the LAT that will beselected, we have modeled our conception of abaseline LAT that meets the baseline requirementsgiven in table 1 of the NRA. Because of the sparsityof the specifications in table 1, we fleshed out aconception of the baseline LAT using our judgmentof the general performance characteristics of anyinstrument in this energy range. For example, theNRA specifies the effective area as 8,000 cm 2 , whileany instrument will have a decreased effective areanear its threshold. At 30˚ off-axis, our assumedeffective area is 5,000 cm 2 at 1 GeV, 3,400 cm 2 at100 MeV, and, just above the 20 MeV threshold,520 cm 2 at 25 MeV.Because the LAT will measure the direction ofincident photons, there is no need to simulate theresponse to the isotropic sky flux. Assuming theGRB to be at high galactic latitude, we simplymultiply the extragalactic diffuse-sky flux measuredwith EGRET (Sreekumar, et al. 1998) by the detectorresponse matrix. In a simple model of sourcedetection, we only use counts within the 68-percentradius of the point spread function. Correspondingly,the background is evaluated for this area andthe effective area is reduced.For each detector a simulated dataset consists oftwo spectra, a background spectrum and a sourceplus-backgroundspectrum. For the GBM detectors,a background-only spectrum is made by addingPoisson fluctuations corresponding to a 500 sobservation of the background count-rate model.Real observations will have background variations.With BATSE, we model the variations with loworderpolynomials. The BATSE experience is thatstatistical fluctuations dominate and these are wellrepresentedby Poisson fluctuations on a constantbackground rate. For the LAT analysis a moresophisticated background model will be necessaryto describe the temporal and directional variations.We assume that this model will produce backgrounduncertainties corresponding to the Poisson fluctuationsof a 10,000-s background observation.The photon model used for these simulations is thestandard Band “GRB” function (Band, et al. 1993),which is one representation of a four-parametermodel in which two power laws are smoothly joined(section 1.1.2.3). The source count rate model iscreated by applying the detector response model tothe assumed photon model. The total model for thesource interval is the sum of the source count ratemodel and the background count rate model. Poissonfluctuations in the counts are simulated basedupon a livetime slightly below the duration of thethe real GRB spectrum that is being modeled.The simulated data are fit using the standard forward-foldingprocedure: a parameterized photonflux model is assumed, the photon model is multipliedby the detector response matrix, and theresulting count model is compared to the detectedcounts using a statistic. For the comparison statistic,in order to correctly treat the small number ofcounts in the LAT channels and the high-energyBGO channels, we use maximum likelihood withthe Poisson probability distribution. Similar resultsare obtained when2with model variances is used.The fits to the data of several detectors are true jointfits, with count models for each detector generatedfrom a single photon flux model. The fitting softwareis directly applicable to real GLAST data.2.5.1.2 Spectral PerformanceGRB 940217 is used as an example because itsspectral parameters are comparatively well determinedfrom observations by BATSE, COMPTELand EGRET. EGRET observed an 18-GeV photon90 minutes after detectable emission had ceased inthe BATSE data (Hurley, et al. 1994). Since it waswell observed with COMPTEL and EGRET, it isnecessarily a bright event, with a 50- to 300-keVfluence placing it in the brightest 0.5 percent observedby BATSE (fig. 21). The time history iscomplex, with a series of pulse complexes spreadover 180 s (fig. 3).Our simulation is based upon the time-integratedspectrum because that is the spectrum for which30

14.9˚ off-axis, and one BGO detector and thebaseline LAT both viewing the source at 30˚ offaxis.In the NaI detector the burst is detected athigh-statistical significance in each of numerouschannels; the BGO detector sees the source in allbins including the 10 to 30 MeV bin, bridging thedifficult few MeV region to the threshold of thebaseline LAT, which detects the burst to about 1GeV. The GLAST mission with the GBM woulddetect a burst like GRB 940217 over 5.3 decades ofenergy.Figure 21.—The fluence distribution of GRB’s.The fluences of GRB 940217 and GRB 990123are shown in the context of all GRBfluences measured with BATSE over 8.5 years.The dashed line is a power-law of slope –3/2indicating the fluence trend of bright bursts.The dotted lines show that more than 160 burstshave fluence values greater than one-tenth thefluence of GRB 940217. Correcting for theBATSE bursts for which data gaps preventfluence determinations, ~24 bursts with fluencesgreater than one-tenth the fluence of GRB940217 are observed yearly.COMPTEL and EGRET spectral results have beenreported. Analysis of the EGRET TASC data gives ahigh-energy power-law index of -2.5 ± 0.08 (Hurley,et al. 1994), while COMPTEL telescope dataindicate an index of -2.6 ± 0.11 (Winkler, et al.1995). We have obtained the four spectral parametersof the Band GRB function by fitting the dataof three BATSE spectroscopy detectors, whichtogether span the energy range 20 keV to 28 MeV,for a 188 s interval which includes essentially allof the burst flux. The COMPTEL and EGRETresponse is better at high energies, so we imposedthe requirement that = –2.6, obtaining from the fitthe parameter values A = 0.0181 photons s –1 cm –2keV –1 , E break= 760 keV and = –1.26.Because this is a very bright burst, the GBM wouldoperate in memory conserving mode and accumulateTTE data from only the best-illuminated NaIdetector. Our simulated spectrum (fig. 22) is thereforefor one NaI detector viewing the source atThe shape parameters obtained from the fit areE break=746 ± 12, = –1.261 ± 0.003 and = –2.68± 0.01. Comparing to the values assumed for thesimulation (listed above), the values for E breakandare in excellent agreement, deviating by 1.1 and 0.3respectively. The disagreement between theassumed and fit value for is small in absolute units(0.08) but large in -units (7.1).We now show that good results are obtained whenwe analyze a burst with the same spectral shape asGRB 940217, but 10 times dimmer. We do not usean actual dimmer burst as an example, because goodCOMPTEL and EGRET results on the spectra ofsuch bursts are lacking. Because the cumulativefluence distribution follows a –3/2 power law downto at least 10 –5 ergs cm –2 (fig. 21), bursts like thedimmed example will occur 32 times more frequently.Even though the –3/2 power law can nolonger be understood as indicating which burstsoriginate in Euclidean space where cosmologicaleffects are negligible, it is still an observational factthat fluence distribution of bright bursts is welldescribed by a –3/2 power law.We show in Figure 23 a simulation of GRB 940217,dimmed by X10 (A = 0.00181 photons s –1 cm –2 keV –1 )but otherwise using the Band GRB function parametervalues given above. Because this is a dimmerburst, GBM would accumulate TTE data from twoNaI detectors and we use simulated data from fourdetectors. Because of the decreased number ofcounts, the error bars on the parameters are larger:E break= 1034 ± 146, = –1.32 ± 0.02 and = –2.78± 0.06, deviating by 1.9 , 3.0 and 3.0 from thevalues assumed for the simulation. The quoted errorbars are for single parameters of interest; the agree-31

Figure 22.—Simulated spectra of GRB 940217.The top box shows the best-fit count rate models (histograms) compared to the simulated count ratedata (points), while the bottom box shows the deconvolved spectra. The channels have been rebinnedinto broader bins for display purposes—the fit is made to the data at full resolution. Points within 1σof zero are plotted as 2σ upper limits.32

Figure 23.—Simulated spectra for a dimmed version of GRB 940217.The spectral shape is that of GRB 940217, but the intensity has been reduced 10-fold.33

ment is better when the parameter cross-correlationsare considered.2.5.1.3 Time-Resolved Spectroscopy PerformanceOur second example burst is GRB 990123, the onlyburst for which prompt optical emission has beendetected (Akerlof, et al. 1999). The wide-bandspectrum observed with all four instruments on theCGRO is shown in figure 6. High-energy flux wasdetected in a low-gain BATSE Spectroscopy Detector,COMPTEL and the EGRET TASC (the EGRETspark chamber was not operating). In the gammarayband, the burst is notable for its high fluence(figure 21) and the high value of E breakreached forthe peak of one pulse, 1470 ± 110 keV (figure 7). Atthe observed redshift of z ≥ 1.61, the gamma-rayemission, if isotropic, is at least 1.6 × 10 54 erg(Briggs, et al. 1999a).Figure 24.—Simulated spectral parameter time history of GRB 990123.The top panel shows the light curve of GRB 990123 observed with BATSE. The dashed histograms inthe remaining panels show the parameter values obtained by fitting BATSE data. Simulations of GBMand LAT data were made assuming the BATSE parameter values. The values from joint fits to GBMand LAT data are shown in red, while the LAT-only measurements of β are shown in blue.34

Our goal is to use this burst as an example of GBMperformance for time-resolved spectroscopy.Because of the coarse time resolution of the datafrom COMPTEL and EGRET, and due to thediffering phasing of the accumulation intervals, thepublished spectrum (figure 6) is for a 32-s intervalencompassing most of the burst flux. We thereforeuse data from two BATSE Spectroscopy Detectorsto determine the spectrum for 8 intervals. The twoSpectroscopy Detectors together cover the energyrange 26 keV to 25 MeV with a flux detection to atleast the 4 to 8 MeV band (figure 7). The timeresolvedvalues obtained for using the BATSEdata cluster about the time-integrated value of ~–3found by the four Compton instruments (Briggs, etal. 1999a table 1). The first row of each trio of rowsin table 8 shows the Band GRB function parametervalues obtained from this BATSE data.Parameter values from the fits to the BATSE dataare assumed for the purpose of simulating GLASTGBM and LAT spectra and therefore become the“true” parameter values to which the simulatedresults can be compared. Results from fitting thesimulated GLAST spectra have both actual noisefrom the BATSE observation and simulated noisefrom the GLAST simulations. The eight spectrahave a wide range of realistic GRB spectral shapesand demonstrate the performance of the GBM andLAT combination for time-resolved spectroscopy.The parameter values obtained by fitting one GBMNaI detector, one GBM BGO detector, and thebaseline LAT are listed on the second row of eachtrio of rows in table 8 and are depicted with redsymbols in figure 24. The agreement between theassumed values (dashed histogram) used to createthe simulations and the best-fit GBM/LAT values isexcellent with only one or two exceptions for .The final row of each trio lists the values of thehigh-energy spectral index obtained from fittingonly the LAT data (blue points in the bottom panelof figure 24). The LAT-only values are also inexcellent agreement with the “true” values; however,the errors are much larger. In two intervalsTable 8.—Time resolved spectroscopy performance.Start time(s) End Time(s) A E break(keV) Alpha BetaBATSE 0.090 19.648 0.015 200 –0.49 –3.3GBM+LAT 203±9.3 0.410±0.067

there no significant flux detection in the LAT andtherefore no constraint on the spectral index.The combination of the GBM and LAT clearlyprovides a much fuller picture of the spectral shapeand evolution of the burst than can be obtained withthe LAT alone. The six measurements of the spectralindex obtained from the simulated data of thebaseline LAT are all consistent with their mean of2.80 ± 0.15, so in this example the baseline LAT isunable to detect spectral evolution. We hope that theselected LAT will be more sensitive than thebaseline; nevertheless any LAT by itself will beunable to measure the crucial parameters E breakand .Burst data from both the GBM and the LAT willconsist of TTE, so the data from a burst could beanalyzed for whichever time intervals seem appropriate,e.g., based upon pulse structure seen with theLAT. The error bars obtained for this example aresmall enough that in an actual analysis, finer binningwould probably be selected.2.5.2 Burst DetectionCapabilities for detecting and locating bursts aredetermined by size, number and orientations of theNaI detectors, and the degree to which the LATblocks their FOV. Our calculations are based onseveral assumptions and approximations, describedbelow.Burst Detection TechniqueOur baseline triggering scheme is similar toBATSE. We require two detectors to be above athreshold specified in standard deviations abovebackground. The energy interval used is 50 keV to300 keV, and the time interval used for sensitivitycalculations is 1.024 s, although other trigger timeintervals may also be employed, as is done withBATSE. The baseline threshold will be 4.5 σ abovebackground, rather than the value of 5.5 normallyused for BATSE. This reduction is possible becauseof the lower sensitivity of the Burst Monitor, whichprecludes triggering on fluctuations from CygnusX–1. The accidental trigger rate, based on Poissondistributed statistical fluctuations in the backgroundrate at a threshold of 4.5 σ, is about 0.05 per year.BackgroundThe background in the NaI detectors is readilyscaled from BATSE, since the orbit will be similarand the detectors are the same thickness. Theaverage rate will be 156 counts/s, with orbitalvariations of about a factor of 3.Detector OrientationsIn our baseline configuration, the detectors areoriented in four banks of three units. The zenithangles of the detectors in each bank are 30°, 60°,and 90°. Each bank is oriented at a different azimuth,equally spaced by 90°. Only the NaI detectorsare considered in these calculations.LAT BlockageWe model the LAT blockage using the conservativeassumption that it blocks all azimuthal angles, inspacecraft coordinates, beyond 90° of the azimuthof the detector normal. This is equivalent to modelingthe LAT as an infinitely long cylinder. There isno blockage of detectors at 90° zenith angle and theamount of blockage approaches half of a detector’sFOV as the zenith angle approaches zero.Detector ResponseBased on BATSE experience, we approximate thedetector response between 50 and 300 keV as asingle conversion factor of 0.8 counts/photon. Withthis value, the simplified calculation of triggerefficiency employed here reproduces the BATSEefficiency quite well. For these calculations, theangular response is considered to be a cosine function.The baseline system has full-sky coverage forsufficiently strong bursts, although the sensitivitydrops rapidly at zenith angles above ~120°. Figure 25shows the projected area, in units of one detector(126.7 cm 2 ), of the second most brightly illuminateddetector as a function of zenith angle for severalazimuthal angles. This plot provides an indicationof the trigger sensitivity, since triggering requirestwo detectors above threshold. Figure 26 shows asimilar plot for the sum of the detectors illuminatedby the burst. This plot provides an indication of thequality of spectra produced by summing detectoroutputs. Averaged over the whole sky, the averageprojected area is 340 cm 2 , equivalent to 2.68536

Figure 25.—The projected area of the secondbest-illuminated detector. In a trigger schemewhich requires a significant signal in twodetectors, the second detector is the critical one.detectors. The effective FOV, defined in the GLASTAO as the sensitivity integrated over solid angledivided by the peak sensitivity, is 8.6 steradians.Locations can be computed for sufficiently strongbursts that illuminate three or more detectors, whichis the case over a solid angle of 11.5 steradians.The burst trigger sensitivity is shown in figure 27.The fraction of the sky over which a burst is detectableis plotted against the peak flux, where peakflux is defined as integrated over 1 s and between50 and 300 keV. The absolute threshold is 0.57photons/cm 2 –s, although lower flux bursts will bedetected at times when the background is lowerthan the average. The flux at which the system is50 percent efficient is 0.74 photons/cm 2 –s, at thisflux level bursts can be detected over 2π steradians.This is a conservative calculation of the threshold,since it assumes the BATSE trigger algorithm,which demands two NaI detectors individuallyabove threshold. For the larger number of detectorsused here, this scheme is clearly not optimum.Lower thresholds can be achieved by summing therates of closely pointing detectors and including theomnidirectional BGO detectors. For example, aburst with a peak flux of only 0.35 photons/cm 2 –swill yield a 5 excess above background in the 1-ssummed rate of the four upward facing detectors.Alternative trigger schemes will be investigated inphase B. Independent of the GBM triggering, burstsidentified by the LAT can be analyzed using the0.256-s resolution BTIME data at fluxes well belowthe trigger threshold.Based on the Burst Monitor sensitivity and the burstintensity distribution determined by BATSE, wefind that the Burst Monitor will trigger on about150 bursts per year. This rate calculation assumesSAA dead time and Earth blockage similar toBATSE and random pointing directions. The rateFigure 26.—The projected area of the detectorsilluminated by the burst.Figure 27.—Trigger sensitivity for bursts.The curve shows the fraction of the sky overwhich bursts of a particular peak flux willtrigger the GBM.37

will be up to 50 percent higher if the GLAST +Zaxis is preferentially zenith pointing, as currentlyplanned. Based on the GLAST Science Requirements,the LAT will trigger on 50 to 100 bursts peryear. Although this estimate is sensitive to thepoorly known distribution of spectral indices above1 MeV, the GBM will probably have a more sensitiveburst trigger than the LAT for all but very hardbursts.2.5.3 Burst LocationsLocating GRB’s by comparing rates in detectorsthat are facing in differing directions was pioneeredby the Russian Konus experiments (Mazets, et al.1981). The method has been very successfullyimplemented in the BATSE instrument. Our experiencewith locating GRB’s using BATSE will guideus in locating GRB’s with the GBM.For the GBM, we plan a three-stage refinement oflocations of GRB’s: On board, ground automated,and ground manual. Each serves a different purposeand represents a different trade between accuracyand speed.The onboard location is used to repoint the spacecraftto allow the LAT to detect delayed high-energyemission. The location must be computed in a shorttime—short compared to the time necessary torepoint. Several seconds is clearly adequate. Accuracyof about 20˚ is sufficient to ensure that thesource is within the large LAT FOV. Simple algorithmsare easily capable of meeting these requirements.BATSE is, in fact, currently providingonboard locations to OSSE that are this good. Ourbaseline algorithm for GBM neglects atmosphericand spacecraft scattering. Computing much betterlocations on board requires more memory and afaster CPU, increasing the costs significantly withlittle scientific return.The ground automated location is computed in nearreal time on the ground and is used to providecoordinates for rapid follow-up by ground-basedinstruments. The model for this capability is theGCN system currently in use on the CGRO. When aburst triggers the GBM, the data needed to computeaccurate locations (TRIGDATA) are immediatelytransmitted to the MOC. The burst location iscomputed automatically at the mission operationscenter (MOC), using a program provided by theGBM team, and sent electronically to any interestedobservers.The ground manual location is determined withhuman intervention after the data are received at theinstrument operations center (IOC). These locationsare used for the burst catalog and to optimize theLAT sensitivity to the burst. The model for thiscapability is the current production of burst locations,by the BATSE team, within a day or two ofthe occurrence of a burst. The operator optimallyselects the source and background intervals, resultingin the best available burst location.Statistical fluctuations in counts received by eachdetector result in location solutions differing fromthe true location. The effect of Poisson fluctuationson GBM location accuracy can be accurately simulatedby computing how much the rates change withazimuth and zenith angles, compared to statisticalfluctuations. The average angular uncertainty is9° for a 1 s burst with flux of 1 photon / cm 2 s, aflux value which is less than a factor of two abovethe trigger threshold. Figure 28 shows a color-codedmap of the angular resolution as a function of zenithFigure 28.—Map of statistical errors in location of a1-s burst with peak flux of1 photon cm –2 s –1 .38

and azimuth angles. At 10 photons/cm 2 –s, thestatistical error is 1.5°.Because statistical location errors arise from fluctuationsin the detected counts, all three stages oflocating GRB’s will have similar statistical errorswhen all of the burst data have been received. Forlong GRB’s, a succession of onboard and groundautomatedlocations will be produced, with decreasingstatistical errors as more counts are received.The ground manual locations will have optimumstatistical errors because of careful human selectionof background and source time intervals.In addition to statistical errors, the location processis also subject to systematic errors. These are moredifficult to estimate a priori. Our BATSE experienceis valuable, allowing us to estimate the systematicerrors of each GBM location method by comparisonto the systematic errors obtained in a similar approachto locating GRB’s with BATSE. The methodsand development of the BATSE locationalgorithm LOCBURST are described by Pendletonet al. (1999). The initial primitive version ofLOCBURST guides us in estimating the accuracyon the onboard locations. The initial LOCBURSTalgorithm, which is the first stage of the currentprogram, inverts the rates of the three detectors withthe highest rates to obtain the location. No complexfitting procedure is used, spectral effects are implementedusing tables indexed by hardness ratio, andatmospheric scattering is ignored. The simplicity ofthis approach makes it feasible to implement onboard at reasonable cost. This approach obtainssystematic errors below 6˚ for 50 percent of thelocations and below 12˚ for 80 percent of the locations(Pendleton, et al. 1999). Combining the12° systematic error which 80 percent of the locationswill meet with the 9° statistical error for aburst with an intensity 1.8 times the trigger threshold,a total error of 15° is obtained, easily meetingthe accuracy requirement for repointing the LAT.The error distribution of locations produced with thecurrent version of LOCBURST, as used to producethe 4Br catalog (Paciesas, et al. 1999) has beenderived (Briggs, et al. 1999b) by comparing a subsetof BATSE locations with the more accurate locationsdetermined with the IPN. The IPN uses arrivaltime information, at widely separated spacecraft, totriangulate the location. With two spacecraft, anarrow annulus of typical width, 10 arcmin, isobtained. With three or more spacecraft, the intersectingannuli give a small error box (Hurley, et al.1999). When intersecting annuli give a small errorbox, one has a direct measurement of the error inthe BATSE location. In more common cases, whereonly single annuli is available, the separationsbetween the BATSE locations and the annuli can beused to constrain the distribution of BATSE totallocation errors. This comparison of BATSE and IPNlocations has produced a two-term model forBATSE location errors. Most of the probability is ina core term with a small systematic error, while asmall fraction of the probability is in a tail term witha larger systematic error. We have also found thatthe systematic error values depend on the spectralresolution of the BATSE data used to determine thelocation.Locations based upon the BATSE 16-channel datatype (CONT) have 82 percent of the probability in acore with σ=1.67° and 18 percent of the probabilityin a tail with σ=5.4°. We believe that the dominatecauses of the systematic error are the circular errorbox approximation, the approximation of the spectrumas a power law, “edge” cases in which theburst is located 90° from the axis of a detector, andimperfections in the response model. Cases inwhich the location is at 90° to some detectors causethe location to be well determined perpendicular tothose detectors, but poorly determined parallel tothem, i.e., highly elliptical error boxes. Additionally,in these cases, the location becomes highly dependenton the power-law index, which controls theimportance of scattering from the Earth’s atmosphere.The GBM design and ground location algorithmswill improve on all of these systematic error causes.The most important improvement is the increasednumber of detectors: 12 NaI detectors viewing~3π steradians instead of 8 detectors viewing 4π.Over most of the FOV there will be enough detectorswith a non-edge view of the burst to constrainthe location in all directions, thereby producingnearly circular error boxes, e.g., bursts over 6.3steradians are within 70˚ of the axis of four or more39

detectors. With BATSE, bursts occasionally presentedlocalization problems when only two detectorsshowed significant flux. The BATSE responsemodel is deficient because the computing power of1990 required the detector response to be averagedover azimuth and because the spacecraft massmodel is too crude. Current computing power willallow a full mapping of the detector response andwe expect that an accurate mass model will bedeveloped for GLAST, which is a simpler spacecraftthan CGRO. We will also use more accurate spectralmodels than the power law currently used by theBATSE algorithm.The ground-automated locations will be based uponthe rates provided in the TRIGDATA datatype,which will probably have the same resolution as theBTIME data, i.e., 4 energy channels and 0.256-stemporal resolution. This automated algorithm willprobably use a power-law spectrum or a smalllibrary of more complex spectra, and will reportcircular error regions. This is quite similar toBATSE locations based upon 4-channel data, whichcurrently have a systematic error distribution with82 percent of the probability in a core of 2.7° and18 percent of the probability in a tail of 5.4°(Briggs, et al. 1999b). We estimate that the groundautomated algorithm will have a systematic errorbetween 2° and 3°, which will enable ground-basedobservations by specialized telescopes with wideFOV, as was done with ROTSE to detect the promptemission of 990123 in response to a BATSE automatedlocation (Akerlof, et al. 1999).The ground manual algorithm will have additionalimprovements, such as using 128 channel TTE datato obtain a more accurate spectral model, betterselection of the source interval and modeling of thebackground, and generation of elliptical error boxes.The detector response model will fully includeresponse variations over direction and energy, anddiffering responses of the detectors because oftheir positions on the instrument. This representsan improvement over the BATSE analysis of16-channel data, which has a systematic errordistribution with 82 percent of the probability in acore of 1.67° and 18 percent in a tail of 5.4°. Weexpect a modest reduction in the width of the coreand a substantial reduction in the fraction of theprobability in a wide tail, resulting in a typicalsystematic error of about 1.5° or less.2.5.4 False Trigger RejectionIf the capability is implemented to slew the spacecraftin response to a Burst Monitor trigger, it willbe important to avoid false triggers arising fromsolar flares, electron precipitation events, etc. Muchof the information required for trigger identificationis generated by the burst location algorithm. Electronprecipitation in the vicinity of the spacecraft ischaracterized by approximately equal rates inoppositely facing detectors and very poor fits to apoint source model. Location and high fluxes below25 keV efficiently identify solar flares. Electronprecipitation events, at a distance from the spacecraft,are somewhat more difficult to identify in realtime. They usually appear as long, slowly risinghumps in the background, near the latitude extremesof the orbit, peaking at slightly different times indifferent facing detectors, with computed locationsnear the horizon. The Burst Monitor will thereforeprovide a slew trigger only if the rate data indicate ahigh probability that the trigger is in fact a GRB.Spacecraft slews that are initiated by false triggerscould seriously impair the scientific return ofGLAST. Fortunately, our team has extensive experiencefrom BATSE in recognizing these false triggers,and the Burst Monitor detectors and triggerscheme are similar to those of BATSE. We cantherefore provide assurance that false burst triggerswill be not be a problem for GLAST.2.6 Spacecraft Interface2.6.1 MechanicalAn important feature of the mechanical design ofthe GBM is a high degree of flexibility in positioningthe components. This is crucial since the physicalcharacteristics of the main instrument and thespacecraft are not yet specified. This flexibility isachieved by using relatively small detectors, butincreasing their number to achieve sensitivity andredundancy, even with the main instrument significantlyreducing any one detector’s FOV. We assumethe conservative case that all Burst Monitor componentsmust be out of the LAT FOV, and only on thetwo sides not occupied by the solar panels. Even40

Table 10.—Power.Component Power/Unit Total PowerBGO detector 0.6 1.2NaI 0.3 3.6DPU 4 4LVPS Losses 4 4HVPS Losses 5 5Figure 29.—Detector placement concept.Detector placement is flexible and will becoordinated with the spacecraft contractor.with these constraints, the Burst Monitor componentsare easily accommodated.Figure 29 (same as 11) shows one possibility formounting the detectors. Two BGO detectors arepositioned on each side of the LAT, providing fullsky coverage. The NaI detectors are mounted infour banks of three, with each bank at a differentazimuth, and at different zenith angles. Each detector,of course, is blocked by the LAT over a largefraction of the sky. It is important to note thatalmost any mounting arrangement is satisfactory aslong as the following conditions are met: 1) TwoBGO detectors must be on opposite sides of theLAT, 2) all detectors are unobstructed in the +Zdirection, and 3) the NaI detector normals mustsample a wide range of azimuth and elevationangles. If the Burst Monitor is descoped to a smallernumber of NaI detectors, their viewing angles willbecome somewhat more constrained to assureadequate sky coverage.Table 9 provides a summary of the mass and size ofthe various components. The total mass is estimatedat 54.5 kg with 20 percent contingency on all itemsexcept the crystal mass. Most of the mass is in thedetector assemblies, and is known quite accurately,based on measurements of BATSE flight PMTassemblies. The electronic boxes are small andpresent no mounting problems. Mounting structure,to be provided by the spacecraft contractor, is notincluded.Table 9.—Mass and size.Component Mass (kg) Size (cm) Number Total Mass (kg)BGO crystal 11.47 12.5 cm dia. × 12.5 cm 2 22.94BGO housing 0.28 2 mm thick 2 0.56NaI Crystal 0.59 12.5 cm dia. × 1.25 cm 12 7.08NaI Housing 0.028 2 mm thick 12 0.34PMT (inc. housing) 0.5 12.5 cm dia. 16 8.0DPU 1.0 10 cm × 10 cm × 10 cm 1 1.0HVPS 1.5 TBD 1 1.5LVPS 1.5 TBD 2 3Cables 6 n/a n/a 6Contingency 20% on all items except crystals 4.1Total 54.541

2.6.2 ElectricalThe electrical interface presents no complications.The Burst Monitor requires unregulated power,clock, command, and telemetry lines. Communicationwith the LAT, if desired, can be achievedthrough the spacecraft using data in the telemetrypacket, or by a direct interface to the LAT. Thepower requirements are summarized in table 10.The total power is 17.8 watts, without contingency.The required telemetry rate is normally 4 kbps,increasing to 9 kbs during bursts, as was describedin more detail in section 2.2.2.6.3 ThermalThe spacecraft will provide thermal control andinsulation. The MSFC Engineering Directorate willuse thermal radiation analyzer system (TRASYS)and system improved differencing analyzer(SINDA) computer models to support the GRBMthermal and thermal/vacuum testing. Preliminaryanalysis shows the following thermal requirements.Detectors0 to 20 °C operational, –10 to 30 °C storage, stable1 °C over one orbit. The stability requirement isderived from the need for short term gain stability.The AGC can maintain stability over longer times.Electronic Boxes0 to 50 °C operational, –30 to 90 °C storage.2.7 Ground System and Operations2.7.1 Requirements and Operations ConceptThe ground system for the GBM is a hardware andsoftware system that accepts instrument data fromthe GLAST MOC and produces scientific data sets.The system also monitors instrument performanceand safety.The GBM IOC receives data from the instrumentvia the MOC, and converts it into standard low-leveldata products for distribution to the GLAST SOC.The system data flow for GBM is shown infigure 30.Figure 30.—System Data Flow for the GLAST Burst Monitor.42

Table 11.—Responsibilities of the GBM IOC.Responsibilities of the GBM IOCNominal instrument operations and monitoringInstrument calibrationProduction and maintenance of operations softwareProduction of data analysis softwareProduction of low-level standard data products useable by the communityVerification of flight dataProcessing of data to support the IPI team’s investigationsSupport of the MOC and Guest Observer FacilityThe IOC is responsible for monitoring the instrumentand generating instrument commands. Toavoid duplication and reduce costs for this secondaryGLAST instrument, we plan to have taskscompleted by the GLAST SOC whenever reasonable.Specific responsibilities of the Burst MonitorIOC are listed in table 11.The GBM operations concept utilizes an approachthat minimizes the costs and risks of computerhardware and software development. The sameinstrument ground support equipment (IGSE) andsoftware developed and employed for instrumentintegration and testing (I&T) is also used for onorbitnominal operations tasks such as instrumentmonitoring, state of health verification, and commanding.By using the same IGSE hardware andsoftware during instrument I&T and normal flightoperations, development costs are lower, and risk isminimized. Software development for instrumentmonitoring and for data reduction and analysis isbased on the prior flight data operations experienceof the BATSE and COMPTEL instrument teams onthe CGRO. Software developed for analysis ofBATSE GRB data serves as the basis for the GBMdata analysis software.2.7.2 Instrument Ground Support EquipmentThe GBM ground system provides instrumentmonitoring and commanding, data reduction andprocessing, and data distribution. The groundsystem consists of three components: Calibrationhardware and software (see section 2.3.1), groundsupport equipment, and operations equipment.Calibration hardware accepts raw analog detectoroutput for input into the calibration software that isused to generate the detector response function. Therequired hardware is available as standard commercial,off-the-shelf products. Ground support equipment,consisting of a PC or workstation andassociated peripherals and software, is used duringI&T to accept output from the instrument DPU orspacecraft simulator. Ground support softwareprovides data verification and analyses of detectorperformance. During nominal flight operations, theIGSE is connected to the MOC network, and receivesdaily operations data. The housekeeping andinstrument status data are extracted and formatted,for visual verification of instrument performance byoperations personnel. The science data are forwardedto the operations equipment for reduction,analysis, and distribution. Operations equipmentconsists of several PC’s or workstations and associatedperipherals and software that receive anddisplay data from the IGSE.Each of the major components of the GBM groundsystem are tested prior to verification of the fullground system. After test readiness of the componentsis validated, the ground system is activated forend-to-end testing of data flow and componentcompatibility, using the IGSE in its instrumentintegration mode, accepting data from the instrumentDPU or spacecraft simulator.Continuity of the ground system software tools isthe central feature of our operations software developmentapproach. This is a low risk, cost-efficientapproach. The same institution that has the responsibilityfor using the IGSE for instrument operationsis responsible for its development from the earliestphase of the project. IGSE software development43

egins with integration with the DPU or spacecraftsimulator. Software development continues throughthe instrument I&T phases and the integratedspacecraft test phase. During those phases, theoperator tools, for commanding state of health andconfiguration verifications, are refined based on theaccumulated experience gained while meeting therequirements of instrument testing and calibration.The same IGSE operator tools are used for integratedmission system development, including endto-endtesting. Prior to and during this phase, thetools to support network communications, planning,stored commanding, anomaly response, and missionsystem administration are developed and verified.The GBM instrument simulator (the instrumentengineering model) is maintained and operated atthe IOC. It is used for testing operational softwareupdates and DPU performance prior to uplink to theflight instrument. It is also available for hardwareand software anomaly investigations.2.7.3 Nominal Instrument OperationsThe IOC has the responsibility for nominal operationof the GBM instrument. Table 12 lists the toplevel ground support software tasks and summarizesthe functions of each module. Software modules tosupport these functions are fully developed, testedand validated prior to launch. The ground supporthardware is a PC or workstation connected to theGLAST MOC network.Most instrument commanding is automated andaccomplished via preplanned or stored commands.The use of the real time commanding capabilities isexpected to be rare after initial on-orbit instrumentactivation, checkout, and tuning. All instrumentcommands and flight software updates, with certainexceptions for anomaly response, are processedthrough the Burst Monitor IGSE at the IOC, andtransferred to the MOC for transmission to thespacecraft. Exceptions include instrument poweron/off and related safe hold commands. These andother related spacecraft safety commands areprepared and uplinked by the MOC using predefinedprocedures.Command data verification is accomplished by preuplinkscreening of commands through a configurationcontrolled database. All flight software updatesare verified prior to uplink using the instrumentsimulator. Instrument status, state of health, andfunctional verifications are accomplished using theIOC IGSE data reduction tools. These tasks areautomated and the output is reviewed routinely bythe instrument operators. Anomalous conditions arethe subject of alerts to the appropriate personnel,further study of the IGSE data, and appropriateresponse and reporting.2.7.4 Instrument Ground CalibrationCalibrations with sources and Monte Carlo simulationswill establish the baseline detectorpreformance before launch, as described in section2.3.1. On orbit, the detector gains are maintainedTable 12.—Instrument operations software tasks.TaskNetwork CommunicationsState of Health VerificationCommandingAdministrativeAnomaly ResponseData TransferDetector StatusReduction and AnalysisArchivingFunction SummaryUse established common network protocols for two-way communications with theMOC; security monitoring; problem alertsHousekeeping monitors; instrument parameter diagnostics; limit alarms and other alertsCommand database; real-time commands; stored commands; command verificationAdministrative messages; operations activity records; anomaly recordsAutomated alerts; responses for predefined telemetry anomaliesFTP transfer of raw data from the MOCFormat and display housekeeping data of detector operational parameters(temperature, voltage, etc)Data reduction and analysis; delivery to SOCArchive raw instrument data44

using the AGC technique described in section 2.4.Ground software verifies the AGC operation anddetector resolution by maintaining an archive ofposition and width of the 511 keV annihilation line,which will be readily visible in both the NaI andBGO detectors. Additional lines in the detectorspectra allow verification of the channel to energynonlinearity.2.7.5 Instrument MonitoringInstrument status monitoring is an automatedfeature of the IGSE software. Details of theinstrument’s commanded status are periodicallytelemetered to ground and compared by the IGSEwith the expected parameter values. Alarms andautomated operator alerts are generated whenselected differences between the expected and thetelemetered status are registered. These records arelogged for subsequent operator analysis andanomaly reporting.Operations personnel generate daily summaryreports, charts, and graphs of instrument status andperformance for visual inspection. Data validation isaccomplished through generation of daily orbit plotsand trigger event data for visual inspection foranomalies.2.7.6 Operations SoftwareDaily operations with IGSE at the IOC includeprocessing of instrument housekeeping data andlimited processing of scientific data, for quick-lookinstrument functional verification, record keeping,and timely response to anomalies, with the goal ofminimizing loss of observing time.Operations software reads raw instrument datareceived from the MOC and extracts housekeepingand science data. Science data are reduced andformatted into flexible image transport system(FITS) format, in preparation for delivery to theSOC. The software creates sets of daily standardplots of both housekeeping and science data. Standardplots include housekeeping instrument parameters(temperatures, voltages, etc.), orbital plots offull-day detector rates (BTIME and BSPEC datatypes) for inspection and analysis, and detailed plotsof GBM trigger data.2.7.7 Data Analysis SoftwareAnalysis of event light curves and count spectra isperformed using software based on mature datadisplay and spectral fitting software, developed foruse in the analysis of GRB data obtained withBATSE. Time-resolved spectroscopy is performedusing the program WINGSPAN, developed by theBATSE science team, for multi-detector timeresolvedspectral analysis. This robust, forwardfolding, spectral fitting package contains the capabilityfor simultaneous spectral fits to data frommultiple detectors. This capability is required forsupport of the GBM IPI team’s science investigations(see section 1.5). Some of the functionsimplemented in this software, such as data andresponse readers, will require modification for theGBM datasets.An event location algorithm will be developed foron-ground processing of trigger event locations onthe sky. The premise for computing the location of agiven event observed with the GBM is based on therelative counting rates, due to the differing projectedarea in those detectors that observe the event. Thistechnique has been successfully used to computelocations of gamma-ray events observed with theeight detector module BATSE system. Additionaldetails are provided in section 2.5.3.The GBM triggered event data are formatted intothe binary platform-independent FITS data format.The data analysis software is based on this dataformat. Raw science instrument data and FITS datafiles are archived to digital video disk (DVD). DVDis a cost effective, high volume, permanent storagemedium, similar to a compact disk (CD), but with amuch larger storage capacity see section 2.7.10.2.7.8 Data ProductsScience data reduction is performed at the BurstMonitor IOC in Huntsville, AL. These tasks prepareflight data for delivery to the SOC, and process datafor analysis by the IPI team in accordance with theproposed science investigations. Reduced sciencedata are transferred to the SOC in the portablebinary FITS format. After a 30–60 day in-orbitcheckout, all data products are delivered to the SOCin useable form within 6 weeks of data receipt, withthe exception of the published GBM Burst Catalog,45

Table 13.—Delivery data products.ProductBackground Data (BTIME, BSPEC)Burst Data (TrigData, TTE)BTIME DisplayBSPEC DisplaySkymapCalibration and ResponseEvent CatalogFITS ToolsGBM User’s ManualGBM Instrument StatusContent SummaryCount spectra in continuous coverage mode; data delivered to theSOC as files in FITS format containing continuous coverage backgroundspectra; each file contains a 24-hr daily datasetCount spectra at high time and energy resolution for triggered events inFITS format; also ±2,000 s of BTIME and BSPEC background spectraSoftware to read Burst Data event files and display lightcurves as a functionof energySoftware to read Burst Data event files and display time integrated countspectraAngular distribution of triggered GRB’sCombine calibration and simulation data to generate instrument responsefunction for all triggered events; response functions are formatted into FITS files, and asoftware response reader is providedCatalog containing parameters of interest for all GBM-triggered events,including parameters such as location, duration, and intensityData readers for all GBM FITS filesUser’s manual for data products; software documentationDocumentation of instrument configuration; operations summary logwhich will normally be released at the end of every2 years of observation. All data will also be continuouslyavailable via the World Wide Web, with newdata added weekly. During the first 12 months ofobservations, the instrument may not be completelycalibrated, and thus any data made available will besubject to later revision. Table 13 lists the dataproducts delivered to the SOC.After the first 12 months, the GLAST observingprogram will be based on a guest observer program(GOP). Data gathered for a selected investigationwill be verified by the guest observer (GO). After a3-month verification phase, the data are delivered tothe SOC.46The FITS format, selected for delivery data products,is a data format designed to provide a meansfor convenient exchange of astronomical databetween installations whose standard internalformats and hardware differ. The FITS standard isthe format adopted by the astronomical communityfor data interchange and archival storage. The FITSsupport office, at NASA’s Goddard Space FlightCenter (GSFC), is responsible for documenting theFITS standard defined by NASA’s Science Office ofStandards and Technology, participating in itsevolution, and advising NASA astrophysics missionson how to present their data in FITS format.Although this format allows for transparent dataaccess from all popular computer platforms, usersmust develop or obtain separate software to readand display the data from the FITS file. The IOCprovides software optimized for GLAST datasets.Investigators can also use software from otheravailable packages such as FTOOLS and XSPECfor GBM data analysis. FITS readers for GBMdatasets are provided as part of the standard delivereddata products.2.7.9 Verification of Flight DataFlight data are verified by daily operations personneland Instrument Principal Investigator (IPI)science team members. Initial verification is performedvia daily summary charts and graphs ofinstrument performance and safety parameters, aswell as raw data that are produced by operationspersonnel as part of GBM daily operations. Thesereports and data plots are examined for indicationsof anomalies.Analysis of GBM data by the IPI science teamprovides secondary data verification. As data arereduced and analyzed by the IPI science team,errors and corrections are propagated to the IOCdaily operations team, for modification of proceduresthat format data for delivery to the SOC.

2.7.10 Hardware RequirementsCalibration and validation equipment consists of acommercial system interface to the analog output ofthe GBM detectors. Spectra are accumulated andanalyzed for detector performance verification. TheGBM IGSE consists of one PC or workstation andassociated peripherals, with interface software to theDPU or spacecraft simulator, and the operationsequipment. Operations equipment consists of two tothree PC’s or workstations, for daily processing ofraw science data (approximately 100 Mb/day), andassociated peripherals, such as disk storage, backup,printers, etc.Expected data rates, including specifications for themaximum rate and size of events, represent importantdesign requirements imposed on the DPU byinstrument simulations. These data rates and instrumentdead time requirements determine the appropriatedata buffer size and processing architecture.The nominal GLAST context instrument data rateof 10 kbs=40 Gb/yr results in a raw data storagecapacity of 200 Gb for a projected 5-year mission.Raw data are archived to nonvolatile storage mediumDVD. Currently available single sided, singlelayer DVD disks have a capacity of 4.7 Gb. Doublesided, dual layer DVD, not yet available, have acapacity of 17 Gb. Twelve 17-Gb disks will hold theprojected 5-year GBM dataset. This storage estimateis conservative, since the data rate for theGBM is 4 kbps nominal, and 9 kbps in triggermode.2.7.11 Staffing PlansThe IOC staffing plan is severely constrained bythe Phase E funding profile. Consequently, we willpredominantly use low-cost UAH student supportfor routine operation tasks. Table 14 presents thefull-time equivalent (FTE) manpower for operationsduring each of the 5 years of the nominal mission.Covered tasks include data receipt, instrumenthealth and safety, archiving, mission operations,Table 14.—Phase E staffing profile.Year of Mission 1 2 3 4 5Scientist 1.7 1.4 1.0 0.5 0.5Student 2.3 2.0 1.7 1.3 1.3and command generation and transmission. Scientificanalysis is not included.2.8 Science Team Roles andResponsibilitiesThe science team selected for the GBM has extensiveexperience with scintillation detector systems,spacecraft instrument development and analysis ofgamma-ray data. The team consists of scientistswho designed, built, and operate BATSE andCOMPTEL on CGRO and who are currently developingSPI on INTEGRAL. The extensive experienceand outstanding track record of these teammembers assures success of the GBM. Additionaldetails of the investigator’s qualifications are presentedin the resumes (Appendix A).Table 15 presents a list of the co-investigators, witha summary of their responsibilities on the BurstMonitor, and previous relevant experience.2.9 Descope OptionsThe primary descope option for the Burst Monitor isa reduction in the number of NaI detectors. Table 16summarizes the loss of scientific capability as thenumber of detectors is reduced. The relative sensitivityin the table is the approximate area, in unitsof one detector, for a burst near the zenith. In eachdescoped case, we have oriented the detectors to tryto obtain burst locations over as wide a FOV aspossible. The locations require observation by aminimum of three detectors. The burst triggerrequires observation by at least two detectors. Theminimum science mission is reached when there areonly two NaI detectors, at which point the sciencegoal of burst locations is abandoned. The minimummission does retain the most important goal of theBurst Monitor, which is time resolved broadbandspectral response for GRB’s. Burst triggers remainpossible, although the sensitivity is significantlydegraded. All descope options include two BGOdetectors.Reduction in the number of NaI detectors reducesrisks to all constrained resources: Cost, schedule,47

Table 15.—GMB Science TeamName & InstitutionDr. Charles Meegan, PINASA/MSFCDr. Giselher Lichti, Co-PIMPEDr. Michael BriggsUAHDr. Roland DiehlMPEDr. Gerald FishmanNASA/MSFCDr. Robert GeorgiiMPEDr. Andreas von KienlinMPEDr. Marc KippenUAHDr. Robert MallozziUAHDr. William PaciesasUAHDr. Robert PreeceUAHDr. Prof. Volker SchoenfelderMPEResponsibilityScientific requirements andoversight, NASA point ofcontactLeadership of MPE effortFlight software, MissionOperations DirectorSoftware and data analysis atMPEDetector performancespecificationsDetector design and massmodelingDetector electronics, detectorperformance test, calibrationSimulations, detector responsematrices, DPU specifications,I&T proceduresOperations Software; Ops andDA hardware, Education andPublic OutreachInterface with LAT teamData Analysis requirementsand softwareMPE coordination with DLRand MSFCExperiencePerformance of balloon flights; BATSE: instrument development,flight software, data analysis, operations, Burst Teamleader; HST Asst. Project. Scientist; ASTRO-2 MissionScientistPerformance of balloon flights; COS-B: leader of Fast-Routine Facility at ESOC; COMPTEL: project manager ofMPG’s hardware development; INTEGRAL: local projectmanagerBATSE: spectral analysis, line search, burst isotropy &location accuracy, Spectroscopy Detector hardwaredevelopmentCOMPTEL: Chairman of Data Reduction Group; InstrumentCalibration and Data Analysis Method development;COMPTEL/OSSE/SMM Spectral Analysis INTEGRAL/SPI:Chairman of Data Analysis GroupPerformance of balloon flights; Gamma-Ray AstronomyTeam Leader at MSFC; Spacelab NRM PI; BATSE: PI,instrument developmentINTEGRAL/SPI: BGO-Shield design and test; SPI simulations,calibration, software development; ACS performance tests;COMPTEL data analysisINTEGRAL/SPI: ACS electronics, ACS performance tests,ACS burst detection system, SPI calibration; Development ofLow Temperature DetectorsBATSE: Rapid Burst Response Team Leader, data analysis,simulations; COMPTEL data analysisBATSE: software development, data analysis, Web sitedevelopment, simulationsPerformance of balloon flights; BATSE instrument development,BATSE Spectroscopy Team leaderBATSE: software development, spectral analysis andinterpretation; theory of gamma-ray emission mechanismsPI of Compton Telescope Balloon Program at MPE; PI ofCOMPTEL aboard CGRO; Co-PI of spectrometer INTEGRAL(SPI)mass, volume, power, and telemetry. Since thesedetectors are provided by MPE at no cost, theNASA cost is reduced primarily by transferringsome tasks from MSFC to MPE holding fixed theMPE costs. Two specific descope cases are consideredbelow.Descope to Eight NaI Detectors.With a reduction in the number of NaI detectorsfrom 12 to 8, the Burst Monitor sensitivity and FOVare degraded, but the scientific return is still goodand the scientific goals are not severely compromised.With this option, MPE would assume re-48

Table 16.—Scientific performance for several descope options.NaI Detectors Burst Locations FOV Burst Trigger FOV Effective FOV Relative Sensitivity(steradians) (steradians) (steradians)12 11.55 12.57 8.61 2.88 8.98 11.47 6.71 2.16 8.24 10.59 5.16 0.92 0 0 * 0.5* We define the effective FOV in terms of the trigger sensitivity, which for two detectors on opposite sides of the LAT is 0 sr. If defined asprojected area, the effective FOV for two detectors is ~2π sr.sponsibility for the cable harness, which wouldprobably not exceed the cost savings of procuringfour fewer detectors. MSFC would realize costsavings of $102.5k for this descope, occurring atany time before Critical Design Review (CDR). Thetotal mass reduces to approximately 46 kg.Descope to Two NaI Detectors.This option is a descope to the performance floor. Inthis case, MPE would accept responsibility for thecable harness and for performing the instrumentI&T, including thermal vacuum tests. MPE wouldrealize offsetting cost savings in the detector andHVPS procurements, and in the reduced preflightcalibration effort. MSFC would retain responsibilityfor test requirements, plans, and procedures. Therewould be additional costs for work on these documents,as a result of transferring implementationresponsibilities to MPE. Significant cost savings arerealized, not only in the manpower and materials forI&T, but also in the DPU and the flight and groundsoftware efforts, since burst locations are not calculated.Cost savings are presented in table 17 and arecalculated for two cases: descope at PDR andTable 17.—NASA cost savings for descope toperformance floor.EffortSaving ($k)(PDR/CDR)1. Cable harness construction 102.5/102.52. I&T performance 90/903. I&T hardware 35/354. Thermal-Vac test, inc. fixtures 32/325. DPU descope 16/166. Flight Software Reductions 51/267. Data Analysis Software Reductions 52/268. Added Documentation 0/–25TOTAL 378.5/302.5descope at CDR. All I&T and harness constructioncosts (items 1–4) are incurred after CDR and aretherefore the same for both cases. I&T performancesavings (item 2) are based on one FTE contractorcost. Savings for items 1, 3, and 4 are taken fromthe cost breakdown in table B–4, Volume 2. DPUdescope savings (item 5) represents only the reducedparts, therefore it is a conservative lowerlimit, and is valid for descoping at any time beforethe request for proposal (RFP) for the DPU. Softwaresavings (items 6–7) are based on UAH researchscientist manpower costs and reflect therequirements and design effort that occur prior toPDR and CDR. The descoped flight software effortis conservatively estimated at 80 percent of the fullyscoped effort and the descoped data analysis softwareeffort is conservatively estimated at 80 percentof the fully scoped effort. The additional documentationcost represents one FTE of civil servicemanpower for descope, occurring after CDR.If the descope to the performance floor occurs earlyenough, we will increase the thickness of the NaIdetectors to regain some of the lost sensitivity. Sinceburst locations will not be determined in the fulldescope case, a more isotropic response for thesedetectors is desirable.The total cost savings to NASA is $378.5k, for adescope to the performance floor at PDR, and$302.5k if descope occurs at CDR. The massreduces to approximately 30–40 kg, depending onthe revised thickness of the NaI detectors.49

3.0 Technical Approach3.1 OverviewThe Burst Monitor project will produce an instrumentwith excellent scientific performance and lowrisk, using flight proven hardware with simpleinterfaces. Most of the hardware will be procuredthrough competitive bid or supplied by MPE atno cost to NASA. No technology development isrequired. All technology needed to produce thisinstrument is similar to the BATSE experimentwhich the developers of this proposal designed,produced, and still operate. A schedule forinstrument development is provided in the factsheet. End items to be provided are:• Flight GRBM instrument• Electrical and mechanical ground supportEquipment for use in integration andoperations.• IOC and equipment• Flight software and documentation• Instrument ground operations command,control, housekeeping software anddocumentation• Data analysis and archiving software anddocumentation3.2 Fabrication/Procurement Plans3.2.1 Sodium Iodide and Bismuth GermanateDetectorsMPE provides these detectors at no cost to NASA.The vendor of the BGO will be Cristmatec. Performancerequirements will be developed by thescience team and design specifications will bedeveloped by the MPE project team during phase B.MPE procures flight qualified detectors by contractadministered through Deutsches Zentrum fuerLuft- und Raumfahr (DLR). MPE retains technicaldirection over the production. Flight qualifieddetector assemblies include PMT’s, high-voltagebleeder strings, and preamplifiers.3.2.2 Power Supplies.MPE provides both HV and LV power supplies atno cost to NASA. These power supplies are similiarto ones that have been previously designed andflown by MPE.3.2.4 Data Processing Unit.MSFC will procure the DPU by competitive bid.The MPE/U.S. science team is working with theEngineering Directorate at MSFC to develop performancespecifications. At least one vendor, AmptekIncorporated, can provide a flight qualified unit, atthe price used for our cost estimates, by makingmodifications to their CEASE radiation monitorsystem. The MSFC engineering team providestechnical, cost, and schedule oversight for thisprocurement.3.2.5 CablesThe MSFC project works with the GSFC projectoffice, the spacecraft contractor, and the maininstrument provider to establish interface controldocuments (ICD’s) for the Burst Monitor with thespacecraft and the main instrument. After detectorplacement and mounting has been determined, thespacecraft contractor specifies flight cable routing.Flight cables are to be fabricated at MSFC. We haveproduced cable harnesses for flight equipmentincluding the lightning imaging sensor, opticaltransient detector, and the solar x-ray imager.3.2.6 SoftwareSoftware development on the GBM project employsmany of the same people as used for BATSE.BATSE software for data acquisition, flight commandand control, and data analysis is still in use.Algorithms for data acquisition and processing foronboard tasks and for data analysis and archivingare used for GBM. The current software providesprototypes for development of GBM software usingmodern software applications and computers. Aformal development process is used for softwaredevelopment with milestones indicated on theschedule (figure 3 in the Management Section).In-process technical management uses softwarestatus walkthroughs on a weekly basis. The softwaredeveloper discusses status and approach withthe science team to assure that coding reflectsrequirements.50

3.3 Calibration PlanMPE calibrates the detectors. The calibrations aresufficient to verify computer simulations of thedetector response. The calibration plan is developedby MPE in consultation with the science teamat MSFC and UAH. All flight detectors will beexposed to radioactive sources to acquire spectra asshown in tables 5 and 6. These spectra will be usedto fine tune the detector response matrices obtainedby Monte Carlo simulations. UAH co-investigators,who have extensive experience in this area, willperform the simulations. A spare flight DPU card,used for data acquisition from the detectors, issupplied to MPE for use in this calibration. Thiscard is returned to MSFC and is used to compareresults of system calibrations using the flight DPUwith MPE results. In-flight calibration and validationuses detector response from astronomicalsources, as is done for BATSE.3.4 Assembly, Integration & Test3.4.1 OverviewThe separate hardware components are shipped toMSFC for instrument assembly and test. The detectors,DPU, and power supplies are interconnectedusing flight cables. The DPU is connected to thespacecraft simulator, which is connected to theelectrical ground support equipment (EGSE).3.4.2 Integration and Test ProcedureGenerationThe subsystem integration control procedure will beprepared by SD71. All procedures shall be submittedto the responsible organizations for review andsignature approval. An integration/test readinessreview (ITRR) will be conducted by SD71 prior tostarting the GBM I&T. The ITRR chairman willissue minutes of the review and verify that all of the“constraints to test,” identified during the ITRR, areclosed prior to starting test operations. The SD71lead engineer will be the test conductor for GBMI&T control procedures. Generated documentationwill include the following:• Original signature procedures• “As-run” test procedures• TDR/DR and TDR log• Data generated during testing• Test report3.4.3 Integration and Test RequirementsGBM subsystem integration activities will beperformed in MSFC building 4481, in a class 10kclean room. All clean room operations will beperformed in accordance with MSFC–STD–246“MSFC Design/Operational Criteria of ControlledEnvironment Areas”. The handling and test operationsfor any hardware classified as electrostaticdischarge (ESD) sensitive will be in accordancewith MSFC–RQMT–2918. This ESD designationwill be specified on the hardware drawings, packinglists, inspection reports, or paperwork accompanyingthe hardware. In addition, an ESD sensitive testarticle will be labeled with a sensitive electronicdevice symbol. All integration and test operationswill be monitored and accepted by the QualityAssurance Office. All nonconformances will bedocumented on MSFC Form 460 in accordancewith MPG 8730.3. The TDR/DR troubleshootingand dispositions shall be in accordance withMSFC–P13.1.3.4.4 Functional TestingFunctional tests will be performed to verify gammarayand housekeeping data from each detector,commands, telemetry, and DPU flight software.Separate hardware components will be shipped toMSFC for instrument assembly and test. Detectorswill be mounted on flight-like mounting structuresand the DPU and power supplies will be interconnectedusing flight cables. The DPU will be connectedto the spacecraft simulator, which will beconnected to the EGSE.The following tests will be performed:1. Functional tests to verify gamma-ray andhousekeeping data from each detector, commands,telemetry, and DPU flight software.2. Thermal vacuum test: The GBM will be exposedto 3–4 days of thermal vacuum functionaltests in chamber V7, at MSFC’s environmentaltest facility. The environmental test facility providesfacilities and engineering/technical support51

for performing environmental testing of spacesystems and components. The facility was firstorganized in the early 1960’s and has providedsupport to all major NASA projects developed atMSFC since that time. Chamber V7 is used forEarth orbital and deep space simulations to testperformance of space systems and subsystems.The chamber is equipped with a shroud that isliquid nitrogen cooled. Heat lamps are used tosimulate the radiance of the Sun or reflected heatfrom the Earth. The chamber is horizontallyoriented with internal dimensions of 8 ft. diameterand 10 ft length. Typical pressures that can beobtained in this chamber are 1×10–6 Torr andlower. The system is equipped with two rotary,oil-sealed roughing pumps and two 24-inch cryopumps. There are six 6-inch ports available on thechamber for connection of instrumentation cabling,power cabling, mechanical feedthroughs,and fluid feedthroughs. A data acquisition system,known as PACRATS, is available for monitoringand recording up to 190 channels of temperature,pressure, and voltage data.3. Vibration Test: GBM experiments will bequalification and acceptance tested to meet thedynamic environment on one of four Unholtz-Dickie T4000 shakers within the MSFC ED27vibration laboratory. Each of the tables is configuredto run vibration tests from 5 Hz to 2,000 Hzand is capable of 40,000-lb force. All standardvibration tests can be generated—random, sine,sine on random, random on random, classicalshock, and rocket separation shock. The vibrationcontrol systems handle 16 input channels andmore can be made available.Burst Monitor experiments will be mounted to theshaker table through the use of a flat aluminuminterface plate between the shaker and the detectorsor CPU’s. A total of 16 tests will be run(14 detector tests and 2 CPU tests). Both of thehigh-energy BGO detectors will receive three axisrandom vibration tests. Each of the 12 low-energydetectors will receive three axis random vibrationtests as well.4. Pyro Test: Each of the high-energy BGOdetectors, low-energy detectors, and the CPU’swill receive pyroshock testing to simulate theDelta launch vehicle separation loads in theMSFC pyroshock test laboratory, also located inbuilding 4619. The aluminum interface plate usedfor vibration testing will also be utilized forpyroshock testing. The shock spectra environmentwill be duplicated to laboratory best tolerances.There will be a total of 16 pyroshock tests.Following I&T the GBM hardware will be shippedto the GLAST spacecraft integration contractor.3.5 Spacecraft IntegrationHardware components will be delivered to thespacecraft contractor, who will be responsible formounting the Burst Monitor hardware to the spacecraft,using procedures jointly developed by GSFC,MPE, and MSFC in conjunction with the spacecraftand main telescope contractors. Functional tests willbe performed after mounting to the spacecraft.MPE, MSFC and UAH will support the spacecraftintegration tests.3.6 Quality Assurance and SafetyThe Safety and Mission Assurance (S&MA) Officeat MSFC oversees the GBM tasks and suppliessafety and mission assurance engineering andinspection according to in-place ISO 9001 certifiedpolicies and procedures.3.7 PartsResponsibility for parts resides with the Parts andPackaging Group of MSFC’s Engineering Directorate.Only flight qualified parts will be used in theBurst Monitor. An electronic electrical electromechanical(EEE) parts plan will be part of the BurstMonitor Quality Assurance approach.52

Uncertainty in DPU costs, sincerequirements are preliminaryTable 18.—Burst monitor risk mitigation.Burst Monitor Risk Mitigation TableInstrument Risk Mitigation Risk Level Responsible PartyDesign-to-cost; reduce redundancy requirements;descope number of detectors6PISchedule recovery from a hardwarefailure during integration and test.a. The schedule has slack after integration andtest to recover.b. Sufficient spares will be available for quickchange out of hardware.4Project ManagerU.S.—Germany (MPE) interface:a. Will export control issues be animpediment?b. Are there any problems with MSFC notdirectly managing the MPE effort?a. Following NASA export control guidelinesb. Team communication, insight into MPEreviews. Previous team experience.c. Control of Interface Control documents withMPE3Project ManagerPhase E performance at funding levelsspecified in AO.Consider transferring some responsibilities,such as data archiving, to MOC; consider largerrole for MPE7PI, Co-PIEMI in relatively long signal and HVcables.Early analysis; enhanced shielding; decentralizeHV2MSFC and MPE SystemEngineersRisk level is defined from 1 to 10 with 10 being the highest and 1 being the lowest.3.8 ISO 9001MSFC is ISO 9001 certified and employs ISO 9001standards to all flight projects..3.9 Risks and Risk Mitigation.The GBM should be considered relatively low riskdue to maturity of the technology and in considerationof the similiar experience of developing asimiliar instrument (BATSE) by the assigned personnelat MSFC.The items in table 18 have been identified as possiblerisk areas.3.10 ReviewsSection 9.0 in the Management Volume details theBurst Monitor project and program reviews.The Burst Monitor team will conduct a series ofinternal design reviews and participate in the NASAmandated formal design reviews. All members ofthe Burst Monitor development team will participatein these reviews. The formal design reviewswill be coordinated with the GSFC project office.MSFC will coordinate and prepare the reviewpackage to be submitted to GSFC prior to thescheduled review. The Burst Monitor team willparticipate in the following reviews:Quarterly GSFC Reviews (per GLAST schedule)–Annual Independent Assessment Reviews (IAR):• System requirements review (SRR)–June 1, 2000• Preliminary design review (PDR)–August 3, 2001• Nonadvocate review (NAR)–August 17, 2001• Critical design review (CDR)–August 15, 2002• Pre-environmental review (PER)–1 month prior to start of environmental tests• Preship review (PSR)–1 month prior to delivery to spacecraftcontractor53

4.0 Phase A/B DevelopmentTechnical Definition PlanThis section details the development of the instrumentdesign prior to the PDR. This includes allscientific and engineering trade studies, calculations,tests and analyses, as well as the preliminaryengineering design efforts for the flight instrument.4.1 Preliminary Design ProcessDuring Phase A/B, the scientific performance tradestudies will include, but not be limited to:• optimum placement of the detectors on theGLAST fight system to meet the scientificobjectives• NaI detector entrance window, seal, and thermalcovering selection.• Detailed calibration plan, preflight and onorbit• Onboard trigger and location algorithmsAnalyses will include the following:• Detailed detector response simulations• Modeling the detector background inorbit• Estimation of dead-time effects for strongGRB’s• Simulations of joint GRB spectral fits forGRB’s with the GLAST LAT and other contemporaneousGRB instrumentsEngineering trade studies will include:• PMT circuit design, including pre-amp, foroptimum resolution and minimizingdeadtime• Evaluation of alternate mounting designs forthe detectors (to be performed jointly with thespacecraft contractor).Phase A/B efforts related to procurement of majorpurchased elements:• Testing and evaluation related to alternativePMT suppliers• Finalizing the DPU specifications and preparationof an RFP• Design, procurement and test of science performancetest detectors• Development of specifications for flight detectorprocurementPrototype hardware will be developed for all flightcomponents and tested to ensure the design willmeet all scientific performance and design requirements.GSE will be designed to the preliminarydesign level.The following sections detail these Phase A/Befforts according to institution.4.1.1 Marshall Space Flight CenterDevelopment Phase A/B:Science Participation in phase A/B:As the P.I. institution, the P.I. at MSFC will assignresponsibilities to Co-I institutions for the phase A/Bactivities, as outlined in Section 4.1 above, andoversee the development and implementation of allphase A/B activities. If needed, the P.I., in consultationwith the GBM P.M.,will direct changes in theseresponsibilities to meet GLAST GBM resourcesand schedule.The MSFC tasks during Phase A/B will also includethe development of Safety and Mission Assurancedocumentation, system engineering documentation,flight software and data processing software developmentplan and preliminary design, and developmentof the Burst Monitor DPU draft contract enditem (CEI) specification. It is anticipated that noflight hardware will be procured.Safety and Mission AssuranceThe MSFC GBM project team will support S&MAactivities in phase A/B with an approach that includes:1) Strong emphasis on S&MA management,2) thorough, experienced-based understanding ofS&MA principles, NASA and MSFC S&MApolicies and requirements, 3) focus on establishingclear goals and expectations for the GBM projectS&MA effort, and 4) applying the appropriate useof existing state-of-the-art S&MA tools and techniquesor, if deemed necessary and/or advantageous,the judicious development of new tools and techniques.A draft safety and mission assurance program plan(SMAPP) has been developed for the GBM projectto ensure risk management, system safety, quality54

assurance, reliability, and maintainability analysis.The draft SMAPP will be finalized and submittedfor approval within the first 60 days of contractaward. The draft SMAPP may be found as AppendixB of the Burst Monitor management plan.Systems EngineeringUsing the Burst Monitor project plan andNPG7120.5A as a guide, a draft ICD will be developedfor MPE-provided hardware, the DPU and theBurst Monitor to spacecraft harness, when detailedinformation becomes available on the spacecraftinterface. A memorandum of understanding (MOU)will be developed with MPE that outlines the MSFCdata requirements (DR’s) that MPE will followduring development of the MPE flight hardware.A draft instrument integration plan and hardwareprocessing flow document, including identificationof ground support equipment (GSE) will be developed.SoftwarePreliminary requirements will be developed inphase B for flight software, operations software, andscience analysis software. For flight software, thefocus will be on the burst trigger and locationalgorithms. Specifically, we will perform tradestudies to determine if there is a cost effective andmore sensitive alternative to the BATSE techniquefor triggering. We will also derive a requirement forprogram and data memory. For the data analysissoftware, the focus will be on revisions and enhancementsto our spectral analysis package WING-SPAN.We will also work with the LAT team to determinewhat products are expected from the Burst Monitor,both on board and on the ground.Data Processing UnitA preliminary CEI performance specification forthe DPU will be produced in phase B. The primaryfocus will be on memory requirements derived fromthe flight software study. A cost/risk trade study willbe performed to determine the appropriate level ofredundancy.4.1.2 Max-Plank Development Phase A/B:MPE is a major collaborator in the development ofthe Burst Monitor. MPE will provide the GBMscintillation detector elements and other flightcomponents, the HVPS, and LVPS to the BurstMonitor Project. MPE provides these detectors at nocost to NASA. The science team will developperformance requirements and design specificationsby the MPE project team during phase B.Flight hardware tasks, undertaken by MPE, can besubdivided into two main parts: development andfabrication of the NaI and BGO detector modulesand fabrication of the HV and LV power supplies.Development and fabrication of the power supplieswill be performed under contract to MPE. Thedetector modules will be designed by MPE, togetherwith industry. At MPE a breadboard (BB) model ofthe detection chain will be built for optimizationand study of the electrical design. Fabrication andintegration of flight hardware and structural testmodels (STM’s) and the electrical models (EM’s) ofthe detector modules will be performed by industry.BGO and NaI crystals, needed for the flight hardware,will be contracted for separately by MPE.MPE scientific personnel will carry out all detectorperformance tests and detector calibrations.Development and fabrication of the detector moduleswill be accomplished in several phases. Theidea is to reach predefined goals and to simplify theorganization of the project at the end of each phase.For each phase a major reassessment of the projectwill be conducted that will allow an effective controlof the costs.The development plan with the different projectphases A/B is listed below with a short descriptionof the contents of each phase. The development ofthe detector modules will be accomplished with thehelp of several prototype and test modules, asneeded for design verification, spacecraft integrationfit and form, structural, thermal, electrical, andfunctional tests.MPE Development PlanPhase A—MPE and Contractors• Preliminary studies by MPE and MSFC/UAH55

• Industrial studies of the structure, thermal, andEMC behavior• Design of voltage divider and preamplifier• Monte Carlo simulations of detector behavior foroptimization of the design, together with UAH/MSFC• Build a breadboard of the detector chain forsystematic investigations of detector behaviorPhase B—Contractors, with MPE Oversight• Definition of the detector module design• Coordination of the mechanical interfaces• Coordination of the electrical interfaces• Verification of the detection chain (PMT, voltagedivider, preamplifier) with the BB• Assembly, integration, and test—planning, definitionof the specifications, test procedures andscientific requirements.Models of the Detector ModulesSeveral structural and engineering models will bebuilt to facilitate detector development:Structure and Thermal Model:A structure/thermal model of an NaI and a BGOdetector module will be manufactured whichwill be representative of the mass, center ofgravity, and moment of inertia of the flightdetectors. Thermal and power characteristicswill be simulated.Tests: vibration and thermal vacuum testEngineering Model:An engineering model of an NaI and a BGOdetector module will be fabricated, which willbe representative in structure and electricaldesign. Only one PMT will be used in the caseof the BGO module. Light-Emitting Diodes(LED’s) will be used for simulation of highenergysignals.Tests: functional and thermal vacuum test, andperformance test.4.1.3 University of Alabama in Huntsville(UAH)As a major co-investigator institution in the GLASTGBM, UAH will participate in most of the analysesand trade studies described above (section 4.1)during the phase A/B effort. An evaluation of testprocedures, software, calibration and data analysissoftware design carry-over from the BATSE/Compton Observatory program will be made byUAH scientific personnel during phase A/B. Scientificsupport will be provided by UAH in the developmentof flight hardware specifications, althoughno flight hardware design work will be performedby UAH.4.2 Trade StudiesThe Burst Monitor design uses flight proven, maturetechnology. It is therefore anticipated that nomajor system level trades will be performed inphase A/B until the details of the Burst Monitor toGLAST spacecraft interface is available. At thattime system level trades will be performed todetermine the optimal location of the Burst Monitordetectors, HV and LV power supplies, and the DPU.Trades also may be performed on the DPU CEIspecifications to optimize cost and performancecharacteristics. A trade study will be performed todetermine the appropriate level of redundancy forthe DPU.4.3 Team InteractionsSection 2.8 describes the management approach andresponsibilities of each team member. During phaseA/B, weekly status telecons will be held to keepteam members informed. An action item trackinglist will be maintained and used to track openissues.Quarterly team meetings will be held in conjunctionwith GLAST project reviews at GSFC. This willoptimize limited team resources. The team willinteract at major reviews such as PDR, CDR, andFRR.Effective communication with MPE will be ofparticular importance because of the distant locationof MPE in Germany. Representatives of MPE willbe included in Burst Monitor telecons and fax, ande-mail will also be used in maintaining communication.It is anticipated that MPE would be present atthe quarterly GSFC reviews and at all major teammeetings.56

Export control of all information to MPE will bemaintained following NASA export control guidelines.5.0 Education and Public Outreach,Small Disadvantaged Businessand New Technology5.1 Education and Public OutreachSeveral characteristics of GRB’s make them anexcellent topic for education and public outreach.They are now known to be the most powerfulexplosions in the universe, they can be seen at verylarge distances and very early times, and their originremains a mystery. They have and will continue toexcite the interest of the science attentive public andwill certainly be a major focus of the GLASTeducation and public outreach (EPO) effort. TheBurst Monitor team is committed to a vigorous andproductive EPO program for GLAST.We concur with the approach outlined in the AO,wherein the secondary instrument and interdisciplinaryscientist EPO efforts are integrated into aunified GLAST effort led by the LAT team. Thecontribution of the GBM will be defined during thedefinition phase. The GBM team expects to provideinput to this definition and will provide financialsupport to the level prescribed in the AO. We havebudgeted a total of $50k for the Burst Monitor EPOeffort. Since the GBM effort will not be an independentplan, but will support the observatory levelplan, we do not provide specific programs or schedules.The GBM team brings a valuable asset to theGLAST EPO program—the NASA/MSFC ScienceCommunications (SciComm) process, now in itsthird year of operation at MSFC. The SciCommprocess, designed and operated by practicing researchscientists, gives researchers the opportunityto directly communicate the results and implicationsof their work to a science-attentive audience and togenerate subsequent communication products thatimprove both peer and nonpeer communicationactivities. Through direct integration of mediarelations, education, and technology transfer functions,the SciComm process also provides for theparallel development of other consistent and scientificallyaccurate communication products, such aspress releases. SciComm has been generating threeto five headline or feature stories per week forInternet distribution since 1997. The subject matterincludes Earth science, microgravity research, andspace science. Their web site was the winner of the“1999 People’s Choice Webby Award” for bestscience site on the Internet. Stories on their web siteare presented to an audience of over 20,000 individualsper day, and have a significant track recordof leveraging additional coverage through popularmagazine articles, newspaper articles, televisionfeatures, and classroom applications.5.2 Small Disadvantaged BusinessMSFC has an impressive record of socioeconomicperformance. The trend has been toward increasedpercentages of Marshall’s procurement budgetgoing to targeted business groups. In FY98 amilestone was reached when, for the first time,double digits were achieved for the percentage ofprocurements with small disadvantaged businesses(SDB). The Center has two general goals of20 percent small business and 8 percent SDBidentified in its implementation plan. Both arecurrently being exceeded. There are other goals andobjectives assigned by NASA headquarters andduring the recent Minority Enterprise DevelopmentWeek, Marshall was recognized for exceeding all ofits FY98 goals.They included:Small Business8a ProgramSmall Disadvantage BusinessWoman Owned Small BusinessNASA 8 percent SDBSmall Business SubcontractingSDB SubcontractingWoman Owned SubcontractingWith the 11 months of FY99 information available,Marshall is meeting its small business goal with101 percent of goal, 8a with 100 percent, SDB with100 percent and Woman Owned with 120 percent.57

The subcontracting data for FY99 is not yetavailable; however, MSFC expects to also achievethese objectives.The GLAST Burst Monitor benefits from theseprograms which, in many cases, are embedded inthe MSFC institutional support contracts. We workwith the Small Business Office at MSFC to ensurethat any of our requirements for subcontracts thatmight be generated are considered for SB/SDB setaside.5.3 New TechnologyWe employ no new technology in the flight hardwaredue to the stringent cost cap for secondaryinstruments and because the scientific requirementsare easily met by established, flight proven technology.To perform the time-resolved spectral analysis,that is central to our scientific investigation, we willbe adapting and expanding the WINGSPAN programdeveloped for BATSE. This package is alreadybeing used extensively by the BATSE team, as wellas CGRO GI’s, and is an excellent candidate fortechnology transfer to the high-energy astrophysicscommunity.WINGSPAN allows interactive spectral fitting in awindowing environment, taking time sequences ofcount spectra from a single detector, modeled by aresponse matrix, as the basic data model. Thisdiffers, in philosophy, from the well-known XSPECpackage, which is commonly used to analyze x-rayspectra one at a time and lacks the ability to easilytrack temporal behavior of spectral model fit parameters.Like XSPEC, WINGSPAN can perform jointfits to spectra from any instrument that has observeda given event, and is not restricted to only data fromBATSE or even just the CGRO instruments. WING-SPAN is written in interactive data language (IDL),published by Research Systems, Inc. and FOR-TRAN. It relies on computer-portable data productsin the FITS format, which has become a standard inthe astrophysics community. We are currentlyextensively revising the program to make it platformindependent and to improve its capability for performinga time sequence of simultaneous fits to datafrom mulitple instruments.58

Appendix AResumes

Dr. Charles MeeganRole in GBM: Principal InvestigatorMarshall Space Flight CenterEducationB.S. Rensselaer Polytechnic Institute, 1966Ph.D. University of Maryland, 1973Role in GBMAs Principal Investigator, Dr. Meegan is responsiblefor the overall scientific direction of the BurstMonitor project. He supervises the effort of thescience team in Huntsville and is the official NASApoint of contact.ExperienceDr. Meegan’s main research interests lie in the areaof gamma-ray astronomy, with particular emphasison gamma-ray bursts. He has been actively involvedin GRB studies for the past two decades and hasplayed a significant role in major developments inthis field. For his doctoral dissertation, Dr. Meeganmeasured the energy spectrum of cosmic rayelectrons. His advisor at the University of Marylandwas Dr. James Earl. In 1974, he joined the gammarayresearch team of Dr. Robert Haymes at RiceUniversity as a post-docatoral research associate.There, he participated in balloon flight observationsof several gamma-ray sources, includingobservations of nuclear lines from the GalacticCenter.Dr. Meegan came to Marshall Space Flight Centerin 1976, first as an NRC Research Associate. Heaccepted a civil service position there in 1978. AtMarshall, he was involved in several balloon flightcampaigns, including a observation of nucleargamma-ray lines from supernova 1987A.Dr. Meegan was co-investigator on the originalproposal for the Burst and Transient SourceExperiment (BATSE). He has been heavily involvedin the design, development, testing, on-orbitoperations and data analysis for BATSE for the pasttwenty years. He currently heads the BATSE BurstTeam, whose primary responsibility is production ofthe BATSE burst catalogs.Dr. Meegan served as assistant project scientist onthe Hubble Space Telescope from 1982 to 1984.From 1991 to 1996 he was the mission scientist forAstro2, a Spacelab mission comprising three UVtelescopes. Dr.Meegan was chairman of theorganizing committee for the 4 th HuntsvilleSymposium on Gamma-Ray Bursts. He has coauthoredover 100 refereed journal papers, the vastmajority of them on gamma-ray bursts.SocietiesAmerican Astronomical Society (AAS)AAS High-Energy Astrophysics DivisionAmerican Physical Society (APS)Sigma Xi, The Scientific Research SocietyVon Braun Astronomical SocietyHonors and AwardsNASA Medal for Outstanding ScientificAchievement, 1993.NASA Exceptional Achievement Medal, 1995.Sigma Xi Research Scientist of the Year, Huntsville- 1993.NASA Directors Commendation, 1993Recent Publications“Spatial Distribution of Gamma-Ray BurstsObserved by BATSE”, 1992, Meegan, C. et al.,Nature Vol. 355, p.143.“Identification of Two Classes of Gamma-RayBursts”, 1993, Kouveliotou, C. et al., ApJ, Vol. 413,p. L101.“Detection of Signature Consistent WithCosmological Time Dilation in Gamma-RayBursts”, 1994, Norris, J. et al., ApJ, Vol. 424, p.540.“Discovery of Intense Gamma-Ray Flashes ofAtmospheric Origin”, 1994, Fishman, G. et al,Science, Vol. 264, p. 1313.“Do Gamma-Ray Burst Sources Repeat?”, 1995,Meegan, C. et al., ApJ, Vol. 446, p. L15.“Gamma-Ray Bursts”, 1995, Fishman, G. &Meegan, C., Annu. Rev. Astron. Astrophys., Vol. 33,p. 415.“The Third BATSE Gamma-Ray Burst Catalog”,1996, C. Meegan et al., ApJ, Vol. 106, p. 65.“A New Type of Transient High-Energy Source inthe Direction of the Galactic Centre”, 1996,Kouveliotou, C. et al., Nature, Vol. 379, p. 799.“BATSE Observations of the Large-ScaleIsotropy of Gamma-Ray Bursts”, 1976, Briggs, M.et al., ApJ, Vol. 459, p. 40.“Transient Optical Emission From the Error Boxof the Gamma-Ray Burst of 28 February 1997”,1997, van Paradijs, J. et al., Nature, Vol. 386, p. 686.“BATSE Observations of Gamma-Ray BurstSpectra. IV. Time Resolved High EnergySpectroscopy”, 1998, Preece, R. et al., ApJ, Vol.496, p. 849.“The Fourth BATSE Gamma-Ray Burst Catalog(Revised)”, 1999, Paciesas, W. S. et al., ApJS, Vol.122, p. 465.

Dr. Michael S. BriggsRole in GBM: Co-InvestigatorEducationA.B. in Physics, Princeton University, 1982M.S. in Physics, University of California, SanDiego, 1983Ph.D. in Physics, University of California, SanDiego, 1991Role in GBM:Primary: Flight software development, MissionOperations DirectorSecondary:Detector calibrations, data analysissoftwareExperienceDr. Briggs has been involved in high-energy astrophysicssince 1978, when he worked for the cosmicray research group at NASA/GSFC. At the Universityof California he worked on the design and testingof the BATSE Spectroscopy Detectors. His thesiswas an all-sky map in the few hundred keV to fewMeV band constructed using archival HEAO A-4data. Since his graduation from the University ofCalifornia, he has worked with the gamma-ray grouplocated at NASA/MSFC as a University of Alabamaresearch scientist, focusing mainly on GRB research.He had a major role in the development ofthe spectral analysis software WINGSPAN and haslead the effort to find spectral lines in BATSE GRBdata. In addition to his participation in the spectralresearch, he has worked on issues related to GRBlocations. He was a co-editor of the Third HuntsvilleGamma-Ray Burst Symposium and is a coauthor ofmore than 60 refereed journal articles. His researchinterests include GRB’s, gamma-ray instrumentation,statistical techniques including Bayesianinference, soft-gamma repeaters and x-ray binariesSocietiesAmerican Astronomical Society (AAS)AAS High-Energy Astrophysics DivisionHonors and AwardsFirst Place in the 37th Annual Science Talent Searchfor the Westinghouse Science Scholarships andAwards,1978-1982University of California Regents’ Fellowship,1982Marlar Fellowship,1987-1988NASA Graduate Student Researcher’s Program,1987-1990NASA Compton Gamma-Ray ObservatoryFellowship, 1991-1994Recent PublicationsM. S. Briggs, D. L. Band, R. M. Kippen, R. D.Preece, C. Kouveliotou, J. van Paradijs, G. H.Share, R. J. Murphy, S. M. Matz, A. Connors, C.Winkler, M. L. McConnell, J. M. Ryan, O. R.Williams, C. A. Young, B. Dingus, J. R. Catelli, &R. A. M. J. Wijers, “Observations of GRB990123by the Compton Gamma-Ray Observatory ”, ApJ,Vol. 524, p. 82 (1999).M. S. Briggs, G. N. Pendleton, R. M. Kippen, J. J.Brainerd, K. Hurley, V. Connaughton & C. A.Meegan, “The Error Distribution of BATSE GRBLocations ”, ApJS, Vol. 122, p. 503 (1999).W. S. Paciesas, C. A. Meegan, G. N. Pendleton,M. S. Briggs, C. Kouveliotou, T. M. Koshut, J. P.Lestrade, M. L. McCollough, J. J. Brainerd, J.Hakkila, W. Henze, R. D. Preece, V. Connaughton,R. Marc Kippen, R. S. Mallozzi & G. J. Fishman,“The Fourth BATSE Gamma-Ray Burst Catalog(Revised) ”, ApJS, Vol. 122, p. 465 (1999).G. N. Pendleton, M. S. Briggs, R. M. Kippen, W. S.Paciesas, M. Stollberg, P. Woods, C. A. Meegan, G. J.Fishman, M. L. McCollough & V. Connaughton,“The Structure and Evolution of LOCBURST: TheBATSE Burst Location Algorithm ”, ApJ, Vol. 512,p. 362 (1999).T. W. Giblin, J. van Paradijs, C. Kouveliotou, V.Connaughton, R. A. M. J. Wijers, M. S. Briggs, R.D. Preece & G. J. Fishman, “Evidence for an EarlyHigh-Energy Afterglow Observed with BATSEfrom GRB980923 ”, ApJL, in press (1999).M. S. Briggs, “Gamma-Ray Burst Lines”, in ASPConf. Series 190, Gamma-Ray Bursts: The FirstThree Mintues, ed. J. Poutanen & R. Svensson(1999).M. S. Briggs, W. S. Paciesas, G. N. Pendleton, C. A.Meegan, G. J. Fishman, J. H. Horack, M. N. Brock,C. Kouveliotou, D. H. Hartmann & J. Hakkila,“BATSE Observations of the Large-Scale Isotropy ofGamma-Ray Bursts ”, ApJ, Vol. 459, p. 40 (1996).

Dr. Roland L. DiehlRole in GBM: Co-InvestigatorEducationDiploma in Nuclear Physics, University of Mainz, 1978;Ph.D. in Physics and Astronomy, with Honors,Tech. University of Munich, 1988;Habilitation Tech. University of Munich, 1998Role in GBM:Calibration of the GBM components. Preparation ofdata analysis methods and software tools, both forthe GBM data and for combined analysis with otherburst measurementsExperienceDr. Diehl’s primary research interests have been ingamma-ray astronomy, specifically nuclear astrophysicswith gamma-ray lines from radioactivities. He haspublished several review articles in refereedjournals on this subject, and is an internationallyrecognized expert in this field. He has been theprincipal scientist in developing the data analysissoftware system for the COMPTEL gamma-raytelescope aboard the NASA Compton Observatory,and for the ground calibration of this instrumentwith radioactive sources and accelerator setups. Dr.Diehl coordinates the data analysis preparations forthe SPI collaboration of the INTEGRAL gammaraymission to be launched in 2001. He chaired theuser committee of the Garching Computing Centreof the MaxPlanck Gesellschaft, and co-chairedMPE’s division for Computing and Data Analysis,being coordinator of computing facilities of MPE’sgamma-ray group. Dr. Diehl has published over 150papers in refereed journals and conferenceproceedings, and has presented about 100 talks atinternational institutes and conferences. Dr. Diehljoined the Max Planck Institut fürextraterrestrischePhysik in 1979 as a member of the MPE Comptontelescope team headed by Dr. Volker Schönfelder.He is staff scientist in MPE’s gamma-ray astronomygroup since 1983.Recent Publications“1.8 MeV Gamma-Rays from the Vela Region”Diehl, R. et al: Astroph. Letters and Communications(1999)“Gamma-Ray Line Astronomical Measurementsand Nucleoynthesis” Diehl, R.: Nuclei in the CosmosV, eds. N. Prantzos (1999)“Gamma-Rays Observations and Massive Stars”Diehl, R.: K.A. van der Hucht, G. Koenigsberger &P.R.J. Eenens (eds.), Wolf-Rayet Phenomena inMassive Stars and Starburst Galaxies, Proc.IAUSymp. No. 193 (San Francisco: ASP),p.631(1999)“Gamma-Ray Line Emission from Radioactivitiesin Stars and Galaxies” Diehl, R.; Timmes. F.X.:PASP, Vol. 110, 748, pp. 637-659 (1998)“Ti Gets a Lifetime” 44 Woosley, S.E.; Diehl, R.:Physics World, Vol. 11, pp. 7, 22 (1998)“Modelling the 1.809 MeV Sky: Tracers of MasiveStar Nucleosynthesis” Diehl, R.; et al.:A&AS,Vol. 120,pp. 4, 321 (1996)“Radioactive 26 Al in the Galaxy:Observationsversus Theory” Prantzos, N.; Diehl R.:Phys.Rep.Vol. 267, 1, pp. 1-69 (1996)“Understanding COMPTEL 26 Al 1.8 MeV MapFeatures” Chen, W.; Gehrels, N.; Diehl, R.: ApJVol. 440, pp. L57-L60 (1995)“The Galaxy in the 26 Al Gamma-Ray Line at1.809 MeV” Diehl, R.; et al.: A&A Vol. 298, pp.445-460 (1995)“Imaging Diffuse Emission with COMPTEL”Diehl, R.: Exp. Astr. Vol. 6, pp. 103-108 (1995)“Response Determinations of COMPTEL fromCalibration Measurements, Models,and Simulations”Diehl, R.; et al.: Data Analysis in Astronomy IV,edited by V. diGesu et al.,Plenum Press New York,pp. 201-216, (1992)SocietiesGerman Physical Society ‚Deutsche PhysikalischeGesellschaft‘ (DPG)German Astronomer Society ‚AstronomischeGesellschaft‘ (AG)American Astronomical Society (AAS), High-Energy Astrophysics Division

Stephen ElrodRole on GBM: Project ManagerEducationB.S.E. with Honors, University of Alabama, in HuntsvilleAwards and HonorsTau Beta Pi.Mr. Stephen E. Elrod is a senior systems engineer in the Space Flight Experiments Group, and hascontinuously served as a U.S. government aerospace engineer and project manager for 23 years. Beginningin 1996, he successfully served as the MSFC project manager and chief systems engineer for the IMAGE/Wide Imaging Camera (WIC). This instrument is a component of the IMAGE Far Ultraviolet (FUV)instrument package and was developed jointly with the University of California, Berkeley. The WICdevelopment and recent delivery was accomplished on an accelerated schedule that required innovative andstreamlined approaches to program management, including tight cost controls, schedule compression anddiverse partnerships with academia, technical institutes and foreign governments. Mr. Elrod has alsorendered significant support to several other Science Directorate efforts, including the on-goingdevelopment of the Microgravity Crystal Growth Demonstration, flight component fabrication/testing forthe Chandra X-Ray Observatory and MSFC ISO 9001 certification.In his career, Mr. Elrod has worked in several project offices, including the TOW Missile (U.S. ArmyMICOM), U.S. Space Station, and the Orbital Maneuvering Vehicle. Additionally, he worked for severalyears as a supporting systems engineer on the Hubble Space Telescope. Mr. Elrod also serves as a U.S.Naval Reserve Civil Engineer Corps officer (18 yrs. service, CDR/05) and is currently assigned to the officeof the Deputy Chief of Naval Operations for Logistics, in Washington, D.C.

Dr. Gerald J. FishmanRole in GBM: Co-InvestigatorNASA/Marshall Space Flight CenterEducationB.S. with Honors in Physics, University of Missouri, 1965M.S., Space Science, Rice University, 1968Ph.D., Space Science, Rice University, 1969ExperienceDr. Fishman’s primary research interests have been ingamma-ray astronomy, nuclear astrophysics, and backgroundradiation in space. He has been the principalinvestigator on a large number of space-borne andballoon-borne gamma-ray astronomy experiments andbackground monitoring experiments. Dr. Fishman haspublished over 250 papers in refereed journals andconference proceedings. He has served on a number ofNASA Headquarters committees, including the NASAGamma-Ray Astronomy Program Working Group andthe Astrophysics Working Group.Presently, he is the Principal Investigator of the Burstand Transient Source Experiment (BATSE) on theCompton Gamma-Ray Observatory.Following graduation from Rice University in 1969 witha Ph.D. in Space Science, he was a senior researchscientist for Teledyne Brown Engineering. Dr. Fishmanjoined NASA/Marshall Space Flight Center in 1975 asan astrophysicist. He spent 1977-78 as a staff scientist inthe Astrophysics Division, NASA Headquarters, beforereturning to MSFC. In 1992 he was named chief of theGamma-Ray Astronomy Branch. From 1994–1998, hewas a senior staff scientist in the Astrophysics Division,serving as the leader of the gamma-ray astronomy teamin the Space Science Laboratory of the NASA/MSFC. In1998, Dr. Fishman was appointed by NASA’s Administratoras NASA/MSFC chief scientist for Gamma-RayAstronomy. This is a senior scientific and technicalposition, the technical equivalent of a government seniorexecutive service (SES) position.SocietiesAmerican Astronomical Society (AAS)AAS High-Energy Astrophysics DivisionAmerican Physical Society (APS)APS Astrophysics Division, Executive CommitteeAmerican Association for the Advancement of ScienceInternational Astronomical UnionSigma Xi, The Scientific Research SocietyHonors and AwardsO.M. Steward Scholar (Physics) University of Missouri1963–1965NASA Medal for Outstanding Scientific Achievement1982, 1991, and 1992NRL – Alan Berman Research Publication Award—1992Sigma Xi Research Scientist of the Year, Huntsville—1993Distinguished Alumnus Award, Univ. of Missouri—1994Rossi Prize, High Energy Astrophysics Division, AmericanAstronomical Society—1994Fellow – American Physical Society—1995Relevant PublicationsFishman, G.J.; and Hartmann, D.: “Gamma-Ray Bursts,”Scientific American, Vol. 277, pp. 34–39, July 1997Galama, T.; Groot, P.J.; van Paradijs, J.; Kouveliotou, C.;Robinson, R.R.; Fishman, G.J.; Meegan, C.A.; et al.:“The Decay of Optical Emission From the Gamma-RayBurst GRB 970228,” Nature, Vol.387, pp. 479–481, 1997Fishman, G.J.; and Meegan, C.A.: “Gamma-Ray Bursts,”Annual Review of Astronomy and Astrophysics, Vol.33, pp. 415–458, 1995Fishman, G.J.: “Gamma-Ray Bursts: An Overview,”Publications of the Astronomical Society of the Pacific,Vol. 107, pp. 1-7, 1995Fishman, G.J.: “Observations of Gamma-Ray Bursts,”The Gamma-Ray Sky With Compton GRO and Sigma(Klewer: Holland), M. Signore et al. (eds), pp. 381–94,1995Fishman, G.J.; and Barthelmy, S.: “Gamma-Ray Bursts:Observational Overview, Searches for Counterpartsand BACODINE” Flares and Flashes, proc, IAUcolloquium No. 151 (Springer:Berlin), J. Greiner, et al.(eds.), 1995Fishman, G.J.: “Observed Properties of Gamma-Ray Bursts”Astronomy and Astrophysics Supplement, Vol. 138,pp. 395–398, 1998Pendleton, G.N.; Paciesas, W.S.; Briggs, M.S.; Preece,R.D.; Mallozzi, R.S.; Meegan, C.A.; Horack, J.M.;Fishman, G.J.; Band, D.L.; Matteson, J.L.; Skelton,R.T.; Hakkila, J.; Ford, L.A.; Kouveliotou, C.; Koshut,T.M.: “The Identification of Two Different SpectralTypes of Pulses in Gamma-Ray Bursts” AstrophysicalJournal,Vol. 489, p.175, 1997van Paradijs, J.; Groot, P.J.; Galama, T.; Douvelioutou,C.; Strom, R.G.; Telting, J.; Rutten, R.G.M.; Fishman,G.J.; Meegan, C.A., Pettini, M.; Tanvir, N.; Bloom, J.;et al.: “Transient Optical Emission From the Error boxof the Gamma-Ray Burst of 28 February 1997”Nature, April 1997, pp. 686–689, 1997

Dr. Robert H. GeorgiiRole in GBM: Co-InvestigatorEducationDipl. Phys. in Physics, Technical University ofMunich, 1989; Dr.rer. nat. with magna cum laude inPhysics, Technical University of Munich,1994.Role in GBMDesign of the GBM detector modules and massmodeling.ExperienceDr. Georgii’s primary research interests havebeen in non-linear dynamics, nuclear physics andgamma-ray astronomy. He joined the MPE in1995 as an astrophysicist and is a Co-Investigatorof the SPI instrument of ESA’s INTEGRALmission. For this mission he is currently workingin the detector development.For COMPTEL he isengaged in the data analysis. During his doctoralthesis he spent 3 years at the ILL in Grenoble,France, working on detector development in nuclearphysics. He stayed half a year each at ENEA,Frascati, Italy and the University of Oxford, Oxford,England and was participating in detector developmentin laser and neutronphysics. Dr. Georgii haspublished about 40 papers in refereed journals andconference proceedings.SocietiesDeutsche Physikalische Gesellschaft (DPG)European Physical Society (EPS)Honors and AwardsFellowship at ENEA within the European UnionHuman Capital and Mobility (HCM) program.Recent PublicationsGeorgii, R.; Diehl, R.; Lichti, G.; Oberlack, U.;Schönfelder, V.; Ködelseder, J.;Rayan, J.: “UpperLimits of 26 Al and 60 Fe From M82”, Proceedings ofthe 2 nd INTEGRAL Workshop, ESA Publication, ESASP-382 (1997), 51.Georgii, R.; Diehl, R.; Lichti, G.G.; and Schönfelder,V.: “Can the INTEGRAL-Spectrometer SPI DetectX-Ray Lines From Local Galaxies?”, 4 th ComptonSymposium, AIP Conference Proceedings 410 (1997),1554.Georgii, R.; Meißl, M.; Hajdas, W.; Henschel,H.;Gräf, H.-D.; Lichti, G.G.; von Neumann-Cosel, P.;Richter, A.; Schönfelder, V.: “Influence of RadiationDamage on BGO Scintillation Properties”, NuclearInstruments and Methods Vol. A413, (1998), pp. 50–58.Vedrenne, G.; Jean, P.; Kandel, B.; Albernhe, F.; Borrel,V.; Mandrou, P.; Roques, J.P.; von Ballmoos,P.;Durouchoux, P.; Cordier, B.; Diallo, N.; Schönfelder,V.; Lichti, G.G.; Diehl, R.; Varendorff, M.; Strong, A.W.;Georgii,R.; Teegarden, B.J.; Naya, J.; Seifert, H.; Sturner,S.; Matteson, J.; Lin, R.; Slassi, S.; Sanchez, F.; Caraveo,P.; Leleux, P.; Skinner, G.K.; Connell, P.: “The SPISpectrometer for the INTE-GRAL Mission”, PhysicaScripta T77, (1998), 35–38.Georgii, R.; Plüschke, S.; Diehl, R.; Collmar, W.;Lichti, G.G.; Schönfelder, V.; Bennett, K.; Bloemen, H.;Knödlseder, J.; McConell, M.; Ryan, J.; “COMPTELUpper Limits for the 56 Co X-rays From SN1998”, 5 thCompton Symposium, AIP Conference proceedings(1999), inprint.

Dr. R. Marc KippenRole in GBM: Co-InvestigatorEducationB.S., with Honors in Physics, University of NewHampshire, 1988; M.S., Physics (Thesis: “MonteCarlo Simulation of the COMPTEL Gamma-RayTelescope”), University of New Hampshire, 1991;Ph.D., Physics (Dissertation: “Locations andSpectra of Cosmic Gamma-Ray Bursts), Universityof New Hampshire, 1995Role in GBM:Simulations for detector optimization, backgroundmodeling and detector response; DPU specifications;integration and test oversight; flight and ground-basedsoftware design/development; analysis of calibrationand flight data.ExperienceDr. Kippen has over 10 years experience in highenergyastrophysics research, including software &algorithm development; spacecraft operations; analysisand interpretation of astrophysical data from gammarayinstruments; imaging, spatial and spectral analysisof cosmic gamma-ray bursts; development, operationand maintenance of rapid gamma-ray burst localizationsystems and counterpart search networks; developmentand implementation of Monte Carlo detectorsimulation systems. His primary scientific interestsinclude gamma-ray bursts, nuclear decay emissionfrom astrophysical sources, the origin of cosmicdiffuse gamma rays and novel gamma-ray imaginginstruments. He has published more than 100 papers inrefereed journals and conference proceedings, and haspresented about 30 talks at a variety of internationalinstitutes, workshops and conferences.Dr. Kippen has worked as a member of theCOMPTEL and BATSE instrument teams, where heparticipated in the development of analysis and operationssoftware and performed scientific data analysis.Presently, he is a senior research associate at theUniversity of Alabama in Huntsville, where he continuesto work on the BATSE instrument team at NASA’sMarshall Space Flight Center. Prior positions includeresearch associate at the University of Alabama inHuntsville (1996–1999) and research scientist at theUniversity of New Hampshire (1995–1996).SocietiesAmerican Astronomical Society (AAS)AAS High-Energy Astrophysics DivisionSigma Xi, The Scientific Research SocietyAssociate of The Committee on Space Research(COSPAR)Honors and AwardsSigma Xi Outstanding Dissertation Award, Durham,NH, 1996Recent PublicationsKippen, R.M.; et al.: “Simulated Performance of theFiberGLAST Gamma-Ray Telescope Concept.” InProc. of the 26th International Cosmic Ray Conf., ed.D. Kieda, M. Salamon & B. Dingus, Vol. 5, p. 148,1999.Briggs, M.S., et al.: “Observations of GRB990123 by the Compton Gamma-Ray Observatory.”Astrophys. J. Vol. 524, p. 82, 1999.Briggs, M.S.; et al.: “The Error Distribution ofBATSE GRB Locations.” Astrophys. J. Suppl. Ser.Vol. 122, p. 503, 1998.Hurley, K.; et al.: “The Ulysses Supplement to theBATSE 4B Catalog of Cosmic Gamma-Ray Bursts.”Astrophys. J. Suppl. Ser. Vol. 122, p. 497, 1998.Paciesas, W.S.; et al.: “The Fourth BATSEGamma-Ray Burst Catalog.” Astrophys. J. Suppl.Ser. Vol. 122, p. 465, 1999.Woods, P.M.; et al.: “Discovery of a New SoftGamma Repeater, SGR 1627–41.” Astrophys. J.Lett. Vol. 519, p. L139, 1999.Kippen, R.M.; et al.: “On the Association ofGamma-Ray Bursts With Supernovae.” Astrophys.J. Lett. Vol. 506, p. L27, 1998.Galama, T.J.; et al.: “An Unusual Supernova inthe Error Box of the Gamma-Ray Burst of 25 April1998.” Nature Vol. 395, p. 670. 1998.Kippen, R.M.; et al.: “Characteristics of Gamma-RayBursts at MeV Energies Measured by COMPTEL.” Adv.Space Res. Vol. 22 (7), p. 1097, 1998.Pendleton, G.N.; et al.: “The Structure and Evolutionof LOCBURST: The BATSE Burst LocationAlgorithm.” Astrophys. J. Vol. 512, p. 362, 1998.Kippen, R.M.; et al.: “The Locations of Gamma-Ray Bursts Measured by COMPTEL.” Astrophys. J.Vol. 492, p. 246, 1998.Ryan, J.M.; et al.: “A Balloon-Borne CodedAperture Telescope for Arc-Minute Resolution atHard X-Ray Energies.” Adv. Space Res. Vol. 21 (7),p. 1009, 1998.Kippen, R.M.; et al.: “The BATSE Rapid BurstResponse System.” In AIP Conf. Proc. 428,Gamma-Ray Bursts:Fourth Huntsville Symposium,eds. C.A. Meegan, R.D. Preece, and T.M. Koshut,(New York: AIP Press), 119, 1998.

Dr. G. LichtiRole in GBM: Co-Principal InvestigatorEducationDiploma in Physics at the Technical University ofMunich, 1972,PhD (Dr. rer. nat.) in Physics at the TechnicalUniversity of Munich, 1975.Role in GBM:Co-PI with main responsibility for the German partof the Burst Monitor.ExperienceDr. Lichti has worked in the field of gamma-rayastronomy since the early 1970’s when he developed,together with Prof. Schönfelder, the first doubleCompton telescope and participated in severalsuccessful balloon campaigns. In August 1975 hewent as a member of the Caravan collaboration tothe European Space Operations Center in Darmstadtwhere he was responsible for the scientific operationand surveillance of the European gamma-raysatellite COS-B. In January 1980 he returned to theMax-Planck-Institut für extraterrestrische Physik inGarching and joined the COMPTEL collaboration.As hardware manager he was responsible for the developmentof the NaI detectors and the anticoincidence subsystemof COMPTEL. After the successful launch ofCGRO he was actively involved in the data analysis ofthe COMPTEL data. In 1994 he joined the SPI team ofINTEGRAL and works since then for this project. Aslocal project manager he has to organize and to coordinatethe manufacturing of the complete BGOanticoincidence shield by the industry. In parallel heis still involved in the analysis of COMPTEL data.SocietiesMember of the Deutsche Physikalische Gesellschaft(DPG).Honors and AwardsNASA public service-group achievement award 1992.Recent PublicationsLichti, G.G.; Balonek, T.; Courvoisier, T.J.-L.;Johnson, N.; McConnell, M.; McNamara, B.; vonMontigny, C.; Paciesas, W.; Robson, E.I.; Sadun, A.;Schalinski, C.; Smith, A.G.; Staubert, R.; Steppe, H.;Swanenburg, B.N.; Turner, M.J.L.; Ulrich, M.-H.; andWilliams, O.R.: “Simultaneous and Quasi-SimultaneousObservations of the Continuum Emission of theQuasar 3C 273 From Radio to Gamma-Ray Energies”,A&A, Vol. 298, p. 711, 1995Lichti, G.G.; Iyudin, A.; Bennett, K.; den Herder, J.W.;Diehl, R.; Morris, D.; Ryan, J.; Schön-felder, V.; Steinle,H.; Strong, A.W.; and Winkler, C.: “COMPTEL UpperLimits on Gamma-Ray Line Emission From Supernova1993J, A&A Suppl. Ser. Vol. 120, p. 353, 1996.Schönfelder, V.; Bennett, K.; Bloemen, H.; Coll-mar,W.; Diehl, R.; Hermsen, W.; Kuiper, L.; Lichti, G.G.;McConnell, M.; Ryan, J.; Strong, A.; and Winkler, C.:“Highlights from the COMPTEL 1 to 30 MeV SkySurvey”, 7 th Texas Symposium on RelativisticAstrophysics and Cosmology, Vol. 759 of the Annalsof the New York Academy of Sciences, p. 226,September 1995.Lichti, G.G.; Schönfelder, V.; Diehl, R.; Georgii, R.;Kirchner, T.; Vedrenne, G.; Mandrou, P.; vonBallmoos, P.; Jean, P.; Albernhe, F.; Durouchoux, P.;Cordier, B.; Diallo, N.; Sanchez, F.; Leleux, P.;Caraveo, P.A.; Teegarden, B.; Matteson, J.; Lin, R.;Skinner, G.K.; and Connell, P.: “The SpectrometerSPI of the INTEGRAL Mission”, Proc. of the SPIEConference (Denver), Vol. 2806, pp. 217–233, 1996.Jean, P.; Gomez-Gomar, J.; Hernanz, M.; Jose, J.;Isern, J.; Vedrenne, G.; Mandrou, P.; Schönfelder, V.;Lichti, G.G.; and Georgii, R.: “Possibility of theDetection of Classical Novae With the Shield of theINTEGRAL-Spectrometer SPI”, Proceedings of the3 rd INTEGRAL Workshop (Taormina), 1998.Weidenspointner, G.; Varendorff, M.; Bennett, K.;Bloemen, H.; Hermsen, W.; S. C. Kappadath, S.C.;Lichti, G.G.; Ryan, J.; and Schönfelder, V.: “The CDGSpectrum From 0.8-30 MeV Measured With COMP-TEL Based on a Physical Model of the InstrumentalBackground”, Proceedings of the 3 rd INTEGRALWorkshop (Taormina), 1998.Bloemen, H.; Morris, D.; Knödlseder, J.; Bennett, K.;Diehl, R.; Hermsen, W.; Lichti, G.G.; van der Meulen,R.D.; Oberlack, U.; Ryan, J.; Schönfelder, V.; Strong,A.W.; de Vries, C.; and Winkler, C.: “COMPTEL OrionResults Revisited?”, Proceedings of the 3 rd INTEGRALWorkshop (Taormina), 1998.Iyudin, A.F.; Bloemen, H.; Diehl, R.; Hermsen, W.;Knödlseder, J.; Lichti, G.G.; Ryan, J.; Schönfelder, V.;Strong, A.; and Winkler, C.: “COMPTEL Constraintson Nova-Produced 22 Na”, Proceedings of the 3 rdINTEGRAL Workshop (Taormina), 1998.Lichti, G.G.; Georgii, R.; von Kienlin, A.; Schönfelder,V.; Wunderer, C.; Jung, H.-J.; and Hurley, K.: “TheGamma-Ray Burst-Detection System of theINTEGRAL-Spectrometer SPI”, Proceedings of the 5 thCompton Symposium, 1999.Weidenspointner, G.; Varendorff, M.; Bennett, K.;Bloemen, H.; Hermsen, W.; Kappadath, S.C.; Lichti,G.G.; Ryan, J.; and Schönfelder, V.: “The Spectrum of theCosmic Diffuse Gamma-Ray Background From 0.8-30MeV Measured With COMPTEL”, submitted to A&A ,1999.

Dr. Robert S. MallozziRole in GBM: Co-InvestigatorEducationB.S., Physics, Pennsylvania State University, 1990;M.S., Physics, University of Alabama in Huntsville, 1992;Ph.D., Physics, University of Alabama in Huntsville, 1996.Role in GBMAs Co-Investigator on the GBM project, Dr. Mallozziis primarily responsible for instrument operations anddata analysis software development and maintenance,including science data analysis tools and instrumentmonitoring software. He is also involved in instrumentsimulation studies and calibration, detector responsefunction generation, and public outreach. Scienceinterests include wide band gamma-ray burst spectralstudies, and multi-wavelength, multi-instrumentgamma-ray burst spectroscopy.ExperienceDr. Robert Mallozzi is currently a Senior ResearchAssociate at the University of Alabama in Huntsville,working at NASA’s Marshall Space Flight Center inthe field of high energy astrophysics. His doctoralresearch involved the study of gamma-ray bursts,which are brief flashes of high energy cosmic radiationthat occur randomly on the celestial sphere. Hisresearch investigated effects of a cosmological originof these events. After receiving his doctoral degree, heaccepted a research position on the science team of theBurst and Transient Source Experiment (BATSE) tocontinue studies of gamma-ray bursts. His currentresearch is focused on spectral properties of bursts,and the phenomenon of Terrestrial Gamma Flashes,which are flashes of gamma radiation that werediscovered with BATSE to originate from within theatmosphere of Earth. Current work encompasses abroad range of tasks, including advanced scientificdata analysis and visualization, numerical modelingand simulation, and three dimensional computergraphics and animation. Although formally trained inthe physical sciences, Dr. Mallozzi also has a stronginterest in Computer Science, and is currently pursuingan advanced degree in that discipline.SocietiesAmerican Astronomical Society (AAS)Selected PublicationsPreece, R.D.; Briggs, M.S.; Mallozzi, R.S.;Pendleton, G.N.; Paciesas, W.S.; Band, D.L.; “TheBATSE Gamma-Ray Burst Spectral Catalog I. HighTime Resolution Spectroscopy of Bright Bursts UsingHigh Energy Resolution Data.”, ApJS, in press 1999.Preece, R.D.; Briggs, M.S.; Mallozzi, R.S.;Pendleton, G.N.; Paciesas, W.S.; and Band, D.L.: “TheSynchrotron Shock Model Confronts a Line of Deathin the BATSE Gamma-Ray Burst Data.” ApJ Letters,Vol. 506, p. L26,1999.Preece, R.D.; Pendleton, G.N.; Briggs, M.S.;Mallozzi, R.S.; and Paciesas, W.S.: “BATSEObservations of Gamma-Ray Burst Spectra IV. Time-Resolved High-Energy Spectroscopy.” ApJ, Vol. 497,p. 849, 1998.Pendleton, G.N.; Paciesas, W.S.; Briggs, M.S.;Preece, R.D.; Mallozzi, R.S.; Meegan, C.A.; Horack,J.M.; Fishman, G.J.; Band, D.L.; Matteson, J.; Skelton,R.T.; Hakkila, J.; Ford, L.; Kouveliotou, C.; andKoshut, T.M.: “The Identification of Two DifferentSpectral Types of Pulses in Gamma-Ray Bursts.” ApJ,Vol. 489, p. 175, 1998.Hakkila, J.; Meegan, C. A.; Horack, J.M.; Pendleton,G.N.; Briggs, M.S.; Mallozzi, R.S.; Koshut, T.M.;Preece, R.D.; and Paciesas, W.S.: “LuminosityDistributions of Cosmological Gamma-Ray Bursts”ApJ, Vol. 462, p. 125, 1996.Mallozzi, R. S.; Paciesas, W. S.; and Pendleton, G.N.: “Effects of Spectral Shape on CosmologicalModels of the BATSE Gamma-Ray Burst IntensityDistribution” ApJ, Vol. 471, p. 636, 1996.Meegan, C.A.; Pendleton, G.N.; Briggs, M.S.;Kouveliotou, C.; Koshut, T.M.; Lestrade, J.P.; Paciesas,W.S.; McCollough, M.L.; Brainerd, J.J.; Horack, J.M.;Hakkila, J.; Henze, W.; Preece, R.D.; Mallozzi, R.S.; andFishman, G.J.: “The Third BATSE Gamma-Ray BurstCatalog” ApJS, Vol. 106, p. 65, 1996.Pendleton, G.N.; Mallozzi, R.S.; Paciesas, W.S.;Briggs, M.S.; Preece, R.D.; Koshut, T.M.; Horack, J.M.; Meegan, C.A.; Fishman, G.J.; Hakkila, J.; andKouveliotou, C.: “The Intensity Distribution forGamma-Ray Bursts Observed with BATSE.” ApJ, Vol.464, p. 606, 1996.Mallozzi, R.S.; Paciesas, W.S.; Pendleton, G.N.;Briggs, M.S.; Preece, R.D.; Meegan, C.A.; andFishman, G.J.: “The _F _Peak Energy Distributionsof Gamma-Ray Bursts Observed by BATSE.” ApJ,Vol. 454, p. 597, 1995.Fishman, G.J.; Bhat, P.N.; Mallozzi, R.S.; Horack,J.M.; Koshut, T.M.; Kouveliotou, C.; Pendleton, G.N.; Meegan, C.A.; Wilson, R.B.; Paciesas, W.S.;Goodman, S.J.; and Christian, H.J.: “Discovery ofIntense Gamma-Ray Flashes of Atmospheric Origin.”Science, Vol. 264, p. 1313, 1994.

Prof. William S. PaciesasRole in GBM: Co-InvestigatorEducationB.S., magna cum laude, Physics, Seton Hall University, 1969;M.S., Physics, University of California, San Diego, 1971;Ph.D., Physics, University of California, San Diego, 1978.Role in GLAST GBMProf. Paciesas is a co-investigator, with primaryresponsibility for interface with the LAT. He is alsothe lead UAH representative to the GBM project.His primary science interest in the GBM is timeresolvedwide-band spectroscopy of gamma-raybursts.ExperienceProf. Paciesas’ primary research interests are in observationalx-ray and gamma-ray astronomy, including gammaraybursts (especially spectroscopy), galactic black holecandidates, low-mass x-ray binaries, and Seyfert galaxies.He has worked on a number of balloon-borne and spacebornetelescopes, and he has been the author or co-authorof approximately 100 refereed and 275 non-refereedpublications. He has been a member of the CGRO User’sCommittee, the CGRO Data Analysis Operations WorkingGroup, and the CGRO Operations Working Group.He is currently a co-investigator on the CGRO Burst andTransient Source Experiment.Prof. Paciesas currently holds the position of ResearchProfessor at the University of Alabama in Huntsville.From 1978–1980, he was a NAS/NRC Resident ResearchAssociate at NASA/GSFC, and during 1980–1982 heworked as a Research Associate at the University ofMaryland. He has been a research faculty member atUAH since 1982, working primarily on CGRO/BATSE,and is the leader of the UAH gamma-ray astronomygroup.SocietiesAmerican Astronomical Society (AAS)AAS High Energy Astrophysics DivisionAmerican Physical SocietyAmerican Association for the Advancement of ScienceInternational Astronomical UnionIEEE Nuclear & Plasma Sciences SocietyHonors and AwardsSigma Pi Sigma (physics honor society) 1968Delta Epsilon Sigma (academic honor society) 1969NASA Group Achievement Award, Spacelab 2Nuclear Radiation Monitor 1986NASA Group Achievement Award, GROOperations Working Group 1992NASA Group Achievement Award, BATSEInstrument Team 1992Selected Recent PublicationsPaciesas, W.S.; Meegan, C.A.; Pendleton, G.N.; Briggs,M.S.; Kouveliotou, C.; Koshut, T.M.; Lestrade, J.P.;McCollough, M.L.; Brainerd, J.J.; Hakkila, J.; Henze, W.;Preece, R.D.; Connaughton, V.; Kippen, R.M.; Mallozzi,R.S.; Fishman, G.J.; Richardson, G.A.; and Sahi, M.: “TheFourth BATSE Gamma-Ray Burst Catalog (Revised)” ApJS,Vol. 122, p. 465.Preece, R.D.; Briggs, M.S.; Mallozzi, R.S.; Pendleton,G.N.; Paciesas, W.S.; and Band, D.L.: “The BATSEGamma-Ray Burst Spectral Catalog I. High Time ResolutionSpectroscopy of Bright Bursts using High Energy ResolutionData.” ApJS, in press, 1999.Preece, R.D.; Briggs, M.S.; Mallozzi, R.S.; Pendleton,G.N.; Paciesas, W.S.; and Band, D.L.: “The SynchrotronShock Model Confronts a Line of Death in the BATSEGamma-Ray Burst Data.” ApJ Vol. 506, p. L26, 1999.Mitrofanov, I.G.; Litvak, M.L.; Briggs, M.S.; Paciesas,W.S.; Pendleton, G.N.; Preece, R.D.; and Meegan, C.A.:“Average Emissivity Curve of BATSE Gamma-Ray BurstsWith Different Intensities.” ApJ Vol. 523, p. 610, 1999.van der Hooft, F.; Kouveliotou, C.; van Paradijs, J.;Paciesas, W.S.; Lewin, W.H.G.; van der Klis, M.; Crary,D.J.; Finger, M.H.; Harmon, B.A.; and Zhang, S.N.: “HardX-Ray Lags in GRL J1719–24” ApJ 519, 332, 1999.Pendleton, G.N.; Briggs, M.S.; Kippen, R.M.; Paciesas,W.S.; Stollberg, M.; Woods, P.; Meegan, C.A.; Fishman,G.J.; McCollough, M.L.; and Connaughton, V.: “TheStructure and Evolution of LOCBURST: The BATSE BurstLocation Algorithm” ApJ Vol. 512, p. 362, 1999.Preece, R.D.; Pendleton, G.N.; Briggs, M.S.; Mallozzi,R.S.; and Paciesas, W.S.: “BATSE Observations of Gamma-Ray Burst Spectra IV. Time-Resolved High-EnergySpectroscopy” ApJ, Vol. 497, p. 849, 1998.Harmon, B.A.; Deal, K.J.; Paciesas, W.S.; Zhang, S.N.;Robinson, C.R.; Gerard, E.; Rodriguez, L.F.; and Mirabel,I.F.: “Hard X-Ray Signature of Plasma Ejection in theGalactic Jet Source GRS 1915+105” ApJ Vol. 477, p. L85,1997.Mallozzi, R.S.; Paciesas, W.S.; Pendleton, G.N.; Briggs,M.S.; Preece, R.D.; Meegan, C.A.; Fishman, G.J.; “The nF n_Peak Energy Distributions of Gamma-Ray Bursts Observedby BATSE” ApJ, Vol. 454, p. 597, 1995.

Dr. Robert D. PreeceRole in GBM: Co-InvestigatorEducationB.A., with Distinction in Mathematics and Physics,University of California, Berkeley, 1982;M.S., Physics, The Ohio State University, 1985;Ph.D., Astrophysics, University of Maryland, 1990Role in GBM: Data Analysis Manager.ExperienceDr. Robert D. Preece is a senior research associate for theDepartment of Physics at the University of Alabama inHuntsville (UAH). His primary research interests havebeen gamma-ray bursts, soft-gamma repeaters andquantum synchrotron emission from astrophysicalsources. He is a member of the instrument team forNASA’s Burst and Transient Source Experiment (BATSE)on board the Compton Gamma-Ray Observatory, workingin the areas of spectral analysis, instrument calibrationand astrophysical modeling of gamma-ray emission. Hejoined the BATSE team while an NRC postdoc atNASA’s Marshall Space Flight Center.With the recent observation of x-ray afterglow in severalGRB’s, it is crucial to understand the correlation betweenthe x-ray and gamma-ray bands. One of Dr. Preece’srecent awarded proposals is to do a time-resolved analysisof the x-ray excess emission (5–20 keV) observed in 15percent of the GRB’s observed by the BATSE spectroscopydetectors. Joint fitting of data between the severalspace-based instruments that have observed GRB’s todate, is feasible with the long-duration mission of CGRO.While any one burst many be seen by a only a fewinstruments, the whole ensemble has been observed by alarge number of detectors, with the CGRO instrumentsserving as the common thread. With Dr. M. S. Briggs atUAH, he has a long-term (5 year) grant from NASA toaccomplish this effort.Dr. Preece was involved in the planning of the 4th HuntsvilleGRB Symposium, held on Sept. 15 – 20, 1997, as amember of the Local Organizing Committee; this alsoincluded being a co-editor with Drs. Charles Meegan andThomas Koshut of the Symposium Proceedings. As anaside to this effort, he participated with the CGRO ScienceSupport Center at Goddard Space Flight Center to createthe second BATSE data CD-ROM, which was deliveredto the gamma-ray burst community at the Symposium.SocietiesAmerican Astronomical Society (AAS)AAS High-Energy Astrophysics DivisionRecent PublicationsPreece, R.; Briggs, M.; Mallozzi, R.; Pendleton, G.;Paciesas, W.; and Band, D.: “The BATSE Gamma-Ray Burst Spectral Catalog. I. High-Time ResolutionSpectroscopy of Bright Bursts Using High-EnergyResolution Data”, ApJ, 1999, in press.Paciesas, W.; Meegan, C.A.; Pendleton, G.; Briggs,M.; Kouveliotou, C.; Koshut, T.; Lestrade, J.P.;McCollough, M.; Brainerd, J.J.; Hakkila, J.; Henze,W.;Preece, R.; et al.: “The Fourth BATSE Gamma-RayBurst Catalog (Revised)”, ApJS, 1999, Vol. 122 p. 465.Briggs, M.S.; Band, D.L.; Kippen, R.M.; Preece,R.D.; Kouveliotou, C.; van Paradijs, J.; Share, G.H.;Murphy, R.J.; Matz, S.M.; Connors, A.; Winkler, C.;McConnell, M.L.; Ryan, J.M.; Williams, O.R.; Young,C.A.; Dingus, B.; Catelli, J.R.; and Wijers, R.A.M.J.:“Observations of GRB 990123 by the ComptonGamma-Ray Observatory”, ApJ, Vol. 524 p. 82, 1999.Galama, T.J.; Briggs, M.S.; Wijers, R.A.M.J.;Vreeswijk, P.M.; Rol, E.; Band, D.; van Paradijs, J.;Kouveliotou, C.; Preece, R.D.; et al.: “The Effect ofMagnetic Fields on Gamma-Ray Bursts Inferred FromMulti-Wavelength Observations of the Burst of 23January 1999”, Nature, Vol. 398 p. 394, 1999.Giblin, T.W.; van Paradijs, J.; Kouveliotou, C.;Connaughton, V.; Wijers, R.A,M.J.; Briggs, M.S.;Preece, R.D.; and Fishman, G.J.: “Evidence for anEarly High-Energy Afterglow Observedwith BATSEfrom GRB 980923”, ApJ, Vol. 524 p. L47, 1999.Mitrofanov, I.G.; Anfimov, D.; Litvak, M.; Briggs,M.S.; Paciesas, W.S.; Pendleton, G.N.; Preece, R.D.;and Meegan, C.A.: “Average Cosmological InvariantParameters of Cosmic Gamma-Ray Bursts”, ApJ, Vol.523 p. 192, 1999.Mitrofanov, I.G.; Litvak, M.; Briggs, M.S.;Paciesas, W.S.; Pendleton, G.N.; Preece, R.D.; andMeegan, C.A.: “Average Emissivity Curve of BATSEGamma-Ray Bursts with Different Intensities”, ApJ, ,Vol. 523 p. 610, 1999.Mitrofanov, I.G.; Anfimov, D.; Litvak, M.; Sanin,A.; Saevich, Y.; Briggs, M.S.; Paciesas, W.S.;Pendleton, G.N.; Preece, R.D.; Koshut, T.; Fishman,G.J.; Meegan C.A.; and Lestrade, J.P.: “The EmissionTime of Gamma-Ray Bursts”, ApJ, Vol. 522 p. 1069,1999.Crider, A.; Liang, E.P.; Preece, R.; Briggs, M.;Pendleton, G.; Paciesas, W.; Band D.; and Matteson,J.L.: “Spectral Hardness Decay with Respect toFluence in BATSE Gamma-Ray Bursts”, ApJ, Vol.519 p. 206, 1999.Preece, R.; Briggs, M.; Mallozzi, R.; Pendleton,G.; Paciesas W.; and Band, D.: “The SynchrotronShock Model Confronts a ‘Line of Death’ in theBATSE Gamma-Ray Burst Data”, ApJ, Vol. 506 p.L2, 1998.

Dr. V. SchönfelderRole in GBM: Co-InvestigatorEducationDiploma in Physics at University of Kiel, 1966,PhD in Physics atTechnische Universität München, 1970,Habilitation inExperimental Physics at TechnischeUniversität München, 1979, apl.Professor of Physics at Technische UniversitätMünchen, 1995.Role in GBMCo-I with responsibility to coordinate GBM issuesbetween MPE and DLR and between MPE andMSFC.ExperienceDr. Schönfelder started his carreer in cosmic rayphysics (cosmic-ray neutrons) and is working in thefield of gamma-ray astronomy since 1971. He ishead of the gamma-ray astronomy group at theMax-Planck-Institut für extraterrestrische Physik since1982. He has been the Principal Investigator of theCompton Telescope Balloon Programs at MPE(from 1971 to 1982), the Principal Investigator ofCOMPTEL aboard NASA’s Compton Gamma-RayObservatory (from 1979 till now), and one of thetwo Co-Principal Investigators of the SpectrometerINTEGRAL (SPI) Instrument (since 1995). He haspublished more than 300 papers in refereed journalsand conference proceedings. He has served on anumber of committees in Germany, of ESA and ofNASA. Since 1995 he is apl. Professor of Physics atthe Technische Universität München and as such,teaching courses on “astrophysics” for graduatesand undergraduates.Recent PublicationsWerner Collmar and Volker Schönfelder: “Evidencefor Massive Black Holes in the Nuclei of ActiveGalaxies from Gamma-Ray Observations.” Proc. ofHeraeus-Seminar, Phys. Soc. Bad Honnef (Aug.1997), BlackHoles: Theory and Observation,submitted to Springer (1998);Igor V. Moskalenko, Werner Collmar, andVolkerSchönfelder: “A Combined Model for the X-Ray to Gamma-Ray Emisssion of CygX-1.” ApJ, Vol.502 pp. 428–436, 1998 July 20;V. Schönfelder: “Gammastrahlung ausdemKosmos,”Physikalische Blätter 54, No. 4,Seite 325–330 (1998);Iyudin, A.F.; Schönfelder, V.; Bennett, K.; Bloemen,H.; Diehl,R.; Hermsen, W.; Lichti, G.G.; van derMeulen, R.D.; Ryan, J.; and Winkler, C.: “EmissionFrom 44 Ti Associated With a Previously UnknownGalactic Supernova.”, Nature, Vol. 396, pp. 142–144,12 November 1998.Aschenbach, B.; Iyudin, A.F.; and Schönfelder, V.:“Constraints of Age, Distance and Progenitor of theSupernova Remnant RXJ0852.0–4622/GRO J0852–4642, A&A, submitted March 3, 1999;Schönfelder, V.: “Prospects for the INTEGRAL SpectrometerSPI.”, LiBeB, Cosmic Rays and Related XCand Gamma Rays, ASP Conference Series, eds. R.Ramaty, E. Vangioni-Flam, M. Cassé and K. Olive,Vol. 171, pp. 217–225 (1999);Schönfelder V.; et al.: “The First COMPTEL SourceCatalogue.”, A&A Suppl., submitted August 1999.SocietiesDeutsche Physikalische GesellschaftAstronomische GesellschaftInternational Astronomical UnionHonors and AwardsNASA exceptional scientific achievement Award 1993,Deutscher Philip Morris Forschungspreis 1997

Dr. Andreas A. von KienlinRole in GBM: Co-InvestigatorEducationDiploma in Physics, University of Heidelberg,Germany, 1987; Dr. rer. nat.in Nuclear Physics, magnacum laude, GSI Darmstadt and University ofMainz,Germany, 1993Role in GBM:Development of the detector module electronics(voltage divider and pre-amplifier electrical design),preparation and execution of detector performancetests and GBM detector calibration.ExperienceDr. von Kienlin has been primarily working on thedevelopment of detectors for astrophysics, nuclear physicsand atomic physics applications. During his Ph.D thesisand Postdoc time at the heavy ion facility GSI inDarmstadt and the nuclear physics group at the Universityof Mainz, he developed a successful new detector type forthe energy-sensitive detection of heavy ions. The observedrelative energy resolution of the so called low temperaturedetectors (LTDs) for different heavy ion species ( 20 Ne, ...,209Bi) at different energies (3 MeV/u, ..., 100 MeV/u) wasin the DE/E =10 -3 range. Compared to conventionaldetectors, this is an order of magnitude improvement forvery heavy ions. In first nuclear physics experiments Dr.von Kienlin has shown the successful use of these newdetectors.During his EC-Fellowship in Genoa Dr. von Kienlin hasdeveloped superconducting transition edge sensors (TES)for high resolution x-raydetection and for the applicationin an neutrino physics experiment (b-decay of 187 Re).Dr. von Kienlin joined the Max-Planck Institutefürextraterrestrische Physik in April 1998 as a member ofthe INTEGRAL team of the gamma-ray astronomygroup.He is working on detector development, detectorperformance tests and the calibration of the spectrometerSPI. Furthermore he is engaged in the ACS burst detectionsystem.Dr. von Kienlin has published about 30 papers in refereedjournals and conference proceedings and presented about15 talks at internationali nstitutes and conferencesSocietiesGerman Physical Society ” Deutsche Physikalische-Gesellschaft” (DPG)Honors and AwardsScholarship of the Studienstiftung des deutschenVolkes – 1984-87European Community EC-Fellowship at thenational Italian nuclearphysics institute INFN inGenoa, Italy – 1996-98Recent PublicationsLichti, G.G.; Georgii, R.; von Kienlin, A.;Schönfelder, V.; Wunderer, C.; Jung, H.-J.; Hurley,K.: “The Gamma-Ray Burst-Detection system ofthe INTEGRAL-Spectrometer SPI”, 5 th ComptonSymposium, AIP Conference Proceedings (1999), inprintJean, P.; Vedrenne, G.; Schönfelder, V.; Albernhe,F.; Borrel, V.; Bouchet, L.; Caraveo, P.; Connall, P.;Cordier, B.; Denis, M.; Cozach, R.; Diehl, R.;Durouchoux, Ph.; Georgii, R.; Juchniewicz, J.; vonKienlin, A.; Knödlseder, J.; Larque, Th.; Lavigne,J.M.; Leleux, P.; Lichti, G.; Lin, R.; Mandrou, P.;Matteson, J.; Paul, Ph.; Roques, J.P.; Sanchez, F.;Schanne, S.; Skinner, G.; Slassi-Sennou, S.; Strong,A.; Sturner, S.; Teegarden, B.; vonBallmoos, P.;Wunderer, C.; “The spectrometer SPI of theINTEGRAL Mission”, 5 th Compton Symposium,AIP Conference Proceedings (1999), in print.von Kienlin, A.; Galeazzi, M.; Gatti, F.; Vitale,S.; “A Monolithic Superconducting Micro-Calorimeter for X-Ray Detection”, Nucl.Instr. andMeth. A 412 (1998) 135–139Meier, H.J.; Egelhof, P.; Henning, W.; Kienlin,A.V.; Kraus, G.; Weinbach, A.; “Low TemperatureBolometers for Experiments with Cooled Heavy IonBeams From Storage Rings”, Nucl. Phys. A 626(1997) 451cvon Kienlin, A.; Galeazzi, M.; Gatti, F.; Meunier,P.; Swift, A.M.; and Vitale, S.: ”X-Ray DetectionWith a Bulk Iridium Transition Edge Calorimeter”,Proc. 7 th Int. Workshop on Low TemperatureDetectors LTD–7, Munich 1997Egelhof, P.; Beyer, H.F.; McCammon, D.;Feilitzsch, F.V.; von Kienlin, A.; Kluge, H.J.;Liesen, D.; Meier, J.; Moseley, S.H.; Stöhlker, T.;“Applications of Low-Temperature Calorimeters forPrecise Lamb Shift Measurements on HydrogenlikeVery Heavy Ions”, Nucl. Instr.And Meth. A 370(1996) 263–265von Kienlin, A.; Azgui, F.; Böhmer, W.; Djotni, K.;Egelhof, P.; Henning,W.; Kraus, G.; Meier, J.; andShepard, K.W.:“High Resolution Detection of EnergeticHeavy Ions With a Calorimetric Low-TemperatureDetector”, Nucl. Instr. and Meth. 370 (1996) 815–818

Appendix BLetters of Endorsement

Appendix CDraft International Agreement

The Gamma-Ray Burst Monitor (GBM) is a NASAsecondary instrument on the Gamma-Ray Large AreaSpace Telescope (GLAST) Mission that is a follow-onto the Compton Gamma Ray Observatory. The primarygoal of the mission is to advance our understandingof the origin of the universe through study ofastrophysical and solar phenomenology at high energiesin the electromagnetic spectrum. The goal will beachieved through the development of advancedinstrumentation, that will observe gamma-ray sourcesover a very broad energy band.Pursuant to this Interim Agreement, DLR will usereasonable efforts to carry out the following responsibilities:• Provide overall program management for theGBM Instrument;• Design and build the detectors, low and highvoltage power supplies, and preamplifiers forthe detectors;• Qualify the flight articles for launch and operationin a low-Earth orbit;Participate in a Co-PI capacity along with the MPEscience team members in the science oversight roleof the GBM Instrument development and in scienceoperations. Provide the Co-PI as a participant in theScience Working Group at the facility level and inthe U.S. PI science team.Pursuant to this Interim Agreement, NASA will usereasonable efforts to carry out the following responsibilities:• Design, build, and launch a complete GLASTfacility to be integrated to a NASA providedspacecraft and launched with a NASA providedlaunch vehicle.Provide instrument integration• Provide mission operations development andmission operations including preliminary missiondesign, operations software development,and spacecraft tracking and operations support.• Provide science support for all mission phases.Once approved by both Parties, this InternationalAgreement will provide the framework under whichMPE will ship the hardware and technical data asdetailed above to NASA for use in GBM activities.Detailed arrangements for shipment and receipt ofthis equipment will be made between the NASAMarshall Space Flight Center (MSFC) and MPE.Once approved by both Parties, this InternationalAgreement will remain in effect until completion ofthe GLAST MissionBelow are the primary NASA and MPE contacts forthis agreement:NASA:Dr. Alan BunnerCode SRNASA HeadquartersWashington, DC 20546MSFC:Mr. Steven ElrodSpaceflight Experiments Group/SD21Marshall Space Flight Center, Alabama 35812MPE:Dr. Giselher LichtiMax Planck Institute for Extraterrestrial PhysicsGarching, GermanyIn order to proceed further both NASA and DLRagree as follows:1) To endeavor to conclude this government-togovernmentInternational Arrangement (MOU)and associated exchange of diplomatic notes asexpeditiously as possible, with a goal of completingthe same no later than June 30, 2000.2) The Parties are obligated to transfer only thosetechnical data (including software) and goodsnecessary to fulfill their respective responsibilitiesunder this agreement, in accordance withthe following provisions:a) The transfer of technical data (excluding software)for the purpose of discharging the parties’responsibilities with regard to interface, integra-

tion, and safety shall normally be made withoutrestriction, except as required by national lawsand regulations relating to export control or thecontrol of classified data. If proprietary but notexport-controlled design, manufacturing, andprocessing data and associated software isnecessary for interface, integration, or safetypurposes, the transfer shall be made and the dataand associated software shall be appropriatelymarked.b) All transfers of proprietary technical data andexport-controlled technical data and goods aresubject to the following provisions. In the eventa party finds it necessary to transfer goodswhich are subject to export control or technicaldata which is proprietary or subject to exportcontrol, and for which protection is to be maintained,such goods shall be specifically identifiedand such technical data shall be markedwith a notice to indicate that they shall be usedand disclosed by the receiving party and itsrelated entities (e.g., contractors and subcontractors)only for the purposes of fulfilling thereceiving party’s responsibilities under theprograms implemented by this agreement, andthat the identified goods and marked technicaldata shall not be disclosed or retransferred toany other entity without the prior written permissionof the furnishing party. The receivingparty agrees to abide by the terms of the notice,and to protect any such identified goods andmarked technical data from unauthorized useand disclosure, and also agrees to obtain thesesame obligations from its related entities prior tothe transfer. Nothing in this article requires theparties to transfer goods or technical data contraryto national laws and regulations relating toexport control or control of classified data.c) All goods, marked proprietary data, and markedor unmarked technical data subject to exportcontrol, which is transferred under this agreement,shall be used by the receiving partyexclusively for the purposes of the programsimplemented by this agreement.Nothing in this agreement shall be construed asgranting or implying any rights to, or interest in,patents or inventions of the Parties or theircontractors or subcontractors.3) All equipment and technical data transferred bythe Parties under this interim agreement shallremain the property of the originating Partyunless specified otherwise in this interim agreement.The Parties shall seek to arrange freecustoms clearance and waiver of applicablecustoms duties and taxes, for equipment andtechnical data imported into their respectivecountries under this interim agreement and, ifunable to make such arrangements, shall arrangeto pay same.4) Release of public information regarding thisprogram may be made by the appropriateagency for its own portion of the program asdesired and, insofar as participation of the otheris involved, after suitable consultation.5) NASA and DLR will each bear the costs ofdischarging its respective responsibilities asdefined in this agreement, including travel andsubsistence of its own personnel and transportationof all equipment for which it is responsible.6) It is confirmed that the Parties will conclude anInternational Arrangement (MOU) which willprovide that activities undertaken pursuant tothis Draft International Arrangement (MOU)shall be governed by the Agreement between theGovernment of the United States of Americaand the Government of Germany ConcerningCross-Waiver of Liability for Cooperation in theExploration and Use of Space for PeacefulPurposes (the “Cross-Waiver Agreement”), andshall be subject to the Exchange of Notesbetween the Governments concerning the Cross-Waiver Agreement and to the arrangementsbetween the Parties regarding subrogated claimsof the Government of the United States ofAmerica and the Government of Germany.a) With regard to activities undertaken pursuant tothe International Arrangement (MOU), DLRconfirms that the appropriate organization willpurchase adequate insurance to indemnify andhold harmless NASA, its employees, its related

entities (e.g., contractors, subcontractors, investigatorsor their contractors or subcontractors),and employees of its related entities againstclaims, including subrogated claims of theGerman Government, for injury to or death ofDLR and MPE employees or employees of itsrelated entities, or for damage to or loss ofDLR’s own property or that of its related entities,whether such injury, death, damage or lossarises through negligence or otherwise, exceptin the case of willful misconduct. NASA waivesall claims, including the subrogated claims ofthe United States Government, against DLR, itsemployees, its related entities (e.g., contractors,subcontractors, investigators or their contractorsor subcontractors), and employees of its relatedentities for any injury to or death of NASAemployees or employees of its related entities,or for damage to or loss of NASA property orthat of its related entities, whether such injury,death, damage or loss arises through negligenceor otherwise, except in the case of willfulmisconduct.(5) claims for damage based upon a failure of theParties to extend the provision as set forth aboveor from a failure of the Parties to ensure thattheir related entities extend the provision as setforth above;(6) contract claims between the Parties based onexpress contractual provisions.d) Nothing in the above shall be construed to createthe basis for a claim or suit where none wouldotherwise exist.b) The Parties further agree to use all reasonableefforts to extend this provision as set forth aboveto their own related entities by requiring them,by contract or otherwise, to waive all claimsagainst the other Party and its related entitiesagainst any claim for injury, death, damage orloss arising from activities undertaken pursuantto this agreement.c) This provision shall not be applicable to:(1) claims between a Party and its own relatedentity or between its own related entities;(2) claims made by a natural person, his/her estate,survivors or subrogees (consistent with paragraph6.(a) above) for bodily injury, otherimpairment of health, or death of such naturalperson;(3) claims for damage caused by willful misconduct;(4) intellectual property claims; or

Appendix DReference List

Reference ListAarnio, P.A.; et al.:1990, FLUKA User’s Guide CERN TIS-RP-190Akerlof, C.; et al.:1999 Nature 398, 400Band, D.L.; et al.: ApJ, 486, 928, 1997Band, D.L.; et al.: ApJ, 413, 281, 1993Barbier, M.: Induced Radioactivity, North-Holland Pub. Co, Amsterdam 1969Baring, M.G.; and Harding, S.K.: ApJ, 481, L85, 1997Barthelmy, S.D.; et al.: AIP Con. Proc. 428, eds. C.A. Meegan, R.D. Preece and T. M. Koshut, 99, 1997Bonnell, J.; et al.: AIP Conf. Proc. 428, eds. C.A. Meegan, R. D. Preece and T.M. Koshut, 349, 1997Brainerd, J.J.; Pendleton, G.N.; Mallozzi, R.; Briggs, M.S.; and Preece, R.D: CDROM, edition of the Proceedingsof the 20 Texas Symposium on Relativistic Astrophysics, 1999Briggs; et al.: ApJ 524, 82, 1999aBriggs; et al.: ApJS, 122, 503, 1999bBrun, R.; et al.: CERN Program Library Long Writeup W5013 (Geneva: CERN), 1993Catelli, J.R.; Dingus, B.L.; and Schneid, E.J.: AIP Conf. Proc. 428, eds. C.A. Meegan, R.D. Preece and T.M.Koshut, 309, 1997Cen, R.: ApJL, 517, L113, 1999Dingus, B.L.; et al.: AIP Conf. Proc. 428 eds. C.A. Meegan, R. D. Preece and T.M. Koshut, 884, 1997Fenimore, E.E.; et al.: ApJ, 448, L101, 1995Fishman, G.: NASA Contractor Report CR–150237, 1977Ford, L.A.; et al.: ApJ 439, 307, 1995Gehrels, N.: NIM, A313, 513, 1992Harris, M.J.; and Share, G.H.: AIP Conf. Proc. 428, eds. C. A. Meegan, R.D. Preece, and T.M. Koshut, 67,1997Hurley, K.; et al.: AIP Conf. Proc. 307, eds. G.J. Fishman, J.J. Brainerd and K. Hurley, 27, 1994Hurley, K.; et al.: ApJS, 122, 497, 1999Kommers, J. M.; et al.: ApJ, 491, 794, 1997Kouveliotou, C.; et al.: ApJ, 413, L101, 1993Lloyd, N.; and Petrosian, V.; ApJ, 511, 550, 1999Mallozzi, R.; et al.: ApJ, 454, 597, 1995Mazets, E.P.; and Golenetskii, S.V.: Astrophys. Space Sci., 75, 47, 1981Norris, J.; et al.: ApJ, 459, 393, 1996Paciesas, W.; et al.: ApJS, 122, 465, 1999Pendleton; et al.: ApJ 512, 362–376, 1999Piran, T.; and Narayan, R.: AIP conf. Proc. 384, eds. C. Kouveliotou, M.S. Briggs, and G.J. Fishman, 233,1996Preece, R.D.; et al.: ApJS, in press, 1999Rees, M.J.; and Meszaros, P.: MNRAS, 258, 41, 1992Sreekumar; et al.: The Astrophysical Journal, 494, 523, 1998Stern, B.; et al.; submitted to ApJ 1999Tavani, M; ApJ, 466, 768 1996Winkler C.; et al.: Ap&SS, 231, 153,1995

Appendix EAcronyms List

Acronym ListACSADCAGCAOBATSEBBBGOBSPECBTIMECDRCEASECEICGROCPUCRDDLRDoDDPUDRDREDVDEEEEGRETEGSEEMCEMIEM’sEPOESDESOCFITSFOVFPGA’sFTEGBMGCNGEANTGIGLASTGOGOPGRBGSEGSFCHETEHVHVPSIARanticoincidence subsystemanalog to digital converterautomatic gain controlannouncement of opportunityBurst and Transient Source Experimentbreadboardbismuth germanatebackground spectroscopybackground timecritical design reviewcompact environment anomaly sensorcontract end itemCompton Gamma-Ray Observatorycentral processing unitcritical design reviewDeutsches Zentrum fuer Luft- und RaumfahrDepartment of Defensedata processing unitdata requirementsdata receiver electronicsdigital video diskelectronic, electrical, electromechanicalenergetic gamma-ray experiment telescopeelectrical ground support equipmentelectromagnetic compatibilityelectromagnetic interferenceelectrical modelseducation and public outreachelectrostatic dischargeEuropean Space Operation Centerflexible image transport systemfield of viewfield programmable gate arraysfull time equivalentGLAST Burst Monitorgamma-ray coordinates networkGEometry ANd Tracking Monte Carlo Programguest investigatorgamma-ray large area space telescopeguest observerguest observer programgamma-ray burstground supply equipmentGoddard Space Flight Centerhigh energy transient explorerhigh voltagehigh voltage power supplyindependent assessment review

ICDIDLIGSEIOCIPIIPNIRI&TITTRLAD’sLATLED’sLVPSMLIMOCMOUMPEMSFCMUX’sNaINARNFIOSSEOWIPDRPERPMT’sPREPSRPVORFPROTSES&MASAASBDSGRSINDASMAPPSMMSOCSPISRRSTM’sTDR/DRTRASYSTTEUAHUTCUVWFCinterface control documentinteractive data languageinstrument ground support equipmentinstrument operations centerinstrument principal investigatorInterplanetary Networkinfraredintegration and testintegration/test readiness reviewlarge area detectorslarge area telescopelight emitting diodeslow voltage power supplymulti layer insulationmission operations centermemorandum of understandingMax Planck Institute for Extraterrestrial PhysicsMarshall Space Flight CentermultiplexerSodium Iodidenonadvocate reviewnarrow field instrumentsOriented Scintillation Spectroscopy Experimentorganizational work instructionpreliminary design reviewpre-environmental reviewphotomultiplier tubeprocessing electronicspreship reviewPioneer Venus Orbiterrequest for proposalRobotic Optical Transient Search Experimentsafety and mission assuranceSouth Atlantic anomalysmall disadvantaged businessessoft gamma repeatersystem improved differencing analyzersafety and mission assurance program plansolar maximum missionscience operations centerspectrometer on Integralsystem requirements reviewstructural test modelstest discrepancy record/discrepancy recordthermal radiation analyzer systemtime tagged eventsUniversity of Alabama in Huntsvillecoordinated universal timeultravioletwide field camera