Future Goals for Gamma-Ray Spectroscopy

Home , Gamma (satellite)

FUTURE GOALS FOR -RAY SPECTROSCOPY

P.von BALLMOOS

Centre d'Etude Spatiale des Rayonnements, Toulouse, France

Abstract. Gamma-ray lines are the ngerprints of nuclear transitions, carrying the mem-

ory of high energy pro cesses in the universe. Setting out from what is presently known

ab out line emission in gamma-ray astronomy, requirements for future telescop es are out-

lined. The inventory of observed line features shows that sources with a wide range of

angular and sp ectral extenthave to b e handled: the scienti c ob jectives for gamma-ray

sp ectroscopy are spanning from compact ob jects as broad class annihilators, over long-

lived galactic radioisotop es with hotsp ots in the degree-range to the extremely extended

galactic disk and bulge emission of the narrowe e line.

The instrumental categories which can b e identi ed in the energy range of nuclear astro-

physics have their origins in the di erent concepts of light itself: geometrical optics is the

base of mo dulating ap erture systems - these metho ds will continue to yield adequate p er-

formances in the near future. Beyond this, fo cusing telescop es and Compton telescop es,

based on wave- and quantum- optics resp ectively,may b e capable to further push the

limits of resolution and sensitivity.

Key words: gamma-ray astronomy { instrumentation

1. Intro duction

With the Compton Gamma-Ray Observatory and the GRANAT/SIGMA

telescop e in orbit, high energy astrophysics has for the rst time ever the

opp ortunity to study all the facets of celestial gamma-rays side by side. In

the energy range of nuclear transitions, the gamma-ray sky has b een mapp ed

on various angular scales and a large numb er of new gamma-ray sources has

b een discovered see reviews by Gehrels et al. 1994 and Paul 1995. Mayb e

the most imp ortant scienti c p otential of the present generation high ener-

gy telescop es is their extremely broad common energy coverage. Together

with the op erating X-ray and high energy gamma-ray telescop es, a quasi-

continuous coverage has op ened the p ossibility for multi-wavelength studies

of continuum sp ectra spanning from the keV- to the GeV-range. Figure 1

indicates that the present area once might b e called the \golden age of

gamma-ray astronomy": never b efore has the high-energy sky b een exam-

ined so thoroughly and over such a broad energy range. For many of the high

energy sources, multi-wavelength studies may b e the only way that leads to

an understanding of their complex source mechanisms. A mo del case is the

sp ectrum of the quasar 3C273 Lichti et al. 1995 that has b een observed

- partly simultaneously - from radio to gamma-ray energies. Nevertheless,

the gamma-ray telescop es on the Compton Gamma Ray Observatory and

on GRANAT also have raised new astrophysical questions and highlighted

those which remain unanswered. The future goals of gamma-ray astronomy

must b e de ned in this context. The progress of SIGMA, BATSE, OSSE and

2 P.von BALLMOOS

Fig. 1. The \golden age of gamma-ray astronomy" ? Never b efore the high-energy sky

has b een examined so thoroughly and over such a broad energy range.

COMPTEL is based primarily on skymaps, excellent timing analysis, and

mo derate to fair sp ectral resolution. The observations have revealed certain

asp ects of the morphology of celestial gamma-ray emitters, yet, the physical

pro cesses at work are often only p o orly understo o d. Frequently, the observed

sp ectra do not suciently constrain the emission mechanisms : explaining

a relatively simple, featureless continuum with a complex multiparameter

mo del can b e ambiguous, moreover, di erent comp onents may blend into

one another, each of them can dep end on various physical parameters in the

emitting region.

In manyways, the present situation resembles the situation of optical

astronomy in the b eginning of the last century: Back then, the available

observational data mainly consisted in images, starcounts, variability's, and

colors. Astrophysics was b orn in 1859 when G. Kirchhof and R. Bunsen

develop ed the sp ectral analysis and explained the Frauenhofer-lines in the

sp ectrum of the sun. The exploration of atomic and molecular lines has

since turned out to b e the most p owerful to ol for the study of the physical

conditions in celestial sources.

Up to to day, only little advantage has b een taken from the fundamental

astrophysical information contained in gamma-ray lines. The reason for this

is the mo dest energy resolution of the existing instruments. High resolution Spectro_ExAstr_95.tex - Date: May 16, 1995 Time: 22:50

FUTURE GOALS FOR -RAY SPECTROSCOPY 3

sp ectroscopy will therefore b e one of the ma jor goals of the next generation

of space b orne gamma-ray telescop es.

The scienti c p otential of nuclear gamma-ray astronomy is outlined in

section 2. In sections 3 instruments for sp ectroscopy in the low and medi-

um gamma-raychannel are presented: mo dulating ap erture systems, Comp-

ton telescop es and di raction lens telescop es. The three techniques actually

re ect our current p erception of light itself - they are based on the principles

of geometrical optics, quantum optics and wave optics, resp ectively.

2. The promise of nuclear gamma-ray astronomy

Ultraviolet, visible and infrared-sp ectroscopy has long b ecome an extreme-

ly valuable technique in astronomy. While optical lines re ect structural

changes in the electron shell of atoms, caused by collisions with energies of

the order of 10 eV T 10 K, transition b etween discrete nuclear energy

7 9

levels imply MeV energies T 10 to 10 K, corresp onding to the bind-

ing energy of nucleons. Collision energies of this order are characteristic for

the temp eratures inside stars, particles accelerated by electromagnetic elds

in solar ares, or interactions of cosmic ray particles with the interstellar

medium. Gamma-ray lines are the ngerprints of nuclear transitions, car-

rying the memory of high energy pro cesses in the universe. High resolution

gamma-ray sp ectroscopy is a unique to ol to identify the presence of exited

nuclei, quantitatively determine their abundance's, and provide insightinto

the physical conditions of the source regions temp erature, density, gravita-

tional or cosmological energy shifts.

The scienti c p otential of gamma-ray sp ectroscopy includes b etter under-

standing of the chemical evolution of the Galaxy by mapping the sites of

recentnucleosynthetic activity novae, SN, massive stars. The observation

of sup ernovae in the galaxies of our lo cal cluster and closeby sup erclusters

Virgo cluster will verify our mo dels of nucleosynthesis in massive stars.

Sp ectral features in solar ares, cyclotron lines in gamma-ray bursts and

pulsars are among the further challenges of gamma-ray sp ectroscopy.

An inventory of observed gamma-ray lines is presented in table I, it

re ects the current status of nuclear gamma-ray astronomy. Note that most

of the knowledge on gamma-ray lines is based on observations made with

scintillation detectors that typically feature energy resolutions of 10. Nev-

ertheless, these elementary sp ectroscopic measurements already indicate the

tremendous p otential of gamma-ray line emission. Performance requirements

for future sp ectroscopy missions b ecome conspicuous when the measured/-

anticipated line uxes are compared to the exp ected angular scales. Figure

2 indicates that emissions with a wide range of angular and sp ectral extent

are exp ected, varying in intensityby several orders of magnitude. The sci-

enti c ob jectives for gamma-ray sp ectroscopy are spanning from compact Spectro_ExAstr_95.tex - Date: May 16, 1995 Time: 22:50

4 P.von BALLMOOS

TABLE I

Observed celestial gamma-ray line features

Physical Pro cess Energy Source Flux Instrument,

2 1

[keV] ph cm s Detector

Nuclear deexcitation

Fep,p', 847 Solar ares 0.05 SMM, NaI

Mg p,p', 1369 Solar ares 0.08 SMM, NaI

Ne p,p', 1634 Solar ares 0.1 SMM, NaI

Si p,p', 1779 Solar ares 0.08 SMM, NaI

12Cp,p', 4439 Solar ares 0.1 SMM, NaI

4439 Orion Comp. 510 COMPTEL, sc

Op,p', 6129 Solar ares 0.1 SMM, NaI

6129 Orion Comp. 510 COMPTEL, sc

Radioactive decay

56 56 3

CoEC, Fe 847, 1238 SN 1987A 10 b Ge det's and

2598 var. scint's,

b COMPTEL, sc 847,1238 SN 1991T 510

57 57 4

CoEC, Fe 122, 136 SN 1987A 10 OSSE, NaI-CsI

44 44 + 5

TiEC Sc 1157 Cas A SNR 710 COMPTEL, sc

26 + 26 4

Al Mg 1809 gal. plane 410 COMPTEL, sc

1809 Vela SNR 1-610 COMPTEL, sc

e e Annihilation

511 Gal. bulge 1.710 OSSE, NaI-CsI

511 Gal. disk 4.510 OSSE, NaI-CsI

480120 a 1E 1740-29 c 1.310 SIGMA, NaI

511 Solar Flares 0.1 SMM, NaI scint.

47918 a Nova Muscae 6.310 SIGMA, NaI

400-500 a Bursts ? c 70 various scint.

44010 a Crab PSR ? c 310 FIGARO, NaI

Neutron Capture

1 2

Hn, H 2223 Solar ares 1 SMM, NaI

56 57 2

Fen, Fe 5947 a 6/10/1974 tr. 1.5 10 SMM, NaI

Cyclotron Lines

20-58 Hercules X-1 310 various scint.

legend: sc liquid scintillator/NaI scintillator; a Redshifted line; b Maximum emission,

c detection uncertain, the feature has not b een con rmed yet

sources as broad class annihilators, over long-lived galactic radioisotop es

with hotsp ots in the degree-range to the extremely extended galactic disk

and bulge emission of the narrowe e line. Spectro_ExAstr_95.tex - Date: May 16, 1995 Time: 22:50

FUTURE GOALS FOR -RAY SPECTROSCOPY 5

Fig. 2. Future sp ectroscopy missions have to face emissions with a wide range of angular

extent, and with intensities di erentby several orders of magnitude. The anticipated

ux for extragalactic SNe of typ e 1 has b een deduced from the COMPTEL detection of

SN1991T Morris et al. 1995 and by scaling its Co 847 keV gamma-ray ux with the

optical p eak magnitude of observed SNIa.

According to our present view of celestial gamma-ray sources in the ener-

gy range of nuclear transitions, narrow lines seem to b e generally emitted

from extended distributions while broad lines tend to b e radiated by com-

pact sources.

3. Instruments for gamma-ray sp ectroscopy

A natural rst ob jective for a future gamma-ray sp ectroscopy mission is the

mapping of the relatively intense sources on the upp er right of gure 2

4 2 1 6 2 1

which are typically emitting 10 ph cm s to a few 10 ph cm s .

Candidate sources of this intensity are mostly galactic and include the sites

of recentnucleosynthesis, regions of e e annihilation and clouds where

nuclear de-excitation by energetic particles takes place. Some of them might

app ear as extended structures: either b ecause of their apparently di use

origin - such as the narrow 511 keV line Purcell et al. 1994 - or b ecause

they are relatively closeby as the nucleosynthesis sites in the lo cal spiral

arm del Rio et al. 1994. An instrument that is adequate for this kind of

6 2 1

ob jectives should provide a sensitivity of several 10 ph cm s , a wide Spectro_ExAstr_95.tex - Date: May 16, 1995 Time: 22:50

6 P.von BALLMOOS

eld of view and an angular resolution in the degree range. Such a pro le

corresp onds to the exp ected p erformances of the SPECTROMETER on

ESA's INTEGRAL mission.

3.1. Spectroscopy with modulating aperture systems

Our present understanding of the low energy <1 MeV gamma-ray sky

has b een acquired mainly by mo dulating ap erture systems. The underlying

concept of this typ e of instrument is geometrical optics, this is, the source

photons are considered as traveling on rectilinear paths only. Light is passing

through op en elements of an otherwise opaque ap erture. The ap erture sys-

tems - consisting of masks or collimators - mo dulate the signal which then

reaches the detection plane designed to discern shadow patterns of some kind

spatial or temp oral. Since the photo electric e ect is the predominantmode

of gamma-rayinteraction in medium- and high-Z materials, these systems

are well adapted to the low energy gamma-raychannel. Pinhole cameras,

rastering collimators, co ded masks and mo dulation collimators b elong to

this category of instruments.

Two main classes of mo dulating ap erture systems can b e identi ed, accord-

ing to whether the signal is enco ded by spatial mo dulation co ded mask

telescop es or by temp oral mo dulation mo dulation collimators. These two

typ es stand for a whole sp ectrum of devices mixing the basic concepts of

spatial and temp oral mo dulation. Mo dulating ap erture systems of b oth

classes can b e used to pro duce images of the sky if the system allows to

observe/resolve an extended region simultaneously or quasi simultaneous-

ly. This multiplex advantage improves the survey sensitivity with resp ect

to plain rastering collimators `on-o ' techniques.

Op erational satellite instruments based on the principles of geometrical

optics include SIGMA, OSSE and BATSE Paul et al. 1991, Johnson et

al. 1993, and Fishman et al. 1989. While the detection plane of all three

of them is based on scintillators E=E 10, di erent ap erture systems

are used. Whereas OSSE and BATSE p erform temp oral mo dulation with

collimators and the earth an \anticollimator" resp ectively, SIGMA is a

multiplexing device using spatial mo dulation with a co ded mask.

High resolution sp ectroscopy with a co ded mask instrument will b e avail-

able by the b eginning of the next decade with ESA's INTEGRAL mission.

The INTEGRAL SPECTROMETER describ ed b elow corresp onds to the

ESA prop osal by a consortium b etween the CESR Toulouse, MPE Garch-

ing, CEA Saclay, IFCTR Milan, IEEC Barcelona, Univ. of Louvain,

Univ. of Valencia, Univ.of Birmingham and US Institutes. Figure 3a gives a

cutaway view of the SPECTROMETER. The instrument features a compact

array of high purity Germanium detectors entirely shielded by a BGO/CSI

anticoincidence system. Spectro_ExAstr_95.tex - Date: May 16, 1995 Time: 22:50

FUTURE GOALS FOR -RAY SPECTROSCOPY 7

Fig. 3. 3 instruments Spectro_ExAstr_95.tex - Date: May 16, 1995 Time: 22:50

8 P.von BALLMOOS

The Ge detector assembly consists of 19 high purity n-typ e germanium

detectors co oled to 85K via pulsed tub es supplied bytwo cryogenic com-

pressors. With a total geometric detection area of 500 cm the detection

plane is the largest one that could b e tted into the overall constraints of

the INTEGRAL mission. The ap erture subsystem features a tungsten co d-

ed mask HURA with 127 elements 171 cm ab ove the detector. The co ded

mask together with the detector plane de nes a fully co ded eld of view of

16 16 .

The narrow line sensitivity shown in Fig. 4 has b een obtained by com-

pletely mo deling the SPECTROMETER for the radiation environment con-

ditions exp ected outside the magnetosphere. The sensitivity is based on mo d-

els for the instrument eciency GEANT and background in Ge detectors

Naya et al. 1995. The anticipated p erformance for narrow line sp ectroscopy

is characterized by an energy resolution in the parts p er thousand range typ-

ically 2 keV at 1 MeV, an angular resolution of order 2 , and a sensitivity

6 2 1

2-510 ph cm s in the energy range relevant for nuclear astrophysics.

5 2 1

The 511 keV line sensitivity is of the order of 210 ph cm s .

3.2. Spectroscopy with Compton telescopes

While the total cross section has its minimum at a few MeV, the ener-

gy range b etween several hundred keV and ab out 30 MeV is dominated

by the Compton e ect. At these energies mo dulating ap erture systems run

into several problems: the eciency of the mo dulation decreases and var-

ious background problems related to heavy shielding b ecome increasingly

prohibiting. Hence, making use of the quantum nature of the photon inter-

actions is an inevitable choice for instruments at medium and high energy

gamma-rays. The idea to make use of the Compton e ect instead of ghting

it with thicker shielding and mo dulators has stimulated several groups and

resulted in a distinct class of imaging instruments.

In a Compton telescop e the incident gamma-ray is identi ed by successive

interactions in the two detector layers D and D . Compton scattering in D

1 2 1

is favored when low Z material is chosen. Total absorption of the scattered

photon in D can b e exp ected when high Z materials are used. Since the

D ^D coincidence condition discriminates against most of the internal n

1 2

background, a Comptom telescop e has an extremely low background. At

the same time, this coincidence condition causes the relatively low detection

eciency. In order to further reduce the background due to upward scattered

events, the time-of- ightbetween the two detectors is usually measured in

Compton telescop es. Neutron interactions can b e identi ed by measuring

the pulse shap e of the scintillation pulse.

From the interaction lo cation in D and D , the direction ; of the

1 2

scattered gamma-ray is obtained; the energy dep osits in D and D are E

1 2 1

and E resp ectively. The scatter direction ; , together with the amounts

2 Spectro_ExAstr_95.tex - Date: May 16, 1995 Time: 22:50

FUTURE GOALS FOR -RAY SPECTROSCOPY 9

of energy dep osited in the twointeractions can b e used to reconstruct the

arrival direction of the gamma-ray. The Compton equation allows to express

the scatter angle ' as a function of the energy-dep osits E and E :

1 2

2 2

m c m c

e e

+ with m c = 511keV ' = 1 cos

E E + E

2 1 2

If E and E are measured without systematic errors E = E + E = E ,

1 2 tot 1 2

the derived scatter angle ' equals the true Compton scatter angle '. The

arrival direction of the incident gamma-ray can then b e con ned to lie on a

cone-mantle with axis ; and op ening angle ' see g 3b.

With COMPTEL on the Gamma-ray Observatory, the concept has de -

nitely proven its unique p otential for MeV gamma-ray astronomy Schonfelder

et al. 1993.

The ATHENA concept Kroger et al. 1995, Johnson et al. 1995, Kurfess

et al. 1994 is based on D and D layers using Germanium planar strip

1 2

detectors providing 2-3 keV sp ectral resolution and spatial resolution of 2

mm. Such detectors, typically 5 cm 5cm1 cm, are available to day and

mightbeintegrated into large panels in the future. The ATHENA concept

2 2

foreseesa1m D layer consisting of one panel, and a 1 m D layer of

1 2

four panels, each panel containing 400 Ge strip detectors. Figure 3b shows

aschematic diagram of a solid state Compton telescop e combined with a

co ded mask for low energy gamma-rays.

In the Compton mo de 300 keV-10 MeV such an instrument can achieve

angular resolutions of 0.2 - 0.3 within a eld of view of typically one stera-

7 2 1

dian, and a narrow line sensitivity of several times 10 ph cm s ab ove

1 MeV see g. 4.

3.3. Spectroscopy with diffraction lens telescopes

Pushing the p erformances even further leads to several apparently unsolv-

able problems. Up to now, b etter instruments generally were bigger instru-

ments. Yet, physical limits for the size of spaceb orne instruments are b eing

6 2 1

attained to day.Achieving sensitivities b etter than 10 ph cm s and res-

olutions b etter than fractions of a degree seems to b e principally imp ossible

with the presently practiced instrumental concepts: even larger collection

areas are synonymous with larger detectors and thus higher background

noise. The ensuing mass/sensitivity dilemma can ultimately only b e over-

come by a radically di erent approach for detecting gamma-radiation.

Being used to accept that it is `imp ossible to re ect or refract gamma

rays', high energy astronomy has never considered relying on the wave-

quality of gamma-ray photons. The fo cussing di raction lens telescop e Smither

et al. 1995, von Ballmo os et al. 1995 makes for the rst time use of the

p erio dic nature of light and its interference with the p erio dic structure of a

crystal. Featuring a Laue di raction lens, it is designed to collect gamma- Spectro_ExAstr_95.tex - Date: May 16, 1995 Time: 22:50

10 P.von BALLMOOS

Fig. 4. Comparison of 3 nar-

row line sensitiviti es:

HEAO3 achieved sensitivity.

For all other instruments

T =10 sec.

obs

The hatched area outlines the

sensitivity of the INTEGRAL

SPECTROMETER. Low-

er and upp er limits are for an

optimistico/p es simi sti cp

mo del: o including the mul-

tiplexing advantage i.e. sensi-

tivity for any galactcic source

during a 10 sec survey of the

inner galaxy; p includin g a

background comp onent.

The shaded area indicates dif-

ferent background estimates

for the di raction lens tele-

scop e.

ray line photons on a large e ective area and fo cus them onto a Germanium

detector matrix with a small equivalentvolume for background noise.

The instrument has rst b een prop osed as a ballo on-b orne telescop e with

the lens tuned to di ract 511 keV photons only von Ballmo os and Smither

1994. Such a con guration makes p ossible the study of galactic `micro-

quasars' and other broad class annihilators in the lightofe e annihilation

during a ballo on ight. The p erformances of this Ge-lens/Ge-matrix system

have b een veri ed in summer 1994 during lab oratory measurements with a

ground based prototyp e Naya et al. 1995.

Ultimately however, the concept should b e put to use in space where

longer exp osures and steady p ointing would result in outstanding sensi-

tivities. A `tunable crystal di raction lens' von Ballmo os et al. 1995 can

observeany identi ed source at any selected line-energy in the cardinal range Spectro_ExAstr_95.tex - Date: May 16, 1995 Time: 22:50

FUTURE GOALS FOR -RAY SPECTROSCOPY 11

of nuclear transitions 200 keV - 1300 keV. The `sequential' op eration mo de

resulting from such a concept makes the sites of explosivenucleosynthesis

natural targets for a tunable crystal di raction telescop e.

While the evidence for p oint like sources of narrow gamma-ray line emis-

sion has b een mostly implicit at this p oint { b esides of the sup ernovae 1987A

Matz et al. 1988 and 1991T Morris et al. 1995 { various ob jects like galac-

tic novae and extragalactic sup ernovae are predicted in the area at the lower

left of gure 2. These sources should have small angular diameters but very

low uxes - mostly b ecause such ob jects are relatively rare and therefore are

more likely to o ccur at large distances. The instrumental requirements for

exploring this kind of sources match with the anticipated preformance of a

crystal di raction telescop e:

The imaging capabilities of a crystal di raction telescop e are de ned by

its b eamwidth which is identical to the eld of view of the lens: for compact

sources discrete p ointings of the ob ject will b e an appropriate observation

mo de, while extended structures as for example the jets of galactic micro-

quasars will b e scanned with the narrow b eam. The eld of view/angular

resolution will typically b e 15 FWHM. Below 1 MeV, the narrow line

7 2 1

sensitivity is exp ected to b e a few 10 ph cm s see gure 4.

4. Conclusion

With the sp ectrometer on ESA's INTEGRAL mission, a co ded mask instru-

ment for gamma-ray sp ectroscopy will b e available by the b eginning of the

next decade. The foremost ob jectives of the INTEGRAL sp ectrometer will

4 2 1

b e the mapping of gamma-ray line sources emitting 10 ph cm s to a

6 2 1

few times 10 ph cm s . Many of these p otential sources will b e galactic.

Some of them might app ear as extended structures - either b ecause of their

truly di use origin or b ecause they are relatively closeby as the nucleosyn-

thesis sites in the lo cal spiral arm. An energy resolution of E=E 500, a

wide eld of view and a mid-scale angular resolution make the INTEGRAL

sp ectrometer adequate for such ob jectives.

On a more distant horizon, exp erimental gamma-ray astronomy has to

nd ways to further push the limits of resolution and sensitivity: At ener-

gies ab ove one MeV, Compton telescop es can provide angular resolutions

7 2 1

of fractions of degrees and sensitivities of a several 10 ph cm s .At

energies b elow the MeV, tunable crystal di raction telescop es can achieve

7 2 1

sensitivities of a few times 10 ph cm s and angular resolutions of the

order of 15 . Spectro_ExAstr_95.tex - Date: May 16, 1995 Time: 22:50

12 P.von BALLMOOS

References

del Rio E. et al., 1994, AIP Conf. Pro c. 304, p. 171, ed. Fichtel C. et al., New York

Fishman G. et al., 1989 in Johnson, W.N. Ed, Pro c GRO Science Workshop, p.239

Gehrels N. et al., 1994, ApJ Sup. S., 92, 351

INTEGRAL ESA SCI931, ESA phase A study rep ort, April 1993

Johnson, W.N. et al., 1993, ApJ Sup. S., 86, 693

Johnson, W.N. et al., 1995, pro c. of workshop `Imaging in High En. Astron.', Capri 1994

Kurfess J.D. et al., NASA prop osal for new mission concepts in Astrophysics, NRA 94-

OSS-15, 1994

Kroger R.A. et al., 1995, pro c. of workshop `Imaging in High En. Astron.', Capri 1994

Lichti G. et al., Astron Astrophys 1995, in press

Matz S.M. et al., 1988, Nature 331, 416

Morris D. et al. 1995, pro c. 17th Texas Symp osium on Relativisti c Astrophysics

Murphy R.J. et al. 1990, ApJ ,358, 290

Naya J., et al., 1995a, submitted to NIM

Naya J. et al. , 1995b, pro c. of workshop `Imaging in High En. Astron.', Capri 1994

Paul J. et al., 1991, Adv.Space Res., 11 6, 289

Paul J. et al., 1995, in Signore M. et al Ed, The Gamma Ray Sky with Compton GRO

and SIGMA, p. 15

Purcell W.R et al., 1994, AIP Conf. Pro c. 304, p. 403, ed. Fichtel C. et al., New York

Schonfelder, V., Hirner, A. and Schneider, K.: 1973, Nucl.Instrum. Meth., 107, 385.

Smither R.K. et al., 1995, pro c. of workshop `Imaging in High En. Astron.', Capri 1994

von Ballmo os P. and Smither R.K. 1994, ApJ Sup. S., 92, 663

von Ballmo os P. et al, 1995, pro c. of workshop `Imaging in High En. Astron.', Capri 1994 Spectro_ExAstr_95.tex - Date: May 16, 1995 Time: 22:50