COMPTEL Reloaded: a heritage project in MeV astronomy
MMem. S.A.It. Vol. 0, 0 c (cid:13) SAIt 2008
Memorie della
COMPTEL Reloaded: a heritage project in MeVastronomy
Andrew Strong and Werner Collmar Max-Planck-Institut f¨ur extraterrestrische Physik, Garching, Germany e-mail: [email protected],[email protected]
Abstract.
COMPTEL was the Compton telescope on NASA’s Compton Gamma RayObservatory CGRO launched in April 1991 and which was re-entered in June 2000.COMPTEL covered the energy range 0.75 to 30 MeV, and performed a full-sky surveywhich is still unique in this range, with no followup mission yet approved. This remainsa major uncharted region, and the heritage data from COMPTEL are still our main sourceof information. Data analysis has continued at MPE however, since the data were neverfully analysed during the mission or in the period following, and improvements in analysistechniques and computer power make this possible.
1. Introduction
COMPTEL was the Compton telescope onNASA’s Compton Gamma Ray ObservatoryCGRO launched in April 1991 and which wasre-entered in June 2000. COMPTEL coveredthe energy range 0.75 to 30 MeV, and per-formed a full-sky survey which is still uniquein this range, with no followup mission yet ap-proved. This remains a major uncharted region,and the heritage data from COMPTEL are stillour main source of information in the MeVrange. Data analysis has continued at MPEhowever, since the data were never fully anal-ysed during the mission or in the period follow-ing, and improvements in analysis techniquesand computer power make this desirable andpossible.
2. COMPTEL Mission and Instrument
COMPTEL was a double-scatter Comptontelescope: incoming γ -rays Compton-scatter in one of the 7 upper organic liquid-scintillatorD1 detectors, and are absorbed in one of the14 the lower NaI D2 detectors, see Fig 1.Both D1 and D2 use photomultipliers to mea-sure the light signal and locate the scatter po-sition using the Anger-camera principle. Theenergy deposits give the Compton scatter an-gle according to the usual formula. Hence theincoming direction is determined to an an-nulus on the sky, whose width depends onthe precision of the energy and position mea-surements. At high energies the absorptionin D2 is incomplete, so the response is cor-respondingly broadened. The angular resolu-tion of the Compton scatter angle is about 2 ◦ .The distance between D1 and D2 is 1.577m,allowing a time-of-flight (TOF) discrimina-tion for upward-moving background γ -rays .A plastic-scintillator anticoncidence dome sur-rounding the instrument reduces the charge-particle background. In addition a pulse-shape-discrimination (PSD) measurement is used forbackground rejection. Nevertheless, the data a r X i v : . [ a s t r o - ph . H E ] J u l trong, Collmar: COMPTEL 1 Fig. 1.
The COMPTEL instrument and princi-ple of operation.
Fig. 2.
COMPTEL sky exposure, cm s.Galactic coordinates, centred on l =
0, b = ◦ ,covering the entire sky, as shown in Fig 2.COMPTEL also used a sophisticatedmulti-user analysis system called COMPASS,based on an Oracle database, which allowedfull traceability of all software and analysisoperations, shared over the community of in-stitutes involved. For full details of the in-strument and mission see Schoenfelder et al.(1993). Remarkably, the full mission wasnever analysed due to new projects takingaway people-power, despite the huge invest-ment which CGRO represented. Hence the fullscience potential of this major mission wasnever realized.The main achievements of the COMPTELmission included Galactic, extragalactic sources (Schoenfelder et al. (2000)), Almaps, GRBs, solar flares, and the extragalacticdi ff use γ -ray background.Meanwhile work has continued, for exam-ple one of the strongest sources discovered byCOMPTEL was in the Galactic plane at l = ◦ .It was long suspected that it could be identi-fied with the binary LS 5039 since this sourcehas particular properties, but a proof was lack-ing. Using the full COMPTEL mission data, itwas possible finally to establish the identifica-tion of this source with LS5039 on the basis ofthe detection of its X-day orbital modulation(Collmar et al. (2014)).
3. Recent new developments in datapreparation
1. The TOF selection window was adapted tothe newly generated TOF values (version TOF-VI), which take into account the electronic dif-ferences of the individual D1 and D2 modules,leading to a generally narrower ToF distribu-tion. Hence a narrower window can be definedwhich improves the background rejection e ffi -ciency.2. The PSD window was optimized as functionof energy.3. New energy ranges were defined; the orig-inal ranges were just ad-hoc numerical values0.75–1, 1–3, 3–10 and 10–30 MeV (apart from Al which was chosen to cover the 1.8 MeVline). New ranges were defined which betteravoid instrumental background lines and timevariations in the response: 0.9–1.7, 1.7–4.3,4.3–9 and 9–30 MeV.The original analyses used a resolution in thecomputed Compton scatter angle (known asphibar) of 2 ◦ . This was a compromise dueto the limitations of computer resources, andleads to some degradation of the response.Now finer binning in phibar is not a problem,for example 1 ◦ . Also the event binning andskymap binning can be finer than the original1 ◦ . Thus there will be no loss of potential an-gular resolution.4. Full mission data: this is a very significantdevelopment, the entire mission never havingbeen fully analysed previously. Only the firstfew years were fully analyzed. CGRO under- Strong, Collmar: COMPTEL went two orbital reboosts which a ff ected theinstrumental background; now we can analyseall data including that after the second reboostwhere the background was maximum.5. Parts of the COMPASS data-analysis systemwas ported to Linux from the original Solarisenvironment. Thus reprocessing of the eventdata to the datasets needed for science can becarried out routinely with new parameters, in afraction of the original time.
4. New developments in imaging
The “classic” maximum-entropy deconvolu-tion method (Skilling (1989)) implementedin the MEMSYS5 software package (Skilling(1999)) which is the basis for most of thepublished images (Strong et al. (1999)), wasadapted to use modern fast convolution-on-the-sphere methods (via spherical harmon-ics), and current parallel architechtures. Theskymaps use the HealPix all-sky equal-areapixelisation, which can be visualized with theCDS Strasbourg Aladin interactive sky atlas,.Images can be generated in a fraction of theoriginal time and with finer angular resolu-tion. The data are maintained in the instrumen-tal system to facilitate accurate response andbackground templates. The background tem-plate is based on a number of high-latitude ob-servations, fitted with time-dependent scalingfactor. Occasional variations in the responsePSF, due e.g. to solar mode, are included onan observation basis.Fig 3 shows preliminary maximum-entropy all-sky images in the four new energyranges. The Galactic plane is the most strikingfeature, while known sources both Galacticand extragalactic are visible. The extendedfeature below the plane in the fourth quad-rant is thought to be a residual backgrounde ff ect. Detailed evaluation of these images isongoing.
5. New source catalogue
The original COMPTEL source catalogue(Schoenfelder et al. (2000)) contained 32steady sources, both Galactic and extragalac-tic. The new full-mission data and analysis are enabling the production of a new cataloguewith significantly more sources and spectraland temporal measurements.
6. Outlook
The skymapping will be completed and ex-tended to include Al and other lines of in-terest. These maps will eventually be madeavailable to the community. Interpretation ofthe di ff use continuum emission in the contextof cosmic-ray models (Bouchet et al. (2011);Strong (2011)) and combined with Fermi,INTEGRAL data is foreseen. Acknowledgements.
We thank Martin Reinecke(MPA Garching) for his assistance in adapting themaximum entropy imaging software as described inthis paper.
References
Bouchet, L., Strong, A.W.., Porter T.A.,Moskalenko., I.V., Jourdain, E., Roques, J.-P., 2011, ‘Di ff use emission measurementwith INTEGRAL / SPI as indirect probe ofcosmic-ray electrons and positrons’, ApJ,739, 29Collmar, W., Zhang, S., 2014, ‘LS5039 - thecounterpart of the unidentified MeV sourceGRO J1823-12’, A&A, 565, 38Schoenfelder, V., et al., 1993, ‘Instrument de-scription and performance of the ImagingGamma-Ray Telescope COMPTEL aboardthe Compton Gamma-Ray Observatory’,ApJS 86, 657Schoenfelder, V., et al., 2000, ‘The firstCOMPTEL source catalogue’, A&AS, 143,145SSkilling J., 1989, ‘Classical MaximumEntropy’, in Maximum Entropy andBayesian Methods, Kluwer:Dordrecht,ISBN 0-7923–2 24-9, p.45Skilling, J., 1999, MEMSYS5 Users Manual.MEDC Ltd.Strong, A.W., et al., 1999, ‘COMPTELSkymapping: a new approach usingparallel computing’, AstrophysicalLetters and Comm unications, 39, 209.http: // / ∼ aws / publications / strong taormina paper15.ps trong, Collmar: COMPTEL 3 Fig. 3.
Preliminary COMPTEL all-sky images using the new maximum-entropy method.Galactic coordinates, centred on l =
0, b =
0. Left to right, top to bottom: 0.9–1.7, 1.7–4.3,4.3–9 and 9–30 MeV.Strong, A.W., 2011, ‘Interstellar gamma raysand cosmic rays: new insights from Fermi-LAT and INTEGRAL’, World Scientific,ISBN: 978-981-4462-40-2, page 473,http: // / doi / abs / //