GRB Trigger The SVOM / ECLAIRs Stéphane Schanne presentation at the 2nd SVOM Workshop 24-28 April 2017 Pingtang, Guizhou, China Stéphane Schanne Service d’Astrophysique CEA Saclay / IRFU UGTS ECLAIRs SVOM

SVOM objectives Space-based multi-band astronomical Variable Objects Monitor devoted to GRBs Launch: Dec. 2021 Operations: 3+2 years Aim of GRB science studies : GRB phenomenon, physics, progenitor GRB use for cosmology & fund. physics  Need of a complete sample of GRBs, with spectral and temporal coverage and distance measurement Scientific requirements of SVOM Trigger on all known types of GRBs (>200 in 3 years) in particular X-ray rich GRBs, high-z GRBs and under-luminous GRBs Provide fast, reliable and accurate positions of GRBs, and send alerts to the community Measure the broadband spectral shape (from visible to MeV) and temporal properties of GRBs prompt emission Study the afterglows and provide quick arcsec positions of GRBs Provide redshift indicators for GRBs [2] – SVOM workshop, April 2017, Qiannan, China – S. Schanne Trigger

SVOM system Trigger Space-based Variable Objects Monitor observation of GRB prompt emission ECLAIRs g imager Trigger GRB detection & localization  Alert via VHF to ground  Satellite slew [3] – SVOM workshop, April 2017, Qiannan, China – S. Schanne GRM g spectrometer GWAC Wide Angle Cameras VHF Alert network 3

SVOM system Space-based Variable Objects Monitor observation of GRB afterglow emission MXT X-ray telescope [4] – SVOM workshop, April 2017, Qiannan, China – S. Schanne VT Visible Telescope GFTs Ground Follow-up Telescopes 4

ECLAIRs gamma-ray imager CNES, IRAP, CEA, APC Allocations: Mass <90 kg, Power <90 W Detection plane: - DetArea: 1024 cm² - Detectors: 6400 CdTe pixels (4x4x1 mm3) - Energy range: 4-150 keV Mask: Ta, 40% open, self supporting Shield: PbCuAl - EffArea ~ 400 cm² @ 10-70 keV - EffArea > 200 cm² @ 6 keV - FoV = 2 sr (total) - Localisation accuracy <12' (at detection limit, 90% C.L.) ECLAIRs trigger & control unit (UGTS) [5] – SVOM workshop, April 2017, Qiannan, China – S. Schanne

ECLAIRs gamma-ray imager CNES, IRAP, CEA, APC Allocations: Mass <90 kg, Power <90 W Detection plane: - DetArea: 1024 cm² - Detectors: 6400 CdTe pixels (4x4x1 mm3) - Energy range: 4-150 keV (images<120) Mask: Ta, 40% open, self supporting Shield: PbCuAl - EffArea ~ 400 cm² @ 10-70 keV - EffArea > 200 cm² @ 6 keV - FoV = 2 sr (total) - Localisation accuracy <12' (at detection limit, 90% C.L.) ECLAIRs trigger & control unit (UGTS) [6] – SVOM workshop, April 2017, Qiannan, China – S. Schanne

ECLAIRs trigger & control unit (UGTS) CNES, CEA Functions Control of instrument (conf, HK, thermal, noisy pix) + Power supply Data acquisition: detected counts  packets  mass-memory Scientific Processing: the Trigger (detection and localization of new source, based on coded mask image deconvolution) Alert messages to ground (via VHF system) Slew requests to satellite Hardware : rad tolerant, itar-free I/O board (x2) + Power supply (x2+4) CPU board (with FPGA and CPU) (x2) FPGA : data acquisition and pre-processing CPU: Leon3 dual-core, 80 MHz, 2×50 Mflops (upgrade from previous Leon2 version) OS : hypervisor for time partitioning UGTS [7] – SVOM workshop, April 2017, Qiannan, China – S. Schanne Image Reconstruction on UGTS CPU: ~ 2 s UGTS CPU

Imaging with ECLAIRs (example) Source position at sky pixel (50,50) (35° off axis) Principle of imaging: Projection of a portion of the mask onto the detectior permits to reconstruct the source sky position Mask « ACS » : Aleksandra Gros, Cyril Lachaud, S. Sch. [8] – SVOM workshop, April 2017, Qiannan, China – S. Schanne

Imaging with ECLAIRs (example) Source position at sky pixel (50,50) (35° off axis) Exposure: 100 s Energy: 4-120 keV Source fluence 100 ph/cm2 (~ 0.5 Crab) Source counts: ~ 10 000 on det CXB counts: ~ 250 000 on det Source reconstructed at sky pixel 50,50 with 30 sigma in image Shadowgram (detector plane image) [9] – SVOM workshop, April 2017, Qiannan, China – S. Schanne

Imaging with ECLAIRs (example) Source position at sky pixel (50,50) (35° off axis) Exposure: 100 s Energy: 4-120 keV Source fluence 100 ph/cm2 (~ 0.5 Crab) Source counts: ~ 10 000 on det CXB counts: ~ 250 000 on det Source reconstructed at sky pixel 50,50 with 30 sigma in image Shadowgram (detector plane image) [10] – SVOM workshop, April 2017, Qiannan, China – S. Schanne

GRB diversity : impact on Trigger GRBs appear randomly on the sky => Search in full FoV low intensity GRB are more numerous => Keep SNRimage Threshold low Spectral diversity: high-z GRBs, XRF, XRR + standard GRBs => 4 Energy strips GRB durations : short GRB ~0.5s, long GRB ~30 s GRB variability : pick-up strong spikes over bkg => Trigger time scales from 10 ms to 20 min 17 powers of 2 4 keV 20 keV 50 keV 120 keV Estrip 0 Estrip 1 Estrip 2 Estrip 3 counts s-1 time (s) -10 2 104 4 104 150 [11] – SVOM workshop, April 2017, Qiannan, China – S. Schanne But image computation on UGTS hadware ~2 seconds => Two separate trigger algorithms: - “Image Trigger” for long time-scales > 20s: systematic sky imaging - “Count-Rate Trigger” for shorter time-scales < 20 s: select first an appropriate time scale, then do the imaging on it. Long GRBs (~33 s) Short GRBs (~850 ms) 1ms 10ms 100ms 1s 10s 100s 1000s T90[s] N BATSE (4b cat) 11

Image Trigger of ECLAIRs Cyclic process every 20.48 s, on 4 energy strips - compute Shadowgramm - fit & subtract background shape (CXB modulated by Earth) - deconvolution  Sky Image (counts & variance) Earth Raw shadowgram Reconstructed sky time [12] – SVOM workshop, April 2017, Qiannan, China – S. Schanne - history of sky images (by image summation) 7 time scales: 20.48 s à 1310.72 s (20 min) - search for maxima in reconstructed sky on portions not obscured by Earth and pixels not in known source catalogue - if SNRimage > Thresh, localization by fitting peak in image, and GRB alert

Count-Rate Trigger of ECLAIRs Cyclic process every 2.5 s First step detect excesses in count-rate 9 detector plane zones (sensitive to out-of-axis GRBs) on 4 energy strips, time-scales 10 ms to 20.48 s, - compute background model B - compute - significant excesses (SNRcnt > ThreshCnt)  stored into a buffer [13] – SVOM workshop, April 2017, Qiannan, China – S. Schanne Second step imaging of best excess - search best excess present in buffer, not yet processed - construct Shadowgram on this energy strip and time-scale - deconvolution Sky Image, search excess away from known sources - if SNRimage > Thresh, generate GRB alert Loop to first step: new SNRcnt not yet processed  refined alert

[1 ex. studied with Sarah Antier & Fei Xie] Trigger of ECLAIRs using GRM info GRB050709 (HETE-2) Foreseen in ECLAIRs trigger algorithm : If GRM detects (in 3 detectors) a spike:  Search for short GRB seen in ECLAIRs image, but not in rate-excess: ECLAIRs Count Trigger (scales< 20s)  insert GRM detection in excess buffer  build ECLAIRs image on time window of GRM trigger.  lower SNRimage Threshold to search for weak source  Search GRB with extended emission: ECLAIRs Image Trigger (scales>20s)  lower SNRimage Threshold during the next ~300 s (predefined) to catch extended emission [1 ex. studied with Sarah Antier & Fei Xie] 70 ms 130 s Chandra [14] – SVOM workshop, April 2017, Qiannan, China – S. Schanne GRB050724 (Swift) 250 ms 120 s NS-BH merger, possibly GW event

Trigger output: Alert messages to VHF If SNRi > AlertThresh (~6.5 s)  Alert msg sequence  GFT If SNRi > SlewThresh (~10 s)  request satellite Slew  MXT & VT Sub threshold excesses  included in VHF recurrent msg  GWAC VHF recurrent message Detection + Localization 1st alert (>AlertThresh) confirmed alert (>1st alert) confirmed alert (>SlewThreshold) Light curves Alert descriptor MXT packets T0 Tb0 Tb1 Tb2 Tsr Ts Slew delay [15] – SVOM workshop, April 2017, Qiannan, China – S. Schanne

Trigger prototype implementation Time (s) SNRimage (s) Cnt/25 ms Earth entry In FoV coded as C++ libraries Scientific Software Model (SSM) : both triggers: standalone programs on Linux  simulate thousands of GRBs check scientific performance Data Processing Model (DPM) : FPGA+Leon2 CPU hardware boards same code compiled on Rtems OS,  check CPU power margin Sim individual photons CXB + GRBs (time, energy) through CxgSim (detector simulation)  time, energy, pixel  input to Triggers Count-Rate Trigger [16] – SVOM workshop, April 2017, Qiannan, China – S. Schanne Image Trigger SNRimg cnt/20 ms Time (ks)

Dynamic trigger sim of expected GRBs Sarah Antier (CEA, thesis 2013-2016) Database of ~2000 GRBs detected by Batse, Swift/BAT, GBM and Hete-II  GRBs with light curves, extrapolated to ECLAIRs band (4-120 keV) Background: 100 periods of 1500 s of CXB with Earth, chosen in 1 yr B1 law For each GRB, chose Bkg period and 70 positions in FoV outside Earth in 1yr B1 law  though CxgSim (counts)  through both Triggers (SSM) Swift/BAT [17] – SVOM workshop, April 2017, Qiannan, China – S. Schanne N / bin Swift/BAT 6.5 sigma Alert Thresh. 10 sigma Slew Thresh. SNRimage Thresh (σ) Fluence in 4-120 keV (ph/cm2)

Expected ECLAIRs GRB rates (GRB/yr) (low fluence LGRBs Swift) GRBs expected triggered by ECLAIRs Sarah Antier (CEA, thesis 2013-2016) using detection efficiency, BATSE GRB/yr norm, ECL FoV = 1.83 sr, Duty cycle = 54% (Earth, SAA) ɛGRB > 20% 50% 70% 90% Expected ECLAIRs GRB rates (GRB/yr) > AlertThr (6.5 ) > SlewThr (10 ) BATSE S+L GRBs 46 – 57 ±8 40 – 49 ±8 bonus low E (XRR GBM, XRF Hete) 4 – 10 ±1 4 – 9 bonus ImageTrig (low fluence LGRBs Swift) 4 – 5 4 - 5 Total 54 – 72 ±10 47 – 63 ±7 [18] – SVOM workshop, April 2017, Qiannan, China – S. Schanne T90 (s) T90 (s)

Known source catalog Nicolas Dagoneau (CEA, ongoing work, 2017) Swift/BAT 14-195 keV monthly in 6 yr (2004-2010) RXTE/ASM 5-12 keV weakly in 16 yr (1995-2012) Sources visible by ECLAIRs in 20 min [19] – SVOM workshop, April 2017, Qiannan, China – S. Schanne Visibility (% of time) Most of sources excluded by B1 pointing law (|b|<10° & Sco X-1 ouside FoV), Not always in B1 pointing: GRB follow-up, ToO, GP > 1/3 of time  Sources rising over Earth horizon  Work to include management of sources in the trigger algorithms (shadowgram differences in Count Trigger, Source modeling in ImageTrigger )

谢 谢 for your attention Outlook Trigger optimization ongoing - refine background simulation (CXB: add Sources, SAA & Earth Albedo) - update background modeling onboard (rate and imaging) - noisy pixels management (implement and test) - optimize trigger parameters (incl. Ebands, thresholds)  maximize GRB detection vs false alert rate CPUboard EM Ph injector Spacewire interface IO interface Development of UGTS in parallel in our lab CPUboard EM: FPGA firmware & CPU software May 2017: UGTS BBM to be received implementation on the “real” hardware, Start FM trigger coding April 2018: UGTS EQM to be tested in China no trigger included yet (no camera), test UGTS-PDPU communication, TC/TM packets, simulated DAQ, VHF sequence [20] – SVOM workshop, April 2017, Qiannan, China – S. Schanne 谢 谢 for your attention

Gamma-Ray Burst science questions GRB physics Acceleration and nature of the relativistic jet Radiation processes, gamma-ray emission The early afterglow and the reverse shock Long GRB-supernova connection Short GRB-merger connection GRB progenitors Fundamental Short GRBs and gravitational waves Origin of high-energy cosmic rays Lorentz invariance test physics Cosmology Star formation tracer Re-ionization of the universe Cosmological parameters Cosmological lighthouses (absorption systems) Host galaxies GRB phenomenon Diversity and unity of GRBs, central engine [22] – SVOM workshop, April 2017, Qiannan, China – S. Schanne Need of a complete sample of GRBs, with spectral and temporal coverage and a distance measurement

SVOM instruments Space 4 space instruments: 3 ground telescopes ECLAIRs gamma-ray imager & trigger GRM gamma-ray monitor MXT X-ray focusing telescope VT visible band telescope 3 ground telescopes GWAC ground wide angle camera F-GFT & C-GFT: ground follow-up telescopes Space [23] – SVOM workshop, April 2017, Qiannan, China – S. Schanne G r o u n d

Temporal & spectral GRB coverage S p a c e [24] – SVOM workshop, April 2017, Qiannan, China – S. Schanne G r o u n d

Other SC mission center GRB coordination distribute GRB observation scenario GCN Other telescopes T0+30s T0+1min Auto slew T0 GRB trigger T0+5min VT,MXT observation Other SC mission center GRB coordination distribute Inside SVOM Outside SVOM On board On Ground VHF network F SC GRB ECLAIRs trigger C SC [25] – SVOM workshop, April 2017, Qiannan, China – S. Schanne

The 3 scientific programs of SVOM Core Program (CP) : follow-up GRB triggers of ECLAIRs General Program (GP) : AGN, ULX, TDE, Galactic sources (CV, XRB, pulsars, magnetars, TGF), background studies (CXB), etc Targets of Opportunity (ToO), 1 / day: follow-up external triggers: multi-wavelength (SKA, LSST, CTA, HAWC) or multi-messenger (GW, neutrino) [26] – SVOM workshop, April 2017, Qiannan, China – S. Schanne Observing time allocation Extended mission (after 3 first years) ToO rate increased (up to 5/day) GP time outside the “B1 law” increased (10%  50%)

SVOM pointing strategy For optimal ground follow-up to determine redshift : favor sky observable from Hawaii, Chile and Canary Islands satellite attitude roughly antisolar towards the night To maintain satellite radiators cold : satellite attitude antisolar within 45° For best GRB detection performance : keep Sco X-1 and Galactic Plane outside the ECLAIRs FoV [27] – SVOM workshop, April 2017, Qiannan, China – S. Schanne ECLAIRs sky exposure (4.5 Ms towards Galactic poles in 1 yr) MXT and VT pointings (in 1 yr nominal mission)

Conclusions SVOM designed to study the diversity of GRBs and get a complete sample, good spectral and temporal coverage of the prompt and afterglow, optimized follow-up strategy to get redshift of a large GRB fraction (~50%) SVOM observation plan and instruments (space+ground-based) suited to detect high redshift GRBs (20 GRBs with measured z>5 in 5 yr) SVOM is prepared for multi-messenger era: ToO follow-up of GW and neutrino alerts [28] – SVOM workshop, April 2017, Qiannan, China – S. Schanne More information: “The Deep and Transient Universe in the SVOM Era: new Challenges and Opportunities, Scientific prospects of the SVOM mission”, J. Wei, B. Cordier et al., arXiv:1610.06892