Gamma-ray satellites Fermi Gamma-ray Space Telescope

Transcription

1 Space Instrumenta.on (ELEC-E4220) Gamma-ray satellites Fermi Gamma-ray Space Telescope Talvikki HovaBa University of Turku

2 Outline Introduc.on Gamma-ray detec.on techniques Photoelectric effect GBM instrument Compton scabering Pair produc.on Fermi LAT instrument Tracker Calorimeter Event analysis From trigger to event reconstruc.on Fermi Science Data analysis demonstra.on

3 Introduc.on

4 Electromagne.c spectrum

5 Gamma rays E > 100 kev, f > Hz, λ < m In astronomy is high-energy non-thermal emission Emission does not depend on the physical temperature of the object Not necessarily from radioac.ve decay Requires extreme par.cle accelera.on Typical emission processes are inverse Compton emission or pion decay

6 Gamma-ray sky Fermi 5-year sky

7 Gamma-ray sources in our Galaxy Diffuse emission from the plane of the Galaxy From interac.on of high energy cosmic rays with interstellar maber Supernova remnants Pulsars Fermi bubbles Remnant jet emission?

8 Extragalac.c Gamma-Ray sources Diffuse emission from unresolved gamma-ray sources Ac.ve galac.c nuclei Starburst galaxies Gamma-ray bursts

9 Credit: NASA Variable gamma-ray sky

10 Fermi gamma-ray space telescope Launched on June 11, 2008 Data available since August 2008 Low Earth orbit at ~565km Orbit period min Full sky coverage in 2 orbits! Planned life.me 5 years S.ll on-going No consumables! 2 instruments: GBM 8 kev 40 MeV (GRB detec.on) LAT 30 MeV 300 GeV (imaging instrument)

11 Launch mass 4303 kg Downlink speed 40Mb/s (1Mb/s available for LAT) FERMI

12 Gamma-ray detec.on techniques

13 Gamma-ray detec.on Book: Handbook of Par.cle Detec.on and Imaging edited by Claus Grupen, Irène Buvat. Scin.llators

14 Photoelectric effect A photon absorbs into material and gives all its energy to electrons in the absorbing material Einstein got a Nobel prize in 1921 for his discovery of the law of the photoelectric effect" Scin.lla.on crystal Gamma-ray absorbed, light emibed Photomul.pliers detect light In gamma-ray instruments a scin.lla.on crystal is used (e.g. NaI or CsI) Instead of electrons a light flash is emibed Photomul.plier tubes detect the light and convert it to an electrical signal Used e.g. in GBM on Fermi and many other X-ray satellites

15 Scin.lla.on Scin.llator = material that absorbs the energy of an incoming par.cle Scin.llates = re-emits the absorbed energy in form of light (low-energy photons) Photoelectric effect By Qwerty123uiop, CC BY-SA 3.0, hbps://commons.wikimedia.org/w/index.php?curid=

16 Gamma-ray Burst Monitor (GBM) 8 kev 1 MeV 150 kev 40 MeV Nearly full-sky view with detectors placed on opposite sides of the satellite Provides a trigger when the photon counts in at least 2 detectors show a significant increase Time resolu.on 2 μs Needed for detec.on of fast GRBs that last < 1s

17 Compton scabering A photon collides with an electron and gives some of its energy to it ScaBering angle depends on the polariza.on of the incoming photon -> could be used to detect polariza.on in gamma-ray bands! Scin.llator In gamma-ray instruments the gamma-ray first interacts with a scin.llator The scabered gamma-ray then gets absorbed in another scin.llator that is viewed by phototubes Used e.g. in the COMPTEL instrument on board CGRO

18 Pair produc.on E > 2 mc 2 e.g., tungsten or lead A photon goes near the nucleus of an atom and gives its energy to an electron-positron pair The energy of the photon must be at least two.mes the rest mass of an electron Dominant mechanism for photons with > 30 MeV energy In gamma-ray instruments there is usually a series of converter foils and tracking material to track the path of the pair The energy of the photon is calculated from the pair in a Calorimeter Used e.g. in LAT and in its predecessor EGRET

19 Fermi Large Area Telescope (LAT)

20 Large Area Telescope (LAT) Tracker tower 1.8m Micrometeor shield 0.72 m An.coincidence shield Calorimeter

21 Design considera.ons Size of the telescope = size of the detector Field of View of LAT is 2.4 sr (20% of the sky) Need a large mass or radia.on length to absorb the gamma rays Mass of LAT is ~2800 kg, Calorimeter is 1800kg An.coincidence shield used to discard cosmic rays For each gamma-ray photon there is at least Cosmic rays entering the telescope Made of plas.c scin.llator.les that are sensi.ve to charged par.cles (protons, electrons) but not neutral photons The trigger system will discard most of the Cosmic ray events before sending the data down

22 LAT tracker Gamma-ray photon interacts with tungsten foil where an electronpositron pair is formed Silicon strip detectors detect the charged par.cles and their path can be followed No consumables! Different to EGRET, which used a spark chamber as a tracker Conversion efficiency (how many gamma-rays pair produce) is ~ 63% above 1 GeV LAT tracker has 16 planes of tungsten The top 12 foils are thinner (3% radia.on length = 0.095mm) than the bobom 4 foils (18% radia.on length = 0.72mm) to op.mize detec.on and angular resolu.on at different energy ranges

23 Cosmic ray spark chamber hbps://

24 Why silicon strip over spark chamber? Only consumable is electricity instead of gas that eventually runs out Efficient par.cle detec.on in a thin layer Reliable once installed Power consump.on went down in last years Price of SSDs went down Each tower has 576 SSDs Mul.-chip-modules (MCM) read them out Atwood et al hbps://doi.org/ /j.astropartphys

25 LAT Calorimeter 4x4 array in the bobom of the tracker The incoming electron-positron pair gets fully absorbed in the calorimeter Each calorimeter unit consists of CsI(Tl) crystal detectors in a 12 x 8 hodoscopic array The electrons / positrons interact with the crystal detector producing a light flash in gamma-ray energies (photoelectric effect) propor.onal to the energy deposited PIN photodiodes read the scin.lla.ng light Each end of the crystal have two photodiodes sensi.ve to different energy ranges (2 MeV-1.6 GeV and 100 MeV-70 GeV) No need for a photomul.plier because the light yield of CsI crystal is larger than e.g. NaI, which requires PMTs

26 Localiza.on within each crystal Loca.on of the event is provided by the loca.on of the crystal in the calorimeter and the light asymmetry measured at the diodes in the opposite ends of the crystal Energy of the event is reconstructed from the amount of light measured at the diodes Atwood et al. 2009

27 LAT event reconstruc.on

28 LAT simulator Test how gamma-ray photons of different energy interact in the tracker / calorimeter hbp://fermi.sonoma.edu/mul.media/latsim/simulator.php

29 Event analysis

30 Hardware is just a small part of the system! Astrophysics Par.cle physics Compu.ng and simula.ons Hardware design

31 LAT on-board data acquisi.on system Collects data from other subsystems Implements mul.level event trigger Provides on-board processing to reduce downlinked events (1 photon / CR) Also to minimize dead-.me from reading out the LAT events Performs on-board science analysis to provide triggers for rapid transients (e.g. gamma-ray burst alerts)

32 Event analysis logical structure

33 Event classifica.on Trigger rates

34 LAT trigger and event reconstruc.on VETO is issued if any of the ACD.les react above a certain threshold = main informa.on used in on-board filtering, combined with informa.on from CAL and TRK Event reconstruc.on done using pabern recogni.on algorithms CAL is triggered If any of the crystals show an event with energy above the threshold of the PIN diode TRK is triggered when at least 3 consecu.ve layers show an event

35 LAT event analysis 1. Selects the best candidate track 2. Calculates the most likely energy for the event 3. Chooses the best direc.on for the event 4. Performs charged par.cle rejec.on 5. Flags likely cosmic rays (CRs) based on tracker signal 6. Flags CRs based on calorimeter signal 7. Es.mates the gamma-ray probability and classifies the event (e.g. transient or source)

36 Instrument response func.ons (IRFs) Describe the instrument performance Effec.ve area, point-spread func.on, energy dispersion as a func.on of energy Monte-Carlo simula.ons on how par.cles interact with the detectors Generated using Maximum likelihood analysis Ini.al model based on EGRET (Pass 6) Updated based on in-flight experience (Pass 7 and Pass 8)

37 LAT angular resolu.on Determined by the Pointspread func.on (PSF) PSF = the probability distribu.on for the reconstructed direc.on of incident gamma-rays from a point source Energy dependent 5 o at 100 MeV 0.2 o at > 20 GeV At low energies dominated by mul.ple scaberings in the tracker Compare to e.g. op.cal telescopes with angular resolu.on of ~ 1

38 Fermi science

39 LAT science LAT detected 3033 sources during the first 4 years Released in the 3rd LAT source catalog (3FGL) Catalog is constructed using seeds from catalogs at other wavelengths Image Credit: NASA/DOE/Fermi LAT Collabora.on

40 3FGL sources > 1100 Ac.ve galac.c nuclei 166 Pulsars (most numerous Galac.c source class) Pulsar wind nebulae, Supernova remnants, globular clusters, star-forming region, binary systems, novae, starburst galaxies, normal galaxies 1009 Uniden.fied sources!

41 EGRET detec.on LAT vs. EGRET Radio light curve EGRET did pointed observa.ons EGRET could only determine if there were gamma-ray photons and their energy LAT Gamma-ray light curve Radio light curve LAT has large field of view, all sources looked at every 3 hours LAT can see the varia.ons in the gamma-ray photon flux

42 Dark MaBer 27% of the Universe consists of so called dark maber, which does not interact with electromagne.c radia.on Some of these weakly interac.ng massive par.cles (WIMPs) may produce gamma-rays through annihila.on (opposite of pair produc.on) Fermi sees an excess of emission in the Galac.c center, which is consistent with simple dark maber models Must be confirmed with more direct observa.ons!

43 Electromagne.c counterpart to a gravita.onal wave event: Neutron star merger

44 Fermi data analysis Data are sent down daily and are immediately available for user s to analyze No proprietary period a{er the first year Allows fast follow-up at other wavelengths for transient / varying objects Fermi Science Support Center provides tools and instruc.ons on how to analyze the data Fermi Science Tools Analysis threads > Allows even non-experts to analyze Fermi data

45 Fermi data demonstra.on You can try this on your own using the instruc.ons and tools provided by the COSMAX project: {p://

46 References and further reading C. Grupen, I. Buvat (eds.), Handbook of Par.cle Detec.on and Imaging Ackermann et al. 2012: Descrip.on of LAT and the instrument response func.ons (based on in-flight data) hbp://iopscience.iop.org/ar.cle/ / /203/1/4/pdf Atwood et al. 2009: Descrip.on of the LAT instrument hbp://iopscience.iop.org/ar.cle/ / x/ 697/2/1071/pdf 3rd LAT point source catalog (3FGL): LAT science hbps://arxiv.org/pdf/ v3.pdf Fermi science tools: hbps://fermi.gsfc.nasa.gov/ssc/data/ analysis/