Multi-Messenger Studies with Gravitational Wave Events Challenges and Opportunities Jochen Greiner Max-Planck Institut für extraterrestrische Physik, Garching

“LIGO: A discovery that shook the world” GW 170817 was breakthrough in GW astrophysics Quality and quantity of EM counterpart observations is staggering https://www.youtube.com/watch?v=-Yt5EmEgz2w

Contemporaneous GW & γ-ray Detection GRB 170817A GW 170817 AT 2017gfo (kilonova) Nomenclature: the event was on 17. August 2017 To be fair: Without GW detection this GRB would be just another among 2000 Fermi/GBM GRBs with nothing than γ-ray data: No kilonova, no afterglow, no distance, no jet structure and geometry details, no wondering about low luminosity  truly multi-messenger event no LIGO auto-detection due to noise glitch manual LIGO check after automatic GBM trigger  after 27 min: the coincidence was recognized New era had begun! Abbott et al. 2017, ApJ …and at least 1000 astronomers knew it at this time

Multi-wavelength follow-up observations automatic multi-directional info exchange via GRB email-network: To+25 s: first Fermi/GBM email To+27 min: LIGO info (NS) 192 emails followed 1/3 of all observational astronomers worldwide were involved optical γ-rays, X-rays Radio Neutrinos Abbott et al. 2017, PRL

Combination of different astrophysics Stellar evolution – NS-NS merger – Neutrino emission - Accretion – Jets – Gamma-ray burst – Nucleosynthesis – Chemical evolution 4 different observational phenomena Gravitational waves (GW) Short Gamma-ray burst (GRB) Kilonova GRB afterglow Plan of talk what did we learn? what else can we learn? prospects Tsujimoto+2014

Short GRBs due to NS-NS merger (Eichler+1989) typical jet opening angles ~5-10o closest known short GRB: 080905 at z=0.1218 (570 Mpc) thus, for any short GRB within the aLIGO horizon we expected (1) 100-1000 LIGO detections of NS-NS mergers without a GRB (2) monster-bright γ-ray emission  Instead, the first LIGO detection of a NS-NS merger comes with a weak GRB! Total surprise! Shows our incomplete/biased knowledge!

Gamma-Ray Burst 170817A Both Gamma-ray instruments developed at MPE Garching Fermi/GBM localization consistent with LIGOs INTEGRAL/ACS’ temporal structure sharper With D~40 Mpc: GRB is underluminous by 10.000x!

GRB jet structure and viewing geometry no afterglow during first 10d: slow rise suggest off-axis geometry superluminal motion of compact radio emission from afterglow together with flux evolution: jet opening angle: 4o off-axis angle of observer: 20o consistent with LIGO limit of <28o for inclination of orbital plane (under assumption of jet being perpendicular to NS-NS orbit) very narrow jet observed off-axis Mooley et al. 2018, Nature Troja et al. 2018

The GRB prompt emission spectrum Constraints on synchrotron emission models Begue, JG et al. 2017, ApJ a. Structured or on-axis top-hat jet Lorentz factor of accelerated e- is γe = κmp/me κ parametrizes uncertainty of acceleration ξ < 1 fraction of accelerated e- α < 1 fraction of E turned into radiation αξ extremely tiny for GRB170817A: impossible ξ«1 small efficiency α«0.01 incompatible with relativ. jet 1047 erg/s 1052 erg/s b. off-axis top-hat jet Γ<20 at θ~10o for Epeak difference incompatible with compactness  Emission mechanism is very unlikely synchrotron

produced by successive n-capture reactions (with intermittant β-decay) Kilonova AT 2017gfo Kasen et al. 2017, Nat 551, 80 produced by successive n-capture reactions (with intermittant β-decay) needs n-rich process  NS Merger early blue component from outflow late red component from very n-rich ejecta

Nucleosynthesis from merging NSs Barnes+2016 Smartt, JG+2017 But no, we have not seen gold which isotopes? heavy (up to A~195) ?? lighter (A<130) ?? (both provide acceptable fits!) Rosswog+2018 light curve and color evolution of KN are consistent with r-process

Radioactivity from Kilonova Relatively slow decay of optical light curve of kilonova: Potentially, gamma-ray line measurements could decide between radioactivity or pulsar/magnetar power,  but likely not in near future (1 Mpc limit vs. GRB distances) (e.g. Dai+2017) SPI ~MeV eV ~MeV ~MeV Hotokezaka+ 2016

What did we learn? Answers to… …expected, old questions: short GRBs are indeed due to merging NS GRB jets are narrow theoreticians have grossly predicted the KN signal correctly (duration; luminosity ~1000x nova) light curve and color evolution of KN are consistent with r-process blue component suggestive of merger into a magnetar; red component is likely combination of tidal material and disk wind …unexpected, new questions/problems: GRBs emit off-axis γ-rays the γ-ray emission mechanism in short GRBs is different than previously thought there are indeed ‘red’ and ‘blue’ components in kilonovae: but they are not mutually exclusive; instead, they come together

Open questions what is the emission mechanism in short GRBs? How does off-axis emission (luminosity, spectrum, variability, polarization) scale with Γ? If off-axis γ-ray emission is typical, there should be many local GRBs! Where are they? Why has Swift not seen a single one at <570 Mpc?? simultaneous blue/red components in kilonovae: How to explain the optical/NIR light curve? Do all KN have this blue/red components? What is the effect of viewing angle? final merger product: NS or black hole? how will BH-NS merger look like? need better atomic data for (light) r-process elements Metzger et al. 2017, Science

Predictions I: Rates of nearby GRBs if GRBs also emit off-axis, then there will be many more of those rate depends on relative luminosity ratio and jet opening angle Burgess, JG et al. 2017 Nominal on-axis emission within 10o, beyond power-law decline up to 90o  at faint flux levels, local off-axis GRBs dominate Janka et al. 2006

NS-NS merger rate LIGO detection suggests rate of 1.5+3.2-1.2 yr-1 (per 100 Mpc3) 3 known channels field binary evolution globular clusters nuclear clusters highest rate is from classical isolated binary evolution: 10-2 yr-1  Either rare event, or unknown binary channel with more frequent NS-NS mergers Belczynski et al. 2018, A&A

Predictions II: GW detectors Advanced LIGO / Virgo Plan Next few years will bring hundreds of of GW triggers GW localizations will improve, but generally 10-50x larger than that of GW 170817 now upgrade 2/2019 O3 Start, incl. Virgo ~2020 full design-sensitivity ~2022 KAGRA (Japan) ~2024 LIGO India ~2032?? LISA

Predictions III: multi-λ instruments Wavelength # src / □o FOV Sensitivity instruments γ-rays ++ - Fermi/GBM, INTEGRAL/ACS X-rays 2 + -- Swift (tiling), MAXI UV 10 Optical/NIR 1000 many IR 50 Radio LOFAR  Largest progress possible: with new, more sensitive γ-ray detector

How would a new gamma-ray detector help? Better localization More accurate time difference between merger (GW) and GRB Better spectroscopy to understand the emission process Possibly measure γ-ray polarization to understand the emission process Measure nuclear lines in nearby SN explosions to understand the basics (unlikely to detect nuclear lines from NS-NS merger even with next generation MeV telescope)

Predictions IV: GW counterparts & science for majority (large off-axis angles): no GRB, no afterglow, possibly kilonova within LIGO volume for few: GRB, KN, but no AG for very few: GRB, AG, KN (as 170817) Likely: Kilonova behavior / statistics (not necessarily r-process details) BH-NS merger Unexpected new phenomenon cosmology: Ho (needs a few dozen GRBs) Possibly: Off-axis radiation (process) of short GRBs / Jet physics Possibly inferences on NS radii and equation of state

BH-BH and BH-NS merger Several BH-BH mergers seen by aLIGO None has shown a convincing EM counterpart (claim for EM of first merger, GW 150914, disproven) this is consistent with many earlier predictions, though 150914 sparked the phantasy of some theoreticians But: we should keep open-minded, and continue searching BH-NS merger: still un-observed by aLIGO/Virgo case for EM counterpart completely open Greiner+2016 Connaughton+2016

Summary New era (multi-messenger & multi-wavelength) more new questions than solved problems (nice!) Urgent need for better γ-ray mission (sensitivity & localization) Atomic data of (light) r-process elements Many expectations: KN, off-axis emission mechanism, BH-NS merger, likely also unexpected surprises Pessimistic view: GW 170817 was unique event; no NS-NS merger(s) in O3 Optimistic view: off-axis emission and 100x more frequent NS-NS binaries are the rule, then handful of new NS-NS mergers in O3 (2019)