Cosmic rays, γ and ν in star-forming galaxies Xiang-Yu Wang Nanjing University Collaborators: X. C. Chang, F. K. Peng, Q. W. Tang

CRs in our Galaxy

Cosmic ray (CR) luminosity in galaxies Energy density 1 eV/cm^3 ， equipartition of the magnetic and CR -- a major energetic component Diffuse Galactic gamma-rays are produced dominantly by CRs CRs represent a major form (but not well-studied) of galactic feedback, regulating the formation and evolution of galaxies

1. Evidence of cosmic rays in other galaxies?

Gamma-ray messenger

Gamma-ray emission from starburst galaxies Abdo et al. 2010

Cosmic rays accelerated by SNRs Cygnus Loop (0.5-10 GeV) Supernova explosions induce shocks (SNRs) Cosmic rays are accelerated across these shock fronts GeV Gamma-rays are produced by Cosmic rays

Are galaxies CR calorimeters? Galaxies are typically calorimeters for lepton CRs, explaining the FIR- radio linear correlation How about protons?

Correlation between gamma-ray and infrared luminosities Several nearby star-forming galaxies detected Gamma-ray and infrared luminosity well correlated Naturally expected if more CR energy is converted into gamma-rays in more luminous galaxies Ackermann et al. 2012

CR calorimeter? “calorimetry fraction limit” Best target: (ultra) luminous infrared galaxies Lacki et al. 2011

GeV emission from LIRG NGC 2146 Tang, Wang XY & Tam 2014 A luminous infrared galaxy – CR calorimeter ? using the 68 month Fermi data 5.5σ detection of gamma-ray emission above 200 MeV

NGC2146—a likely calorimeter ? assuming ESN,51 η0.05=1, for proton calorimeter limit : L0.1-100GeV/L8-100μm=1.5e-4. NGC 2146 is likely a proton calorimeter ! cosmic rays accelerated in NGC 2146 lose most of their energy into secondary pions Tang, Wang XY & Tam 2014 2018/12/5

Arp 220- the nearest ULIRG: must be calorimeter! A prototype of ULIRG: LIR=1.4*1012Lsun D=78Mpc n~104cm-3 Possible AGN SN rate: 4+-2/yr Long predicted to be GeV sources (e.g.,Torres 2004; Lacki+ 2011; Yoast-Hull+2015) tpp<tescape Hubble image of Arp 220 作为对比，银河系的红外光度为1.4*10^10 Lsun，超新星爆发率为0.03/yr，恒星形成率为1 Msun/yr 天文年会报告

Fermi observation- PASS 8 Peng, Wang XY et al. 2016, ApJL

LC and SED of Arp 220 Favor cosmic Rays origin

Efficiency of powering CRs Cosmic Rays injection power GeV emission luminosity from CRs Efficiency of powering CRs of SRNs

Is there direct evidence for cosmic ray origin?

Pion-decay evidence in supernova remnants Science, 2013 p+p->p+p+ p0 p0->γ+γ

Gamma-ray count map around LMC in 60 MeV-2.45 GeV Tang, Peng, Liu, Tam, Wang XY, 2017 LMC is the brightest external galaxies in gamma-ray emission LMC is near enough that individual star-forming regions can be resolved The high Galactic latitude of LMC also leads to a low level of contamination due to the Galactic diffuse gamma-ray emission.

Spectrum of the LMC disk component Tang, Peng, Liu, Tam, Wang XY, 2017

fit with pion-decay model Tang, Peng, Liu, Tam, Wang XY, 2017

3. Neutrinos from CRs in galaxies ?

Arriving directions isotropic distribution dominated Favor extragalactic origin

Starburst galaxy scenario Loeb & Waxman 2006 Cosmic rays are accelerated by SNR shocks Normalized with the local 1.4 GHz energy production rate But, Normal SNRs can only accelerate CR to PeV, while IceCube neutrinos need 100 PeV CRs ? Eν ~ 0.04 Ep: PeV neutrino ⇔ 20-30(1+z) PeV CR proton

Hypernova remnant scenario (Liu, Wang et al. 2014) Normal SNRs can hardly accelerate CR to PeV Hypernovae can accelerate CR protons to 10^18 eV (Wang et al. 2007) Semi-relativistic (v~c, or Γβ≥1) outflow Proton ISM Hypernova remnant

PeV Neutrino production (Liu, Wang XY et al. 2014) CR protons collide with the surrounding gas and produce neutrinos Proton ISM Two escape ways: 1) diffusion 2) advection pp efficiency in star-forming galaxies & starburst galaxies See also He et al. 2013

Detailed calculation Chang, Liu & Wang 2015 Use infrared luminosity function obtained by Herschel PEP/HerMES (Gruppioni et al. 2013) Sum up contributions by different galaxy populations Star-forming galaxies also contribute significantly to the diffuse gamma-ray background

Normal SNR + HNR (Chakraborty and Izaguirre 15; Senno+ 15) With normal SNRs, can explain the low energy IceCube data (softer spectrum) IceCube (Chakraborty and Izaguirre 15)

Tension with the gamma-ray background Scenarios： 1）overestimate blazar contribution to EGB ? 2）hidden sources in gamma- rays (e.g. choked jets) arXiv:1511.00688

Internal Υ-ray absorption in starbursts Chang & Wang XY 2014

Taking into account the synchrotron loss of secondary pairs Chang & Wang XY 2014

High-redshift sources Assuming sources are at z=4-10 Considering EBL absorption Xiao et al. 2016

Fermi Search for GeV Flares Associated with the IceCube Track-like Neutrinos Peng & Wang XY 2017 Fermi/LAT upper limits:

Put constraints on the source density Favor low-luminosity, high-density candidates high-density source case (e.g. 4x10-4Mpc-3; starburst galaxies) low-density case (e.g. 4x10-9Mpc-3 FSRQ) Peng & Wang XY 2017

Summary GeV-emitting star-forming/starburst galaxies are huge reserviors of CRs Starburst/ULIRGs (e.g. Arp 220) are likely calorimeters of CRs Star-forming/starburst galaxies may also be, at least partly, the sources of HE neutrinos observed by IceCube