Anomalies in the gamma-ray diffuse emission of the Galaxy and implications for the interpretation of IceCube results

Several independent analyzes of Fermi-LAT results found evidences of an excess of γ-ray diffuse emission along the inner Galactic plane and of a related spatial dependence of the cosmic ray (CR) proton spectral index. These features are not accounted for by conventional models of CR transport. We show that a phenomenological model accounting for those results in terms of spatial dependent CR transport also reproduces the γ-ray excess found by Milagro at 15 TeV in the inner Galactic plane and by H.E.S.S. in the Galactic center. We then use that model to compute the neutrino emission along the Galactic plane finding that is significantly larger than expected on the basis of conventional models. This emission is compatible with ANTARES upper limits and may soon be detected by IceCube or, more likely, by Km3NeT.

A large number of parameters to be fixed against data ! Ginzburg & Syrovatsky, 1964 The CR trasport equation The CR Galactic population Commonly, propagation parameters are fixed on the basis of local observables! e.g. the diffusion coefficient ! D(E) = D 0 (E/E 0 ) -δ ! is fixed on the basis of the secondary/primary CR nuclei ratio (the B/C most importantly) and assumed (for conventional models) to be spatially uniform! warning !! ! due to CR vertical escape and nuclear inelastic scattering onto the interstellar gas secondary nuclei probes only few kpc around us.
Propagation may behave differently in the central region of the Galaxy

ApJ 2008
The Milagro anomaly in the inner Galactic Plane • the excess is present also respect to updated conventional models tuned on CR data and all-sky Fermi-LAT data • this holds also accounting for the CR hardening at ~ 250 GeV/n assuming it is a large scale effect.
(the proton and He spectra were assumed to match CREAM data up to 100 TeV/n) conventional model without hardening conventional model with hardening

The Milagro anomaly in the inner Galactic Plane
• the excess is present also respect to updated conventional models tuned on CR data and all-sky Fermi-LAT data • this holds also accounting for the CR hardening at ~ 250 GeV/n assuming it is a large scale effect.
(the proton and He spectra were assumed to match CREAM data up to 100 TeV/n) The KRA model -implemented with the DRAGON code -adopts a radial dependent diffusion coefficient δ(R) = A R + B for R < 11 kpc, const. above such that δ(R sun ) = 0.5 and convective velocity for R < 6.5 kpc The model is tuned to reproduce the proton spectrum measured by PAMELA (including the hardening @ 250 GeV/n) up to 1 TeV, the B/C (antiprotons also matched by secondary prod.) as well as updated diffuse -ray Fermi data dV C dz = 100 km s 1 kpc 1 The KRA model: Radial dependency of CR transport

Gaggero, Urbano, Valli & Ullio arXiV: 1411.7623 PRD 2015
The project started in 2008, more than 20 peer reviewed papers based on this code. The present version use (among other options) the same nuclear cross sections and gas distribution as in GALPROP Main innovative features respect to previous codes:! • spatial dependent diffusion coefficient(s) (both normalization D 0 (R,z) and rigidity dependence index δ(R,z) )! • anisotropic diffusion in 3D! will soon be released.
A first (of a series) of technical papers to appear in a few days.

Gaggero, Urbano, Valli & Ullio arXiV: 1411.7623 PRD 2015
The KRA model reproduces the full-sky Fermi spectrum and angular distribution. It also provides a better fit in the inner GP region  The KRA model nicely matches MILAGRO consistently with Fermi data (point sources cleaned) no further tuning is required ! The KRA model nicely matches MILAGRO consistently with Fermi data (point sources cleaned) no further tuning is required ! Beside inhomogeneous diffusion the CR hardening at ∼ 250 GeV/n must also be accounted for. This result suggests that hardening is not a local feature but it is may also be related to unconventional diffusion ! -CR advect/diffuse in self-generated Alfven-waves below/above ∼ 50 GeV -harder CR (hence -ray) spectrum if advection dominate the effect is larger in the inner Galaxy, larger D ➞ larger p at which diffusion dominate Note however that this is a low energy effect !! This is at odd with Fermi data and Milagro anomaly (HAWC may soon confirm ! ) proton spect. index diffusion coefficient The case for a spatial dependent δ! A possible interpretation ? What all this implies for neutrinos ?
IceCube found evidence for 28 (2 years) then 37 HESE events (3 yrs) with reconstructed direction above 30 TeV corresponding to a 5.7σ excess respect to the atm. bkg. " angular distribution compatible with isotropy (see however below)" composition compatible with a equal mixture of e, μ, τ as expected for astrophysical generated neutrino" Best fit spectral index above 60 TeV" -2.3 ± 0.3  Even in the quasi-linear theory D || and D ⊥ have opposite dependence on the turbulent power. This is confirmed by ray tracing simulations in strong turbulence regime Marco, Blasi & Stanev 2007 (see also Casse et al. 2002) D k (E) / E 1/3 D ? (E) / E 0.5÷0.6

IceCube measured ν events
for Kolmogorov turbulence