An alternative solution to the -ray Gradient problem

The Fermi-LAT collaboration recently confirmed EGRET finding of a discrepancy between the observed longitudinal profile of -ray diffuse emission from the Galaxy and that computed with GALPROP assuming that cosmic rays are produced by Galactic supernova remnants. The accurate Fermi-LAT measurements make this anomaly hardly explainable in terms of conventional diffusion schemes. Here we use DRAGON numerical diffusion code to implement a physically motivated scenario in which the diffusion coefficient is spatially correlated to the source density. We show that under those conditions we are able to reproduce the observed flat emissivity profile in the outer Galaxy with no need to change the source term, the diffusion halo height, or the CO-H2 conversion factor (XCO) with respect to their preferred values/distributions. We also show that our models are compatible with gamma-ray longitudinal profiles measured by Fermi-LAT, and still provide a satisfactory fit of all observed secondary-to-primary ratios, such as B/C and antiprotons/protons.