Title:Interpreting the $γ$-ray emission from Westerlund 1: hadronic or leptonic?

Abstract:High-energy $\gamma$-ray emission has been detected from several star clusters, indicating that these sources are sites of cosmic-ray acceleration. However, it is still unclear whether the $\gamma$-ray emission is of hadronic or leptonic origin. We present a comprehensive study of the $\gamma$-ray emission from Westerlund 1, under the assumption that high-energy particles are accelerated at the cluster WTS. We investigate both the hadronic and leptonic scenarios by fitting simultaneously the spectrum and morphology of the $\gamma$-ray emission. We numerically solve the transport equations for non-thermal protons and electrons for a given $K_\rm{ep}$ and derive their energy and spatial distributions. $\gamma$ rays are produced either by hadronic pp interactions of protons with the target gas or by inverse Compton scattering of relativistic electrons (both primary and secondary) with the ambient photon fields. We also calculate the synchrotron emission from these electrons. The predicted SEDs are compared with the $\gamma$-ray measurements by Fermi-LAT and HESS and with the X-ray upper limit derived by eROSITA. For reasonable model parameters, we find that in the case of Bohm diffusion, both the leptonic and hadronic scenarios can account for the observed multi-wavelength spectrum and $\gamma$-ray morphology. However, in the case of Kraichnan diffusion, better motivated in star clusters, only the hadronic scenario can explain the observations, while the leptonic scenario fails to reproduce the highest-energy $\gamma$-ray data points. For the leptonic scenario, the magnetic field immediately downstream of the shock can be constrained to be $\lesssim$ a few $\mu$G thanks to the eROSITA upper limit. On the other hand, in order to explain the $\gamma$-ray observations with hadronic emission alone, the gas density inside the bubble should be $\gtrsim 1.2 \, \rm{cm}^{-3}$.