2021MNRAS.501.2332B

We detail a method to measure the correspondence between dark matter (DM) models and observations of stellar populations within Local Group dwarf spheroidal galaxies (LG dSphs) that assumes no parametric stellar distribution. Solving the spherical or cylindrical Jeans equations, we calculate the consistency of DM and stellar kinematic models with stellar positions and line-of-sight velocities. Our method can be used to search for signals of standard and exotic DM distributions. Applying our methodology to the Fornax LG dSph and using statistical bootstrapping, we find: (i) that oblate or prolate cored DM haloes match the stellar data, respectively, ≃60 or ≃370 times better than oblate or prolate cusped DM haloes for isotropic and isothermal stellar velocity dispersions, (ii) that cusped spherical DM haloes and cored spherical DM haloes match the Fornax data similarly well for isotropic stellar velocity dispersions, (iii) that the semiminor to semimajor axial ratio of spheroidal DM haloes are more extreme than 80 per cent of those predicted by Lambda cold dark matter with baryon simulations, (iv) that oblate cored or cusped DM haloes are, respectively, ≃5 or ≃30 times better matches to Fornax than prolate cored or cusped DM haloes, and (v) that Fornax shows no evidence of a disc-like structure with more than two per cent of the total DM mass. We further note that the best-fitting cusped haloes universally favour the largest mass and size fit parameters. If these extreme limits are decreased, the cusped halo likelihoods decrease relative to those of cored haloes.