Modeling biases from constant stellar mass-to-light ratio assumption in galaxy dynamics and strong lensing

A constant stellar-mass to light ratio $M_{\star}/L$ has been widely-used in studies of galaxy dynamics and strong lensing, which aim at disentangling the mass density distributions of dark matter and baryons. In this work, we take early-type galaxies from the cosmological hydrodynamic IllustrisTNG-100 simulation to investigate possible systematic bias in the inferences due to a constant $M_{\star}/L$ assumption. To do so, we construct two-component matter density models, where one component describes the dark matter distribution, the other one for the stellar mass, which is made to follow the light profile by assuming a constant factor of $M_{\star}/L$. Specifically, we adopt multiple commonly used dark matter models and light distributions. We fit the two-component models directly to the {\it total} matter density distributions of simulated galaxies to eliminate systematics from other modelling procedures. We find that galaxies in general have more centrally-concentrated stellar mass profile than their light distribution. This is more significant among more massive galaxies, for which the $M_{\star}/L$ profile rises up markedly towards the centre and may often exhibit a dented feature due to on-going star formation at about one effective radius, encompassing a quenched bulge region. As a consequence, a constant $M_{\star}/L$ causes a model degeneracy to be artificially broken under specific model assumptions, resulting in strong and model-dependent biases on estimated properties, such as the central dark matter fraction and the initial mass function. Either a steeper dark matter profile with an over-predicted density fraction, or an over-predicted stellar mass normalization ($M_{\star}/L$) is often obtained through model fitting. The exact biased behaviour depends on the slope difference between mass and light, as well as on the adopted models for dark matter and light.