Lattice strains obtained from in operando synchrotron X-ray diffraction measurements during metal additive manufacturing are being increasingly used to deduce temperature and cooling rates occurring during the process. The underlying assumption behind this deduction is that lattice strain evolution occurs solely due to temperature changes caused by the passage of the heat-source. However, this assumption crucially neglects the role of mechanical constraints on the sample and the evolution of internal stresses generated due to microstructure evolution. The ensuing elastic strains can have a significant contribution to lattice strains and failure to decouple them from the contribution of thermal strains may induce significant errors in temperature and cooling rate estimates. In this study, we quantify these errors using a combined experimental and numerical approach. First, lattice strains are obtained from in operando synchrotron X-ray diffraction measurements during additive manufacturing of a 316L stainless steel thin-wall. Then, the contributions of elastic and thermal strains to the lattice strain are deconvoluted using experimentally validated fast large-scale thermomechanical simulations of the entire AM process. Results reveal that even in the simple case of AM of a single-phase material such as 316L and a thin-wall geometry, significant errors in both temperature amplitudes (up to 30\%) and cooling rates (up to 27\%) occur when the contribution of elastic strains is neglected. Furthermore, mechanical and thermal effects cannot be trivially decoupled from X-ray diffraction data; fast large-scale thermomechanical simulations become necessary to perform this decoupling.