Temperature monitoring is a critical issue for lithium ion batteries. Since only the surface temperature of the battery can be measured, a thermal model is needed to estimate the core temperature, which can be higher and hence more critical. In this paper, an on-line parameter identification scheme is designed for a cylindrical lithium ion battery thermal model, by which the parameters of the thermal model can be identified automatically. An adaptive observer is designed based on the on-line parameterization methodology and the closed loop architecture. A linear battery thermal model is explored first, where the internal resistance is assumed to be constant. The methodology is later extended to address temperature dependent internal resistance with non-uniform forgetting factors. The capability of the methodology to track the long term variation of the internal resistance is beneficial for battery health monitoring.