Data assimilation provides a way to improve deterministic model accuracy by combining model predictions with observations. The increasing availability of remotely sensed land surface variables has led to the increased use of data assimilation methods for hydrologic applications. The data assimilation methods, however, are primarily designed to correct random errors in observations and model predictions. Therefore, the presence of bias errors must be separately addressed for the success of a data assimilation system. Several strategies for addressing biases have been employed in prior studies, including (1) offline bias correction, where the observations are scaled to the model's climatology prior to data assimilation, (2) online-bias correction, where biases are estimated dynamically within the data assimilation system and (3) model calibration, where the model parameters are calibrated to reduce the inherent biases in model predictions. In this study, we compare the impact of these different bias mitigation strategies in the assimilation of surface skin temperature. The study is conducted using the Land Information System (LIS) data assimilation testbed, which is an interoperable framework for sequential data assimilation that can employ multiple land surface models, multiple observations and multiple data assimilation algorithms. A suite of Observing System Simulation Experiments (OSSEs) is conducted using the Noah and Catchment land surface models and using the Ensemble Kalman Filter (EnKF) algorithm for the assimilation of skin temperature. We will present a quantified analysis of the impact of these three bias correction strategies for skin temperature assimilation.