Abstract: The spatiotemporal varying of environmental parameters, especially the varying of the sound speed profile caused by internal tide, can lead to a mismatch which is a critical challenge for matched field processing of passive localization of the sound source in shallow water. According to the characteristics of the spatial and temporal variations of the sound speed profile caused by internal tide in shallow water, a passive localization method based on matched field processing is proposed. The range-dependent varying of the sound speed profile are approximately represented by multiple adjustable linear varying profiles, and these equivalent profiles are reconstructed by the empirical orthogonal functions which contain the time-varying properties. The results of both the simulated and experimental data show that this method can mitigate the effect of the model mismatch of the sound speed profile on source localization and improve localization performance of match field processing in shallow water environment with internal wave. Comparing with the traditional matched field processing methods which ignore the range dependence of the sound speed profile, it has better performance of source localization and smaller estimation error of source range.