The Pinghu Formation in the Pingbei slope belt of Xihu sag is a tidal-affected delta deposit characterized by thin interbedded sediments of sand, mud and coal. The coal layer is well developed and thin (< 1 m). The widespread development of thin coal seam leads to the complex seismic reflection characteristics of sand, and the sand prediction in thin coal seam is more difficult. This paper firstly analyzes and summarizes the characteristics of thin coal seam in well data of the research area. According to the characteristics of lithologic combination in wells, geological models are created to carry out forward modeling, and the forward modeling results are compared and analyzed. The frequency attributes are used to predict the distribution of sand. Then, create a new method for petrophysical modeling of thin coal seams, fit correction curves, and conduct prestack inversion to describe the sand superposition relationship. The results indicate that amplitude response of sand in thin coal strata is greatly affected by thin coal strata, while the frequency attribute is less affected. The frequency attribute can be used to effectively identify the sand in thin coal strata. The inversion data with the participation of the new petrophysical model can effectively predict the sand and eliminate the influence of some thin coal strata. The sand prediction of this method are in line with the sand transport and sand pooling model in this area. At the places where the fault strike changes and the fault activity is weak, the channel sand body branches or changes direction, which is an effective sand transport channel, forming a medium and low belt sand body scale unloading area, which is consistent with drilling. The research shows that seismic forward modelling analysis can provide fast and effective guidance for seismic data interpretation of sand and thin-coal, and proposes a new petrophysical method to eliminate the influence of thin coal strata and realize high-precision sand prediction, especially for the complex reservoirs. The research results provide strong practical significance for thin coal measures formation reservoir prediction ideas.