Abstract:The calculation schemes of the postprocessing system of the GRAEPS model are studied in detail, and the sea level pressure calculation scheme and the temperature vertical interpolation scheme in the GRAPES postprocessing system, and the extrapolation scheme in the vertical interpolation of potential height are improved. GRAPES mostly adopts the cubic spline function method to vertical interpolation, and the interpolation precision can be guaranteed. However, the sea level pressure calculation scheme and the extrapolation scheme below the terrain in the previous GRAPES postprocessing system is not detailed enough. In this paper, assuming that the temperature profile varied with terrain height and the model atmosphere satisfied the static equilibrium condition, the sea level pressure calculation scheme and the original interpolation layer of sea level pressure and temperature used for vertical interpolation is optimized, so that the accuracy of sea level pressure and highlevel temperature is further improved. The statistical test scores are improved in different regions, and the sea level pressure and the highlevel temperature field in the tropics and East Asia improved significantly and passed the significance test. The extrapolation schemes of potential height are optimized, and the lowlevel potential heights are improved significantly in large terrains such as the QinghaiTibet Plateau and the western South America. Statistical results show that this scheme has significant improvements in the anomaly correlation coefficient and root mean square error in the lowlevel fields in various regions of the world. These provide more reliable numerical model products for users.