This study aims to study the applicability of cloud model forecasting products in artificial rain enhancement operations in complex terrain areas of Guizhou based on the quantitative forecasting products of CMA-CPEFS (China Meteorological Administration-Cloud and Precipitation Explicit Forecasting System) model developed by CMA Weather Modification Centre. The fusion diagnosis method using multi-source observation data of ground-based radar, geostationary satellite and sounding profile is adopted to test the model forecasting performance of 30 cases of aircraft-based rain enhancement in Guizhou from 2020 to 2021. The results are as follows. (1) The model accuracy of distinguishing cloud phase is 93% (28/30), and the matching degrees of the operation height layer and the catalyst type configuration are 93% (28/30) and 100% (30/30), respectively, confirming that the model’s ability to characterize the cloud physical parameters meets the operational standard. (2) There is a systematic positive deviation in the height of 0 ℃ layer, the mean bias error is +108 m, which passes the 95% confidence level test, and the confidence interval is (+108±97) m with p＜0.01. The mean absolute error is +205 m. The forecast accuracy of cloud system’s horizontal moving direction is 80%, slightly higher than that of cloud system’s horizontal moving speed (77%). (3) The threat score (TS) of the potential area for rain enhancement is 0.67, but there is a significant deviation in spatial expansion (the forecasting range is 18%±3% larger than that of the actual situation), which is significantly correlated (R²=0.82, p＜0.001) with the subjective analysis error and overestimation of cloud water content by the model (the deviation is +0.15 g·m^-3). The results provide key physical constraints for the optimization of numerical models for weather modification in the karst landform areas of Guizhou.