Abstract: Previous studies on the parameterization of microphysical process schemes based on the WRF model for the Yangtze River Basin have not optimized the cumulus convection parameterization schemes for multiple meteorological elements in the middle and lower reaches of the Yangtze River in this specific region. In this study, on the basis of the optimal microphysical process and boundary layer parameterization schemes, the middle and lower Yangtze River Basin is selected for the first time as the research object, and three cumulus convection parameterization schemes, Kain-Fritsch (KF), Betts-Miller-Janjic (BMJ), and Grell-Freitas (GF), are selected for the precipitation and temperature, and at the same time from the perspective of We also compare and analyze the reasons for the differences between the three schemes from the perspectives of different altitudes and water vapor sources, so as to select the appropriate parameterization schemes for different weather patterns. The results show that: (1) the three selected cumulus convective parameterization schemes perform differently in precipitation and air temperature simulations: the KF scheme performs better in precipitation simulations, with the correlation coefficients of daily precipitation simulations ranging from 0.73 to 0.77, and the GF scheme performs better in air temperature simulations, with the correlation coefficients of daily temperature simulations ranging from 0.71 to 0.77. (2) the performances of the three schemes differ significantly at different elevations, with the KF and BMJ schemes performing better than the KF and BMJ schemes. The KF and BMJ schemes better show the correspondence between precipitation and topography along the Wuling-Dabashan Mountains. In the longitude profile, the simulation errors of precipitation in June 2015 and 2017 of the KF scheme are 5.96% and 6.06%, respectively, while the GF scheme describes the effect of topographic uplift too strongly, resulting in a larger variation of rainfall in the profile. (3) The sources of water vapor in the simulation results of the three schemes are different. The KF scheme shows that the Indian Ocean monsoon brings abundant water vapor, with less hydromorphic substances and concentrated cloud-water mixing ratios, which is more suitable for the simulation of precipitation in the middle and lower reaches of the Yangtze River, while the GF scheme shows that the South China Sea warm and humid airflow is stronger, with more hydromorphic substances and vigorous cloud development, which is more suitable for the simulation of precipitation in the areas of heavy convection frequent weather. (4) Different water vapor sources have little effect on the accuracy of the simulation results of the three cumulus convective parameterization schemes. Although June 2017 was more affected by water vapor from the Western Pacific than June 2015, the precipitation simulation results still show the best performance of the KF scheme.