Abstract: The performance of the CMA-CPEFS(Cloud and Precipitation Explicit Forecasting System) v2.0 model in forecasting supercooled cloud water was evaluated by comparing its outputs with aircraft observations collected over the Danjiangkou basin on 10 May 2024, supplemented with Fengyun-4B satellite data. Assuming that the simulated cloud type is generally consistent with observations, microphysical characteristics—such as droplet spectra, effective particle diameter, and liquid water content—are compared between aircraft observations and model forecasts. Results show that the model accurately reproduces the occurrence time, location, and altitude of supercooled water, as well as the vertical variation trends of liquid water content and effective particle diameter. However, it systematically underestimates the overall content of supercooled cloud water and the effective diameter of supercooled cloud droplets and significantly underestimates the droplet number concentrations—especially for droplets with diameters around 14 μm—at altitudes of 5.9 and 6.3 km (MSL). In contrast, the modeled droplet spectrum at 6.9 km agrees more closely with observations. An analysis of the vertical distribution and source–sink terms of cloud hydrometeors reveals that ice-phase particles, particularly graupel, dominate the cold cloud. The 5–8 km layer contains some supercooled water, together with small quantities of snow particles and cloud ice, with condensation serving as the main source and riming as the main sink for supercooled droplets. Overall, the study identifies key deficiencies in the model’s microphysical parameterizations of supercooled cloud water and provides a scientific reference for future improvements to microphysical schemes and operational cloud seeding activities..