Hebei, Shanxi provinces in North China suffered from severe drought in the spring of 2017. The drought had a certain impact on crop sowing and growth. Cloud seeding is helpful to alleviate drought. To carry out cloud seeding scientifically and accurately, based on the Cloud and Precipitation Explicit Forecast System (CPEFS_v1.0) and observational data, a typical low-trough cold-front cloud system precipitation process that occurred in North China from May 22 to 23, 2017, is studied. Choosing the Xingtai station in Hebei Province for an example, the cloud system structure at different precipitation stages and cloud seeding conditions are explored. The results show that when Xingtai is located behind the ground cold front and in front of the 700 hPa trough, strong rainfall occurs due to uplift of the front. The maximum rainfall in 10 minutes exceeds 2.0 mm, and the water vapor flux below 850 hPa at this stage is greater than 21 g·hPa^-1·cm^-1·s^-1. As the cloud anvil moves over Xingtai, the high-level cold cloud composed of ice and snow crystals transforms into a cold-warm mixed cloud. The mixing ratio of low-level cloud water, middle-level supercooled water and graupel is significantly higher. The main process of rainfall formation includes the collision of raindrops and cloud droplets, the melting of graupel. When Xingtai is located behind the 700 hPa trough and in front of the 500 hPa trough, the precipitation transforms into light rain formed by deep stratus clouds. The content of water vapor and cloud water in the middle and low layers decreases apparently at this stage. The rising movement mainly appears in the cold area and weakens obviously. The cloud system is still a mixed cold-warm cloud, but the mixing ratio of supercooled water and graupel decreases. The precipitation is mainly produced by the melting of graupel. The precipitation gradually dissipates after the 500 hPa trough transits. The cloud seeding time mainly occurs in the frontal uplifting-heavy rain period and the deep stratus cloud-light rain period. The seeding height is located at 4.0-7.9 km. And the strong seeding area is at the height of 4.0-5.5 km, the area with abundant supercooled water.

Based on FY-2G satellite retrieve products, the normal upper level and surface data from MICAPS, houly precipitation and L-band sounding data, a hailstorm in Lhasa of Tibet Plateau from 19:00 BST to 20:00 BST on 22 June 2016 was examined. The results are as follows: (1) The cloud precipitation model could predict the precipitation area of Tibet, but there were some deviations for heavy precipitation center and rainfall intensity. (2) The evolution of clouds could be predicted truly by this model, the simulations of clouds’ moving direction and velocity were basically consistent with the satellite monitoring. (3) The model could forecast the macroscopic characteristics of convective cloud in the plateau well, although the vertical development of the convective clouds was weaker than that of the retrieval based on FY-2G satellite data. The deviation of the cloud top height was about 1.0-2.0 km and that of the cloud top temperature was about 10-20 ℃. (4) The prediction of the cloud vertical characteristics was in good agreement with monitoring results by FY-2G satellite and the upper-air sounding. The property of the cloud vertical structure and the height of characteristic temperature layer were close to the observations.