Abstract: In order to use and develop the cloud water resources in dry season in Hainan Province effectively, and increase the surface rainfall to alleviate the drought, aiming at Hainan typical cloud, the author studied the nonrandomized effect testing method based on the optimum combination of statistical test and physical test. Through analog deviation discrimination method, the author selected the compared cloud for impact assessment from the software automatically, analyzed the catalytic effect of every rocket convective cells of the western part of Hainan and evaluated the non- convective cells by the method of non- randomized regional comparative analysis. The results showed that, in the echo top aspect, the catalytic effect of convective cells was better, in terms of the increase of liquid water content, the catalytic effect of regional comparative clouds was obvious. Through the curve of radar echo difference of each parameter, the growth trend of target cloud could be clearly showed by the slope, as well as the demise time of compared cloud. As a whole, although the analysis methods and objects were different, it could complement and demonstrate each other, which greatly improved the accuracy of Hainan localized rainfall enhancement effect.