Improvement On Statistical Evaluation Method Of Artificial Precipitation Enhancement Effects

With the economic development and the population increase, consuming water of industry and life is increasing greatly. Artificial precipitation receives more and more attention because it is one of important measures to resolve the crisis. Although the key point of the field is precipitation enhancement effect of cloud seeding all the while, it is still hard to calculate accurately for the complicated physical processes of cloud and precipitation.In order to advance the inferential accuracy of natural rainfall, in this paper, the cluster analysis with statistical test was used based on the original Cluster-Analysis-based Floating Control historical regression Method (CA-FCM). Then, considering that precipitation type in Henan Province's springtime and autumntime is almost supercooled stratiform cloud events, the paper increases supersaturation water vapour content with respect to ice (SSWVCi) as a new controllable variable to more validly evaluate efficiency of Cloud seeding operation.Comparing the new method with the original based on the cloud seeding operation cases in April in Henan Province with the covariates of rainfall, atmospheric precipitable water, and SSWVCi the results show that the evaluation efficiency can be improved greatly.Then, four combined types of covariate ((l)uniform precipitation;(2)uniform recipitation and atmospheric precipitable water;(3)uniform precipitation and SSWVCi;(4) uniform precipitation, atmospheric precipitable water, and SSWVCi) were compared and analyzed with the 20 cloud seeding operation cases in April and October in Henan Province. The conclusion is that the new controllable variable is beneficial to improve the evaluation efficiency.Finally, a "blank" test was conducted. The modified new method——the originalCluster-Analysis-based Floating Control historical regression Method having the cluster analysis with test of significance was used to evaluate enhancement efficiency of natural rainfall of 20 raindays without artifical seeding with three covariates of rainfall, atmospheric precipitable water, and SSWVCi, the result shows that the significance level of rainfall increment is bad.