Research And Application Of Methods Based On Artificial Precipitation Effect Of The Cluster Of Floating Contrast Historical Regression In Statistical Tests

In this paper a new method of evaluating the efficiency of non-randomized artificial precipitation enhancement-Cluster-Analysis-based Floating Control historical regression Method (CA-FCM) was presented, which made use of cluster analysis and grid interpolation of precipitation.First, based on the historical data of 20 years of Henan Province , the cloud seeding operation cases in April and October in the central region of Henan Province were evaluated by cluster-analysis-based floating control historical regression method with uniform precipitation and atmospheric precipitable water as meteorological covariates, cluster-analysis-based floating control historical regression method with uniform precipitation as meteorological covariate , and floating control historical regression method(FCM).Then, six evaluation methods (double ratio analysis evaluation method using regional rainfall tendency control for single cloud seeding operation case, regression analysis evaluation effects on the bases of regional correlation and developing tendency of rainfall, multiple regression analysis evaluation effects on the bases of regional developing tendency of rainfall, cluster-analysis-based floating control historical regression method with uniform precipitation and atmospheric precipitable water as meteorological covariates, cluster-analysis-based floating control historical regression method with uniform precipitation used as meteorological covariate, and float ing control historical regression method ) were compared and analyzed with the case of the cloud seeding operation on 5 April 2002 in Henan Province.In the end, cluster-analysis-based floating control historical regression method with uniform precipitation and atmospheric precipitable water as metoorological covariates could more validly evaluate efficiencyof cloud seeding operations and significant level of CA-FCM method was higher than the other methods, because it adopted cluster analysis which highly improved the correlativity between rainfall distributions in the control area and target area, and used grid interpolation which enhanced exactness of calculating precipitation rainfall, and chose atmospheric precipitable water as the covariant which increased the inferential accuracy of natural rainfall on the cloud seeding operational area.It was shown from the result of analysis and comparison that the evaluation efficiency for the non-randomized cloud seeding operations could be improved by using appropriate physical covariate as control factor and increasing the correlativity between rainfall distributions in the control area and target area .