Research On Cloud Fearture With Remote Sensing Methods Such As Satellite

Cloud plays an important role in the radiative budget of climate system, it is not only an indicator, but also a conditioner in the process of weather and climate change. Cloud is the object of artificial precipitation enhancement.Use many remote sensing observation techniques to study the distribution of temperature and humid profile, cloud liquid water and the relationship between cloud and precipitation etc., which are important for identification of operational rainfall enhancement condition, efficient capture of the extending cloud area, scientific standards of artificially seeding clouds and catalytic.The research in this paper, use FY2C satellite inversion cloud parameter products, ground-based MWRP microwave radiometer, Cloudsat cloud parameter products, surface rainguage data and support vector machine(SVM) to study the relationship between cloud optical thickness retrieved by FY2C and liquid water content from MWRP, the distribution about relative humidity inside and outside cloud and its threshold value, the relationship between cloud and precipitation.The results reported here show that:1. Use statistical fitting methods to study the relationship between MWRP microwave radiometer and cloud optical thickness retrieved by FY2C on May to November,2008 in ShouXian, after removing the sample points of precipitation. Meanwhile, use cloud top temperature retrieved by FY2C to divide 0℃～-40℃into four categories for all sample points and statistical fitting independently.The results show that:Correlation coefficient is above 0.5 commonly which is a good one to 0.68 for warm cloud, it implys that cloud optical thickness retrieved by FY2C has a good indicator role in vertical integral liquid water content, the extent of cloud thickness in the vertical direction and so on about clouds features.2. Precipitation forecasts using SVM is based on cloud products retrieved by FY2C.Take the cloud products retrieved by FY-2C geosynchronous satellites to be a feature component and the class label is Micaps rainfall data of an hour with which makes a prediction model about precipitation and nonprecipitation. Results show that all of the models have a good accuracy rate to non-precipitation forecasts and a lower accuracy rate to precipitation forecasts.When liquid water path from microwave radiometer as a kind of feature component put into the model, the accuracy rate of precipitation forecasts category will be a little raising especially.3. The studies have been carried out by statistical analysis of the distribution about relative humidity inside and outside cloud, the distribution about connected cloud mass relative humidity and the distribution about different height relative humidity inside and outside cloud with Cloudsat 2B-GEOPROF product and ECMWF-AUX product in global different area.Take relative humidity threshold value under various conditions and the accuracy rate according to accumulate frequency intersections of relative humidity inside and outside cloud. The research about statistical relative humidity in different area show that the max or min relative humidity threshold value to identify cloud is corresponds to the max or min relative humidity accumulate frequency value of statistical analysis in different categories.Europe has a max relative humidity threshold value of 69%. Asia has a min relative humidity threshold value of 47%, the difference is up to 20%.The research about statistical connected cloud mass relative humidity in different scale show that relative humidity threshold value is all above 75%, the discriminant accuracy rate is above 80%, there is a higher enhancement in the relative humidity statistical results compared to unconnected cloud mass.A bigger scale of connected cloud mass, a higher relative humidity discriminant threshold value. The research about statistical relative humidity in different height show that when accumulate frequency value is up to 25%, the corresponding relative humidity is 77%, lower than 79% of low-lever cloud, higher than middle-level cloud.It is the same for warm cloud when its accumulate frequency is up to 25%.The relative humidity threshold value has a reducing trend according to the raising of the cloud top height, there is little distinction of relative humidity between low-lever cloud and middle-level cloud, but a significantly reducing in high-level cloud.The discriminant accuracy rate is above 80% in different height.The relative humidity threshold value in low-lever cloud and middle-level cloud is bigger than relative humidity threshold value of no distinguish between height, smaller than relative humidity threshold value of connected cloud mass of no distinguish between height.