Font Size: a A A

Fine-scale Spatial Estimation Of Precipitation Based On GPM IMERG Precipitation Product

Posted on:2020-02-23Degree:MasterType:Thesis
Country:ChinaCandidate:Q Q HeFull Text:PDF
GTID:2370330623457224Subject:Geography
Abstract/Summary:PDF Full Text Request
Spatialized precipitation information is of great significance for the research and management of water resources.However,it is difficult to obtain the spatial distribution of high precision and fine precipitation in mountainous areas under the current technical means directly.To solve this problem,this paper takes the Fujian-Zhejiang-Jiangxi region as the research area.Based on the modified water vapor factor,the GPM IMERG precipitation products are statistically downscaled to construct a precipitation background field of high resolution in the study area.Then,from the perspective of terrain precipitation mechanism,FNL reanalysis data and digital elevation model are used to construct terrain precipitation incremental model and obtain the spatial distribution of terrain precipitation increment in the study area.Finally,the precipitation information fusion model is constructed based on the GWR algorithm,and the spatial distribution of fine precipitation under the complex terrain of the Fujian-ZhejiangJiangxi region in 2015 is obtained.The conclusion is as follows:(1)TRMM and GPM are consistent in the monthly variation of two kinds of precipitation products,which can reflect the time and space changes of precipitation in the study area accurately.GPM is better than TRMM(the previous generation)precipitation products in study area.In addition,higher resolution makes it more suitable for precipitation background field.(2)From the point of downscaling results,the GWR model is better than the OLS model in terms of fitting effect and final accuracy.MOD05 water vapor data is more suitable as a downscaling factor than MOD13 vegetation index,but the accuracy of water vapor products will affect the fitting effect and the result of the downscaling model,so the error correction has to be carried out before use.(3)There is a large difference in the increment of terrain precipitation in each month of the study area.The high value distribute in the coastal hilly,mountainous areas and the surrounding low-lying mountainous areas under most circumstances.The distribution of the windward slope,the rain shadow area and the precipitation increment are consistent with the prevailing wind direction,wind speed,water vapor conditions and the distribution of local terrain.Therefore,the application of the precipitation increment model constructed in this paper is feasible in the study area.(4)The accuracy index of the result of fusion model constructed in this paper has been improved significantly compared with the background field and the MRE decreases by 41.4% monthly.Monthly MRE is less than 20% except August,with most of the months being less than 15% and the smallest in December being 6.75%.The precipitation estimation model,which combines the terrain precipitation increment,has clear physical meaning and accurate estimation results.In short,it can provide reliable spatial precipitation data and estimation schemes for hydrology and climate research in the Fujian-Zhejiang-Jiangxi region.
Keywords/Search Tags:GPM, Terrain Precipitation, Downscaling, Precipitation Information Fusion, GWR
PDF Full Text Request
Related items