Advanced Search
QU Xuebin, FU Ya'nan, YUAN Xiuzhi, XIE Xiaoli, BAI Ming. 2020: Applicability analysis of GPM-IMERG daily precipitation data in Inner Mongolia. Torrential Rain and Disasters, 39(3): 293-299. DOI: 10.3969/j.issn.1004-9045.2020.03.010
Citation: QU Xuebin, FU Ya'nan, YUAN Xiuzhi, XIE Xiaoli, BAI Ming. 2020: Applicability analysis of GPM-IMERG daily precipitation data in Inner Mongolia. Torrential Rain and Disasters, 39(3): 293-299. DOI: 10.3969/j.issn.1004-9045.2020.03.010

Applicability analysis of GPM-IMERG daily precipitation data in Inner Mongolia

  • The applicability of GPM-IMERG daily precipitation data is analysed by comparing daily precipitation data of GPM-IMERG and 119 meteorological stations in Inner Mongolia from 2015 to 2017, in order to provide scientific basis for the application of GPM-IMERG data in Inner Mongolia. Analysis shows that:(1) The accuracy of GPM-IMERG to capture precipitation days is 74.3%, and to ignore non-precipitation days is 85.5%. The overall error of GPM-IMERG for distinguishing precipitation and non-precipitation increases with latitude. (2) The number of stations with positive and negative average errors of GPM-IMERG precipitation in Inner Mongolia is approximately the same. The average absolute error is 4.1 mm. Spatially, it shows a decreasing distribution from southeast to northwest. (3) GPM-IMERG overestimates in low precipitation while underestimates in high precipitation. When the precipitation is greater than or equal to 25 mm, GPM-IMERG precipitation is 30% less than the precipitation of meteorological stations on average. (4) The correlation coefficients of GPM-IMERG precipitation and station precipitation can pass the significance test of 0.01 level in each city, among which, Chifeng City has the highest correlation coefficient and Wuhai City has the lowest. For different seasons, the correlation coefficient is the highest in summer and the lowest in spring. For different vegetation types, the slope of GPM-IMERG in high vegetation coverage area is closer to 1, and the correlation coefficient is also higher.
  • loading

Catalog

    Turn off MathJax
    Article Contents

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return