基于不同植被指数的TRMM数据降尺度及误差校正研究

Downscaling and Error Correction of TRMM Data Based on Different Vegetation Indices

[目的]对不同时间尺度的热带测雨卫星(TRMM)数据进行空间降尺度及误差校正研究,为华中地区洪涝灾害监测等提供科学参考。[方法]主要借助增强型植被指数(EVI)和归一化植被指数(NDVI)分别运用地理加权回归(GWR)模型实现2001-2019年华中地区TRMM数据的空间降尺度,并结合地理差异分析(GDA)和地理比率分析(GRA)对年、季和月的降尺度结果进行误差校正,通过气象站数据对校正前后的数据进行对比分析。[结果]①TRMM数据和气象站数据的决定系数(R 2)在年(0.630)、季(0.710~0.865)和月(0.637~0.875)尺度都表明了TRMM数据在华中地区具有较好的适用性;②通过GWR模型实现了TRMM数据空间分辨率由0.25°到1 km的降尺度转换,且TRMM EVI数据精度优于TRMM NDVI数据,说明华中地区TRMM数据与EVI的关系比NDVI更为密切;③对优选的TRMM EVI数据分别进行GDA,GRA校正,结果表明GDA校正结果优于GRA校正,且在降雨量越多的月份校正效果越好。[结论]在华中地区,EVI比NDVI更加适合TRMM数据降尺度研究。降尺度数据采用GDA校正比GRA校正效果更为显著。

[Objective]The spatial downscaling and error correction of tropical rainfall measuring mission(TRMM)data at different time scales were researched in order to provide references for flood disaster monitoring in Central China.[Methods]This article mainly used geographically weighted regression(GWR)model to achieve spatial downscaling of TRMM data from 2001 to 2019 with the help of enhanced vegetation index(EVI)and normalized difference vegetation index(NDVI),and compared and analyzed the annual,seasonal and monthly downscaled data through meteorological station data.Then combined with the geographic difference analysis(GDA)and geographic ratio analysis(GRA),the downscaling results of the year,quarter and month were corrected for error,and the data before and after the correction were compared and analyzed.[Results]①The coefficient of determination(R 2)of TRMM data and meteorological station data in year(0.630),season(0.710~0.865)and month(0.637~0.875)all showed that the TRMM data had better applicability in Central China.②The spatial resolution of the TRMM data was downscaled from 0.25°to 1 km through the GWR model,and TRMM EVI data had better accuracy than TRMM NDVI data,indicating that TRMM data in Central China had a closer relationship with EVI than NDVI.③GDA and GRA corrections were performed on the optimized TRMM EVI data.The GDA correction results were better than the GRA corrections,and the correction effect was better in months with more precipitation.[Conclusion]In Central China,EVI is more suitable for TRMM data downscaling research than NDVI,and downscaling data using GDA correction is more effective than GRA correction.