基于二次多项式回归模型的黑河流域TRMM数据降尺度研究  

Downscaling Study on TRMM 3B43 Data of the Heihe River Basin Based on Quadratic Polynomial Regression Model

作　　者：樊东[1,2] 薛华柱[1] 董国涛[2] 蒋晓辉[2] 张文鸽[2] 殷会娟[2] 郭欣伟[2] 

机构地区：[1]河南理工大学测绘与国土信息工程学院,河南焦作454000 [2]黄河水利科学研究院水利部黄土高原水土流失过程与控制重点实验室,郑州450003

出　　处：《水土保持研究》2017年第2期146-151,共6页Research of Soil and Water Conservation

基　　金：国家自然科学基金(91325201;41301496;41301030);水利部公益性行业科研专项经费项目(201401031)

摘　　要：在不同空间尺度下分别建立TRMM 3B43降水数据与数字高程模型(DEM)和归一化植被指数(NDVI)的二次多项式回归模型,将2001—2013年黑河流域TRMM降水数据的空间分辨率从0.25°提高到1km,并利用流域内9个气象站点实测数据对降尺度结果进行了检验。结果表明:降尺度方法不仅提高了TRMM数据的空间分辨率,数据的精确程度也有所提高;与传统线性回归模型降尺度方法相比,基于二次多项式回归模型获得的降尺度结果更接近于实测值,其结果更为准确;模型建立的尺度对最终降尺度结果精确性具有较大影响,0.50°是基于DEM和NDVI对黑河流域TRMM降水数据进行降尺度的相对最优尺度。Precipitation data with high accuracy and high spatial resolution are important to ecology,hydrology and meteorology.In this study,we established a Quadratic Polynomial Regression Model（QPRM）between TRMM 3B43 precipitation,Digital Elevation Model（DEM）data and Normalized Difference Vegetation Index（NDVI）on four different scales（0.25°,0.50°,0.75°and 1.00°）,and the TRMM 3B430.25°×0.25°precipitation fields were downscaled to 1km×1km for each year from 2001 to 2013.The downscaled precipitation estimates were subsequently validated against the in-situ observation data obtained from nine rain gauge stations in the period of 13 years in the Heihe River Basin.The results showed that both spatial resolution of data and the data quality were significantly improved.Compared with multiple linear regression model downscaling method,the downscaled result obtained by QPRM is more accurate and closer to the measurements from rain gauges.The modeling scale has a great influence on the accuracy of the downscaled results,and0.50°is the optimal scale to obtain high spatial resolution precipitation by downscaling TRMM 3B43 products using DEM and NDVI data.

关 键 词：TRMM数据降尺度 二次多项式回归模型 DEM NDVI 黑河流域 

分 类 号：P332.1[天文地球—水文科学]