基于MODIS数据和气象观测数据的气温空间插值方法比较

Comparation of the Temperature Spatial Interpolation Methods Based on MODIS Data and Meteorological Observation Data

本文以福建省福州市为研究区域,利用MODIS影像数据反演福州市地表温度;同时利用相同时间的福州市及其周边的32个气象站日平均气温数据,结合数字高程模型(DEM)中提取出的地形因子,研究温度与经度、纬度、海拔高度等的相关关系,提出了福州市地形因子与气温相关模型,利用该模型插值得到福州市的地表温度;再利用气温垂直递减规律进行插值得到地表温度。对比3种结果可知,对于气象数据容易获得,气象站点较多且均匀分布在研究区时,利用气象数据插值地表温度较快,但计算结果比实际值略高;对于气象数据不易获得或站点较少时,可以用影像数据来反演地表温度,其精度受提取参数的精度的制约,在参数使用正确情况下比插值方法反演精度高。

This study makes use of MODIS image data and the daily average temperature data from the 32 meteorological stations to reckon the surface temperature in Fuzhou, and by combining terrain factors extracted from the digital elevation model DEM to explore and study the correlations among the temperature, longitude, latitude, altitude and other factors, and establishes a model to interpolate the surface temperature in Fuzhou. Using meteorological data to interpolate surface temperature is faster, but the calculation is slightly higher than the actual value. Using image data to reckon land surface temperature is affected by the accuracy of parameters. When using correct parameters, the calculation of land surface temperature from the images has higher accuracy than the interpolation by the measured meteorological data. The parameters limit the accuracy of land surface temperature estimation, and the future research should focus on selection of parameters. Based on the topographic effects on temperature, we build a correlation model between topographical factors and temperature, and the accuracy is relatively high. In the future research, we will try to bring other factors such as the terrain, slope, aspect and slope position into the model to improve the accuracy.