青海省农业生态区参考作物蒸散量的遥感反演

根据青海省农业生态区15个气象站2003年气象资料,应用FAO推荐的Penman-Monteith方程计算参考作物蒸散量(ET0),利用MODIS高程(DEM)、地表温度(LST)及法国SPOT卫星的归一化植被指数(NDVI)遥感影像资料,提取遥感数据并耦合到时间分辨率为旬,空间分辨率为1km,将其与计算所得ET0进行相关分析,运用MATLAB软件进行模型拟合,获得该地区的ET0遥感反演模型及其适用条件,使用Arcgis9.3对利用该模型反演的结果进行了空间分布规律分析。研究结果表明,2003年7—11月各旬遥感因子DEM、NDVI和LST与ET0的线性关系显著,其中DEM与ET0呈显著负相关关系,NDVI和LST与ET0呈极显著正相关关系;获得了该地区基于遥感数据的旬ET0三元线性遥感反演模型,该模型通过α=0.01的F和t显著性检验,模型效果极显著,适用期为7—10月;对2004年计算及反演结果表明,ET0空间分布从西北往东南方向递增,低海拔处往高海拔处递减;模型反演平均相对误差为-1.9%,有较好的反演结果。

Inverse Reference Crop Evapotranspiration in the Ecological Agriculture Area of Qinghai Province Based on RS

Based on the meteorological data collected from 15 weather stations in 2003 in the ecological agriculture area of Qinghai Province,the reference crop evapotranspiration(ET0) was calculated by the FAO Penman-Monteith equation.Using the remote sensing image of MODIS elevation(DEM),land surface temperature(LST) and the normalized difference vegetation index(NDVI) were got by the French SPOT satellite,the remote sensing data could be extracted from them and coupled to the time resolution of ten-day,spatial resolution of 1 km.Then these data were used for the correlation analysis with ET0,and the inversion model of ET0 could be generated by using MATLAB software.At the same time,the regional ET0 remote sensing inversion model and its application conditions were got.The spatial distribution analysis of the ET0 calculated by the model from July to November in 2004 was finished with Arcgis9.3.The results of the study showed that: from Jul.to Nov.in 2003,the linear relationships between ET0 and the remote sensing factors of DEM,NDVI and LST were significant,and the linear relationship between ET0 and DEM was obvious negative correlation,and the linear relationships between ET0,NDVI and LST were very significant positive correlation.From the study,the multiple linear model of regional ET0 remote sensing inversion model was obtained,and passed the F and t significant inspection(α=0.01),so the model was extremely significant.The calculation and inversion results of 2004 illustrated that in space distribution ET0 increased from northwest to southwest direction,and decreased from low altitude to high altitude;the model inversion average relative error was 1.9%,which showed the model had a good inversion result.