克里格法的土壤水分遥感尺度转换

 尺度效应往往会制约着定量遥感反演的精度,对地学信息进行空间尺度转换是生产实践的必然要求,而常用的尺度转换模型多利用光谱数据进行差值计算,不适合升尺度和降尺度转换。由于土壤含水量数据具有区域变化量的随机性和结构性特点,本文以15m分辨率的ASTER图像像元为基本单元,采用点克里格法完成ASTER 15m至7.5m分辨率的土壤含水量数据降尺度转换,从分维数的相似程度上来看,转换结果是合理的;并利用块状克里格法对地面实测样点数据进行点到7.5m分辨率的面数据升尺度转换,将升尺度和降尺度转换结果与实测样点均值相比较,结果表明:7.5m分辨率的实测样点土壤水均值误差在1.5782-5.019之间,块状克里格法获取的升尺度土壤含水量数据与点克里格法获取的降尺度土壤含水量数据之间误差则为1.2825-5.0481,可见克里格法考虑了点与周边的关系,所获得的土壤含水量值要优于未考虑空间异质性的土壤含水量平均值。

Remote Sensing Scale Transformation of Soil Moisture Based on Block Kriging

The reasonable remote sensing sale transformation method is very important for RS data utilization.In order to find a suitable scaling model for ascending scale and downscaling transformation with spectral data,given the randomness and structural of soil moisture data as regional variation,point Kriging method accounting for spatial heterogeneity is used in this study to transfer from ASTER soil moisture data with 15m resolution to 7.5m resolution.The similarity degree of fractal dimension is shown that the transformation result is reasonable.Block Kriging method is engaged in the upscaling transformation for the field measured point soil content to 7.5m spatial resolution.The 7.5m resolution upscaling scale conversion results and average of measured sample point soil moisture with downscaling 7.5 m resolution remote sensing data were compared respectively.The result shows that 7.5m resolution soil moisture mean error of the measured sample points is in the range of 1.5782 to 5.019,the error between 7.5m resolution upscaling scale conversion results and downscaling 7.5 m resolution remote sensing data is from 1.2825 to 5.0481,which indicates that soil moisture value acquired by Kriging is much better than the soil moisture average without considering spatial heterogeneity under the same scale because of Kriging method taking account of the relationship between the points and the surrounding.