论文部分内容阅读
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。
The scale effect often restricts the accuracy of quantitative remote sensing inversion. The spatial scale conversion of geoscience information is a necessary requirement of production practice. However, the commonly used scale conversion model mostly uses the spectral data to calculate the difference, which is not suitable for scale-up and scale-down conversion . Because of the stochastic and structural characteristics of the regional variation of soil moisture content data, ASTER image pixels with a resolution of 15m are taken as the basic units and the point Kriging method is used to complete the data of soil moisture content with a resolution of 15m to 7.5m Scale conversion, from the similarity degree of fractal dimension, the conversion result is reasonable; and the block Kriging method is used to scale up the surface data measured from point to 7.5m resolution, The result of scale and scale-down conversion is compared with the average value of measured samples. The results show that the mean error of soil water in the sample with 7.5m resolution is between 1.5782-5.019, and the data of scaling soil moisture obtained by block Kriging And the point Kriging method to obtain the scale of soil moisture content of the data between the error is 1.2825-5.0481, shows that the kriging method considers the relationship between the point and the surrounding, the obtained soil moisture content is better than did not consider the spatial difference The average soil moisture content of the average.