荒漠化程度遥感定量化评价中的尺度问题

基于遥感手段的荒漠化程度定量化评价是当前遥感应用研究中的一个热点问题。在荒漠化评价中应当坚持主导性、综合性等原则。传统的荒漠化自动评价都是在离散的像元的尺度上进行的，未能综合考虑像元所处的环境，使得评价结果带有一定的片面性，评价结果图像中出现大量的噪音。若直接采用滤波方法消除这些噪音，则又会使评价结果偏离真实值。以地学原则为指导，在前人研究的基础上，探讨了以植被盖度和空间信息确定地理单元，提出了以地理单元为尺度进行荒漠化评价的方法。实践证明，基于地理单元的荒漠化定量化评价克服了传统方法的缺陷，评价结果更接近于地理事实。

Scale of quantitative desertification evaluation by remote sensing

Study of desertification quantitative evaluation by remote sensing information is a hot spot of remote sensing application research. The principles about dominance and integrity should be insisted in the desertification evaluation, but traditional desertification automatic evaluation was carried out on the scale of discrete pixel and ignored environment of the pixel, so the result of evaluation had one-sidedness and noise. This paper studies the method of quantitative desertification evaluation in the scale of geographical unit based on the former researcher. The result shows that the quantitative desertification evaluation based on geographical unit is more close to geographical substance.