基于小流域尺度的土地石漠化信息遥感提取

为增强石漠化信息提取的客观性，文章提出了基于小流域的图斑确定方法，并以贵州西部为研究区，基于2013-06-16的Landsat遥感影像和DEM（30 m）数据，分析小流域图斑分割在石漠化信息提取中的应用；并以2013-06-16的Landsat遥感影像为基础提取土地利用边界，以其为地理单元提取石漠化信息。在此基础上构建石漠化稳定指数（DSI），分析不同地理分割单元对石漠化信息提取的影响。结果表明，研究区2种地理单元的中度及以下石漠化面积比率＞80%，而无明显、轻度、中度、重度和强度石漠化提取结果差异分别为66.5%、26.8%、72.4%、75.4%和55.9%，石漠化稳定指数仅0.15，不同地理分割单元对石漠化信息提取有显著影响。基于小流域图斑分割的方法，具有明确的地理意义和良好的稳定性，是石漠化信息提取中理想的分割方法。

Extraction ofRemote Sensing Information ofKarst Rocky Desertificationon the Scale ofSmall Watershed

A segmentation approach based on small watershed patches to karst rocky desertification information extraction was presented to improve the objectivity of karst rocky desertification mapping. Taking the western region of Guizhou Province as a study case, this paper tested the segmentation approach based on small watershed patches to rocky desertification information extraction by using DEM (30 m) and an image of Landsat 8 captured on June 16, 2013. Consequentially, desertification stability index (DSI) was developed to analyze the effects of different patches on karst rocky desertification information, which were extracted from land use patches based on the image captured in 2013. The results showed that the percentage of moderate desertification or better extracted from two kinds of patches was over 80% of the karst area in study case. There was a variable difference in rocky desertification based on the two categories of patches, and the values of non-obvious, slight, moderate, severe and intense desertification were 66.5%, 26.8%, 72.4%, 75.4% and 55.9% respectively. The value of DSI was just 0.15. The segmentation method based on small watershed patches had a significant effect on karst rocky desertification information extraction. Finally, the image segmentation method based on small watershed patches indicated a geographical significance and stability, which was an effective method of karst rocky desertification information extraction.