不同方法确定内蒙古地区伏沙地空间分布的分析——以内蒙古西乌珠穆沁旗分析区为例-analysis of different methods to determine the spatial distribution of v - sand land in inner mongolia - a case study of the analysis area in west wuzhumuqin banner, inner mongolia.docxVIP

Different method to determine the spatial distribution of sandy land in Inner Mongolia --The case in West WU AHU MU QIN banner study area Abstract At present the most serious global ecological environment problem is one of the land desertification. It not only threaten the entire human survival environment, but also restricting global economic development and affect social stability important factors. One of the main reasons desertification is the sinking sandy land excessive exploitation and utilization, Cause sinking sandy land with its incomplete soil structure bear agriculture and animal husbandry dual pressures and ecologically fragile belt truly. This paper, taking remote sensing image interpretation for rely on, Field investigation for important means, Establish sinking sandy land explain marks, Sure sandy land space and Comparative analysis of remote sensing image classification technology, Explore a set of suitable for Inner Mongolia. In West wu ahu mu qin banner sinking sandy land space distribution method, Sure sinking sandy land distribution, meanwhile, In-depth analysis sinking sandy land formation and development background，Explore natural factors and human activities on the sinking sandy land influence. Research shows that, Suitable for Inner Mongolia in West West wu ahu mu qin banner area sinking sandy land the best band combination for 453,This band combination can strengthen. sinking sandy land information Establish interpreting marks .After person and computer interactive interpretation classification, Use supervised classification and decision tree classification method Contrast and precision analysis verify informed: person and computer interactive interpretation can accurately differentiate ground feature between type and The overall classification accuracy reached 89%,accord with application scope and higher than other classification method. Supervised classification because of remote sensing image exist the same object different spectrum or