利用遥感(RS)监测手段和地理信息系统(GIS)技术,获取了黄河源区在1975、1990、和2005年3个时期的土地利用/覆盖数据,并通过叠加分析,获得了该地区土地利用/覆盖的时空转移特征.此外,通过县级行政区对源区各县域内的土地利用/覆盖变化特征进行了统计分析,以期为该地区的环境管理提供科学依据.结果表明:在1975-2005年间,黄河源区的环境退化非常明显,土地利用/覆盖变化主要表现为耕地、沙地、滩地和水库、坑塘面积增加;沼泽地面积减小;高覆盖度草地面积减小,中、低覆盖度草地面积增加.从县级行政区划上来看,耕地的增加主要分布于贵南、同德和泽库三个县;林地面积在玛沁和甘德两县有大面积的减小趋势;新增的沙地主要分布在玛多、共和、曲麻莱和若尔盖县;新增水库、坑塘则主要分布于共和、和贵南两县;沼泽地面积的减少绝大部分发生在若尔盖县;高覆盖度草地在甘德和玛沁两县增加非常明显,但在玛曲、玛多、达日、兴海、阿坝、若尔盖、红原等县均有大面积的减小趋势.因此,在黄河源区开展环境保护或环境治理工作,应根据不同区域的具体环境问题采取相应的治理措施.
The vector data of used land and covered land of the study area in 1975, 1990 and 2005 were gained by means of RS and GIS technologies. Through overlay analysis, the spatial and temporal transfer characteristics of land-use and land-cover change (LUCC) were obtained. For providing scientific basis for environmental management, a statistic analysis was also made on the characteristics of change in land-use and land-cover change of every county. It is found that the environmental degradation in the source regions of the Yellow River had been very obvious from 1975 to 2005, and the main changing characteristics of land-use and land-cover change had been: cultivated land, sandy land, beaches, the reservoir and pits area increasing; marsh area reducing, high coverage grassland area reducing, medium and low coverage grassland area increasing. From the view of statistical data, the cultivated land mainly increased in Guinan, Tongde and Zeku Counties; woodland area significantly reduced in Maqin and Gande Counties; the newly increased sandy land mainly appeared in Maduo, Gonghe, Qumalai and Zoige Counties; the newly increased reservoir and pits mainly appeared in Gonghe and Guinan Counties ; the marsh area mainly decreased in Zoige County. High covered grassland increased obviously in Gande and Maqin Counties, but in Maqu, Maduo, Dari, Xinghai, Aba, Zoige and Hongyuan Counties there were a large area of decreasing coverage. Therefore, in the source regions of the Yellow River, environmental protec- tion or environmental management should be taken as targeted measures depending on specific environment in different areas.