论文部分内容阅读
植被覆盖度是衡量地表植被状况的重要指标,其时空分布和变化对政府进行区域规划和决策起着重要作用。MODIS植被指数是区域植被覆盖度提取的重要数据源。利用2000年1月到2009年10月的MODIS 250m归一化植被指数16d合成产品(MODIS NDVI)和其他MODIS辅助数据估算三江源区的植被覆盖度并分析时空格局变化趋势。在混合像元二分模型的基础上改进了NDVI参数的确定方法,然后利用重构后的MODIS数据估算植被覆盖度。与2007年8月的野外采样点数据比较,估算精度是87.72%,相关系数r为0.889 7,表明该模型估算大面积植被覆盖度是可行的。将获得的植被覆盖度分为5个等级,从年最大化植被覆盖度Mfc的角度进行10a里植被覆盖度的年际变化趋势分析和时空格局变化分析。趋势分析结果表明,2000～2009年三江源区植被覆盖波动式变化,东北部地区得到改善,西部地区在退化,总体呈现退化趋势;时空格局变化分析结果表明,植被覆盖破碎度降低,趋于集中化分布;不同等级覆盖度分布的复杂化程度降低。
Vegetation coverage is an important indicator to measure the status of surface vegetation. Its spatial and temporal distribution and changes play an important role in government’s regional planning and decision-making. The MODIS vegetation index is an important data source for the extraction of vegetation coverage in the region. The MODIS NDVI and other MODIS auxiliary data from MODIS 250m from January 2000 to October 2009 were used to estimate the vegetation coverage in the Three Rivers and to analyze the trend of spatial and temporal patterns. The method of determining NDVI parameters is improved on the basis of the binary mixture model, and then the vegetation coverage is estimated using the reconstructed MODIS data. Compared with the field sampling data in August 2007, the estimation accuracy is 87.72% and the correlation coefficient r is 0.889 7, which shows that it is feasible to estimate the vegetation coverage in a large area. The vegetation coverage obtained was divided into five grades. From the perspective of annual maximum vegetation coverage (Mfc), interannual variation trend and temporal and spatial pattern change of vegetation cover in the area of 10a years were analyzed. The trend analysis shows that the vegetation cover in the Three-River Source Region fluctuated from 2000 to 2009, the northeastern region was improved and the western region degenerated, showing a general trend of degeneration; the spatial and temporal pattern analysis showed that the degree of vegetation cover fragmentation decreased and tended to concentrate Distribution; different levels of coverage of the complexity of the degree of reduction.