雅鲁藏布江流域NDVI变化与风沙化土地演变的耦合关系

Abstract

运用1982-2010年的两种NDVI数据集（Pathfinder AVHRR和SPOT VEGETATION），以及1975、1990、2000和2008年4期遥感数据，通过GIS技术、人工目视解译和灰色关联分析方法，研究了雅鲁藏布江流域NDVI变化和风沙化土地演变的耦合关系，结合1957-2007年降水和气温逐日气象资料，探讨了气候变化对其耦合关系的影响。结果表明：（1）流域内1982-2010年NDVI的年际变化总体上呈波动式增长的趋势。NDVI空间分布呈现由下游向中上游逐渐降低的趋势，以米林宽谷最大、马泉河宽谷最小。（2）2008年流域内共有风沙化土地273 697.54hm²，呈现由江源区马泉河宽谷向中下游递减的趋势。1975-2008年流域内风沙化土地呈缓慢增长趋势，以1990-1999年增长率最高，2000-2008年的增长率最小。（3）对于NDVI年变化和植被生长季（7-9月份）的变化，马泉河宽谷受平均气温的影响最大，日喀则宽谷和山南宽谷受年降水量的影响最大；米林宽谷NDVI的年变化受风沙化土地扩展的影响最大，植被生长季变化受年降水量的影响最大。（4）不同宽谷段NDVI与风沙化土地年变化的关联度自下游向中上游呈总体减小的趋势。流域尺度NDVI植被生长季变化主要受平均气温和年降水量的影响，非植被生长季（10月-翌年6月）变化主要受风沙化土地扩展的影响。;Based on two NDVI datasets (Pathfinder AVHRR NDVI and SPOT VEGETATION NDVI) from 1982 to 2010 and four series of aeolian sandy land data visually interpreted from the Landsat images of 1975, 1990, 2000, and 2008 with the aid of geographic information system (GIS) technology, the coupling relationship between variations of NDVI and change of aeolian sandy land in the Yarlung Zangbo River basin of Tibet, China were analyzed using grey correlation analysis method. Thereafter, their coupling relationship with climate factors (temperature and precipitation) were discussed extracted from meteorological data collected from eight monitoring stations in or around the study area during the period of 1957-2007. The results showed that: 1) The annual mean NDVI in the Yarlung Zangbo River basin demonstrated a fluctuating growing trend during the period of 1982-2010. Meanwhile, the spatial distribution of annual mean NDVI presented a gradually decreasing trend from Milin wide valley in the downstream to the middle and upper reaches of the river basin gradually, with the largest NDVI value found in Milin wide valley and the smallest value in Maquanhe wide valley. 2) There were 273 697.54 hm² of aeolian sandy lands in the river basin in 2008, which showed a spatially decreasing trend from the Maquanhe wide valley in the headwater area to the middle and lower reaches. The area of aeolian sandy lands in the river basin increased slowly from 1975 to 2008, with the quickest increase during the 1990-1999 period and slowest increase during the 2000-2008 period. 3) The main factors which influenced the variations of annual mean NDVI and variations of NDVI in growing season (July to September) was the annual mean temperature in Maquanhe wide valley, while being annual precipitation in Shigatse wide valley and Shannan wide valley, respectively. However, in Milin wide valley, the variations of annual mean NDVI were mainly influenced by the expansion of aeolian sandy land, while that of NDVI in growing season were mainly influenced by annual precipitation. 4) The degree of correlation between variation of annual mean NDVI and annual change of aeolian sandy land in different wide valleys showed a decreasing trend totally from the downstream to the middle and upper reaches of the river basin. Besides, the variation of NDVI in growing season in the basin was mainly affected by mean temperature and annual precipitation, while in non-growing season (October to June) it was mainly affected by the expansion of aeolian sandy land.