基于MODIS数据的博斯腾湖流域植被变化及其与气候因子的关系

Variation of vegetation cover and its relationship with climatic factors in the Bosten Lake Basin using MODIS data

以博斯腾湖流域为研究区,基于2001-2016年时间序列的MODIS NDVI数据分析了研究区植被的时空变化趋势,并结合流域气象站点的气温、降水、日照时数和相对湿度数据分析了植被生长季累积NDVI和16天NDVI与气候因子之间的响应特征.结果表明:(1)流域植被覆盖变化呈改善趋势,生长季累积NDVI年变化率为0.014 a-1,16天NDVI变化率均为正值,植被改善趋势显著区域主要分布在高山草原湿地和农业灌溉区边缘的新增农田.(2)植被生长季累积NDVI主要受降水和相对湿度影响,植被总体生产力与水分条件关系最密切,生长季逐16天NDVI与同期气温和日照时数在植被生长初期和末期关系显著,而与降水没有显著的相关性,说明植被短期瞬时长势对热量条件更为敏感.(3)在植被生长不同阶段对气候变化具有不同的滞后效应,其中植被生长初期和末期对气温有0.5~1个月的滞后,生长盛期对降水有0.5~3个月的滞后、日照时数有1.5~2.5个月的滞后、相对湿度有0.5~2.5个月的滞后,揭示了植被不同生长阶段水热条件对其生长韵律的控制差异.

The response of vegetation to climate at different scales provides synthetic understandings of environmental changes. In this paper,the spatio-temporal variation of the vegetation and its relationship with climatic factors in the Bosten Lake Basin was studied based on MODIS NDVI dataset. The correlation and time lag between the NDVI（ 16-day NDVI and cumulative NDVI in growing season） and the climate factors（ air temperature,precipitation,sunshine duration and relative humidity） were investigated during 2001-2016. The results showthat：（ 1） The cumulative NDVI in growing season exhibited an increasing trend by 0.014 a-1.The most significant increasing areas are mainly located in the lower edge of permafrost belts and the newirrigated farmland areas.（ 2） The interannual variation of vegetation cover was strongly related with the water conditions. The significant correlation between16-day NDVI and the climate factors had highlighted the significant impacts from temperature and lower correlation with precipitation,which indicated that the short-term growth of vegetation is mainly sensitive to thermal conditions.（ 3） The time lag correlationbetween 16-day NDVI and the climate factors showed that vegetation had different time-lag effects on climate change during different growth stages. There was a 0.5-1 month time lag effect on temperature during the early and late stages of the growing seasons. However,the most significant effect during the fast growing seasons was the 0.5-3 month time lag on precipitation. This difference revealed the mechanism rhythms of hydrothermal conditions on the vegetation growth at different stages.