雅鲁藏布江流域干湿转换特征及植被动态响应

为准确识别气候变化条件下高寒流域干湿特征及植被动态响应,该研究以青藏高原东南部的雅鲁藏布江流域为例,基于全球陆面数据同化系统(global land data assimilation system,GLDAS)数据集,计算标准化降水蒸散指数(standardized precipitation evapotranspiration index,SPEI),结合归一化植被指数(normalized difference vegetation index,NDVI)探讨了雅鲁藏布江流域干湿转换特征及其对植被的影响。结果表明:GLDAS的降水与气温数据与实测数据一致性较好,可以用于分析雅鲁藏布江流域干湿转换特征;1982-2015年间雅鲁藏布江流域总体呈现湿润化趋势,但2000年前后流域干湿时空特征发生反转现象,即时间上,2000年以前流域呈现湿润化趋势,2000年以后呈现干旱化趋势,空间上,流域内干旱地区逐渐变湿,湿润地区逐渐变干;流域内约71.83%的区域SPEI与NDVI呈正相关,流域植被受流域干湿条件影响较大,降水和气温是植被动态变化的主要驱动因素;流域内92.17%以上的区域SPEI与土壤含水量呈极显著正相关,表明土壤含水量亦是影响流域干湿特性不可忽视的因素。该研究结果可为辨识高原及高寒区水循环变化过程及其驱动机制提供科学依据。

Characteristics of dry and wet conversion and dynamic vegetation response in Yarlung Zangbo River basin

Abstract: Understanding the mechanisms underlying drying-wetting cycles and their consequence for vegetation dynamic is important for sustainable eco-environmental development in alpine regions. We investigated these in this paper by taking the Yarlung Zangbo River (YZR) basin in southeast Qinghai-Tibet Plateau as an example. The standardized precipitation evapotranspiration index (SPEI) in the region was calculated using the global land data assimilation system (GLDAS) from 1982 to 2015 to represent the drying-wetting cycles. The spatiotemporal variation of the dry-wet cycles and the responsive vegetation dynamic was investigated using the remotely sensed NDVI (normal difference vegetation index) from GIMMS (global inventory modeling and mapping studies). The results showed that: 1) Spatiotemporal variation of both precipitation and surface air temperature calculated from GLDAS agreed well with the ground-truth data. 2) The spatiotemporal changes in the dry-wet cycles calculated from the SPEI from 1982 to 2015 showed that the YZR basin became increasingly wet from 1982-1999 but changed course in 2000 and has been become increasingly drier since. In particular, the arid areas showed a tendency of wetting whereas the humid areas tended to become drying. 3) The overall vegetation calculated from the NDVI had been in increase from 1982 to 2000 but changed course in 2000 and has been in decline since. In terms of spatial distribution, areas with higher NDVI value represented vegetation degradation, while areas with lower NDVI represented an improvement in vegetation. 4) Approximately 71.83% of the areas saw a positive correlation between the SPEI and NDVI, mainly in the middle and low reaches of the basin which have a high vegetation coverage. Areas showing negative correlation between SPEI and NDVI were small and not statistically significant. The high consistency between spatiotemporal variation of the NDVI and SPEI indicated that the drying-wetting cycles played an important role in vegetation dynamics. 5) The driving forces of the dry-wet cycles were precipitation, surface air temperature, potential evapotranspiration and soil water content, which were consistent with that estimated from SPEI. In addition, the SPEI showed that soil water content was the dominant factor impacting the drying-wetting in 92.17% of the areas in the region. Results of this study have important implications for evaluating water cycles and the associated vegetation dynamics in alpine regions.