黄土高原作为我国典型的生态敏感区，总体植被覆盖率低，其植被变化与陆地水文要素紧密关联。本项目以揭示黄土高原植被变化对陆地水储量的多尺度响应为目标，拟开展以下研究内容：采用归一化植被指数（NDVI）和时间序列分析方法，研究黄土高原植被的时空变化特征；联合遥感NDVI数据、GRACE数据和土壤湿度等资料，定量分析不同时空尺度下植被—陆地水和植被—土壤湿度的时滞相关性；利用GRACE多尺度小波分解信号、GLDAS水文模型以及NDVI数据，分别构建逐步回归模型，分析不同时间尺度的水储量变化信号与土壤湿度以及植被在垂向上的对应关系。预期研究成果可为黄土高原的生态治理和水资源保护提供理论支撑和科学依据。

The Loess Plateau is a typical ecologically sensitive region in China, with a series of environmental problems such as desertification and soil erosion. Analyzing the vegetation activity and its connection with hydrological conditions is critical for better understanding the ecological changes in Loess Plateau. In this study, using multiple sources of data, including satellite normalized difference vegetation index (NDVI), GRACE terrestrial water storage (TWS) estimates and soil moisture (SM) data, we carry out a comprehensive analysis of vegetation changes in Loess Plateau at different temporal and spatial scales. By comparing the connection of NDVI-TWS and NDVI-SM, we aim to explore the different influence of TWS and SM on vegetation cover. Time series of GRACE TWS anomalies are decomposed into different temporal frequencies using a discrete wavelet transform, and analyzed against time series of the GLDAS SM and MODIS NDVI anomalies in a stepwise regression. The question we seek to address are (1) how does the sensitivity of the NDVI in response to TWS changes vary spatially, and (2) which temporal components of TWS are most significant in influencing the NDVI for different land use type in Loess Plateau. This study could provide theoretical support and scientific basis for regional ecological management and water resource protection.