Understanding vegetation dynamics and its response to climate changes is important for revealing the mechanisms of terrestrial ecosystem behaviour, predicting future vegetation growth, and thus guiding environmental management. The Jing River Basin (JRB) and the Beiluo River Basin (BLRB), two typical ecoenvironmentally vulnerable regions on the Loess Plateau in China, were selected as case study regions. Based on long-term Normalized Difference Vegetation Index (NDVI) datasets, the time-lag relationships between NDVI and climatic factors (precipitation/temperature) as well as teleconnection factors (large-scale modes of climate variability and solar activity) were revealed. Additionally, ridge regression models were established to quantitatively explore the response of vegetation dynamics to climate change. Results indicate that: (1) NDVI in autumn showed significantly increasing trend (p < 0.01), whereas that in spring and summer was insignificant; (2) there was a time-lag of more than one month between spring/winter NDVI and precipitation/temperature behaviour, and summer NDVI exhibited no lag with temperature but a one month lag with precipitation; (3) regarding the time-lag effects, precipitation was the driving factor of NDVI variations in spring, whereas sunspots dominated NDVI variations in autumn; (4) when time-lagged teleconnection factors were considered, the explanation of the climate effect on the vegetation dynamics in three seasons all relatively increased by >95%, which indicates that the prediction accuracy of NDVI was significantly improved; (5) in summer, time-lagged climatic and teleconnection factors explained <20% of NDVI variations, whereas when soil moisture and base flow were considered, the explanation of NDVI changes in the JRB and BLRB relatively increased by 37.4% and 65.1%, respectively. These findings highlight that considering the time-lag effect of climatic and teleconnection factors has important significance for the accurate monitoring of underlying surface dynamics under changing environment.

了解植被动态及其对气候变化的响应对于揭示陆地生态系统行为的机制，预测未来植被的增长并从而指导环境管理非常重要。选择了黄土高原两个典型的生态环境脆弱地区-靖江流域（JRB）和北罗河流域（BLRB）作为案例研究区域。基于长期归一化植被指数（NDVI）数据集，揭示了NDVI与气候因子（降水/温度）以及遥相关因子（气候变化和太阳活动的大规模模式）之间的时滞关系。此外，建立了岭回归模型以定量探索植被动态对气候变化的响应。结果表明：（1）秋季NDVI呈显着增加趋势（p <0.01），而春季和夏季则不显着；（2）春季/冬季NDVI与降水/温度行为之间存在一个多月的时滞，夏季NDVI随温度没有滞后，但有一个降水滞后一个月；（3）关于时滞效应，降水是春季NDVI变化的驱动因素，而黑子则是秋季NDVI变化的主导因素。（4）当考虑时滞遥相关因子时，对气候对三个季节植被动态影响的解释都相对增加了> 95％，这表明NDVI的预测准确性得到了显着提高；（5）在夏季，时滞的气候和遥距因子解释了NDVI变化的<20％，而考虑土壤水分和基础流量时，JRB和BLRB中NDVI变化的解释分别增加了37.4％和65.1％。这些发现突出表明，考虑气候和遥距因素的时滞效应对于在变化的环境下准确监测地下表面动力学具有重要意义。