干旱灾害是一种具有出现频率高、波及范围大、持续时间长等特点的自然灾害，但不同于其他自然灾害，干旱灾害的发生具有区域不定性，这对于旱情监测预报预警能力是一种严峻的挑战。因此深入了解大尺度干旱事件的演变规律和变化趋势，对于评估旱灾风险、提升抗旱管理水平具有重大现实意义。

近年来，受到气候变化及人类活动的影响，长江流域极端干旱灾害事件时有发生，导致流域内土地荒漠化、江河湖泊水位下降以及草场植被退化等生态环境问题日益凸显。因此，本文以长江流域陆地植被生态系统为研究对象，从植被生态系统供需水量平衡的角度出发，结合气象、遥感等数据模拟流域内不同植被覆盖类型下的供/需水资源量，并分析其时空分布特征；在辨识植被水分亏缺动态特征的基础上，耦合SPEI (Standardized Precipitation Evapotranspiration Index) 的旱度评价模式，构建适用于长江流域植被生态系统的生态干旱评价模型SSDI (Standardized Water Supply–Demand Index)，并实现长江流域生态干旱的动态监测。同时采用游程理论分析流域内近20年的SSDI，实现流域生态干旱特征指标的提取，进一步探讨了长江流域陆地植被生态干旱的空间格局与演变规律。本研究的主要结论如下：

(1)2000-2019年间长江流域多年平均蓝/绿水资源以及绿水系数分别为422.78mm、685.73mm和68.59%，绿水资源约是蓝水资源的1.62倍。近20 a来长江流域蓝/绿水资源和绿水系数呈不显著增加趋势，2000-2019年间蓝水与绿水资源变化速率分别为3.26mm/a 和2.27mm/a。从年内分配上看，蓝绿水资源在7-8月份较多，占全年的29-32%；在1-2月较少，占全年的5-6%。从时间分布来看，蓝水资源呈现东南高西北低的分布格局，绿水资源则呈现由西向东逐渐增加的趋势，而绿水系数在西北部相对较高而在东南部相对较低。

(2)长江流域多年平均的归一化植被指数(Normalized Difference Vegetation Index, NDVI)与生态需水量的空间分布整体上均呈现出由西北向东南逐渐递增的趋势，江源地区的NDVI较低，而长江中下游平原地区的NDVI普遍较高。从时间分布上看，长江流域多年平均生态需水量约为368.89mm，且流域内生态需水量呈现逐渐减少的趋势，减少幅度为-0.81mm/a。从年内分配上看，长江流域的月平均生态需水量约为30.74mm，其中在植被生长季(5~9月)中，生态需水量大约可占到全年总生态需水量的68.52%。

(3)本文构建的SSDI所监测的干旱事件与历史干旱事件有很好的吻合，且SSDI与SPEI有着较为相似的变化趋势，相关系数达到0.836，其中38.1%的区域的SSDI与SPEI呈现强相关关系。SSDI与NDVI的相关系数(R²=0.94) 要明显高于SPEI与NDVI的相关系数(R²=0.88)，且长江流域57.23%的地区的SSDI和增强型植被指数(Enhanced Vegetation Index, EVI)之间的相关系数可达0.6以上。以上结果表明SSDI在长江流域生态干旱的监测中具有较好的适用性，能更好地反映植被状况的变化，且具有较强的可靠性，可用于长江流域的生态干旱监测。

(4)长江流域的植被生态干旱主要发生在冬春季节，且在流域北部的干旱多呈现短期高频的发展态势，而流域南部则呈现长期低频的发展态势，干旱强度较高的地区多分布在中下游平原的江河湖泊流域附近。此外流域内近年来的SSDI整体呈现上升趋势，干旱灾害发生的风险有降低的态势。长江源区、中下游平原以及四川东南部大部分区域有湿润化的趋势，这与植被生态系统的恢复以及流域植被覆盖度的变化有着密切的关系。

Drought disaster is a natural disaster with characteristics of high frequency of occurrence, large ripple area and long duration. Unlike other natural disasters, the occurrence of drought disaster is regionally unpredictable, and it can happen anywhere, which is a serious challenge for our drought monitoring and forecasting early warning capability. Therefore, it is significant to understand the evolution patterns and trends of large-scale drought events in order to assess drought risks and improve drought management.

Effected by climate change and human activities, extreme drought events have occurred frequently in the Yangtze River basin (YRB) in recent years, which led to increasing ecological and environmental problems such as land desertification, declining water levels of rivers and lakes, and degradation of grassland vegetation in the basin, bringing great influence on the ecological environment of the YRB. Therefore, based on the supply and demand balance of vegetation ecosystem, terrestrial vegetation ecosystem in the YRB was taken as the object of this paper, the water supply/demand of different vegetation cover types in the basin was simulated by combining meteorological and remote sensing data, and its spatial and temporal distribution characteristics were analyzed. Based on the identification of the dynamic characteristics of vegetation water deficit, the ecological drought evaluation model SSDI (Standardized Water Supply-Demand Index), which is applicable to the vegetation ecosystem in the Yangtze River basin, was constructed by coupling the SPEI (Standardized Precipitation Evapotranspiration Index) to achieve the dynamic monitoring of ecological drought in the Yangtze River basin. At the same time, the theory of run was used to analyze the SSDI in the basin for the past 20 years to extract the characteristic indicators of ecological drought in the basin, and the spatial pattern and evolution pattern of ecological drought of terrestrial vegetation in the YRB were further explored. The main results of this study are as follows:

(1) The average blue/green water resources and green water coefficient in the YRB during 2000-2019 were 422.78 mm, 685.73 mm, and 68.59%, respectively, and the green water flow is about 1.62 times of the blue water resources in the YRB from 2000 to 2019.The blue/green water resources and green water coefficient in the YRB showed no significant increasing trend in the past 20 years, and the change rates of blue/green water resources were 3.26mm/a and 2.27mm/a during 2000-2019, respectively. In terms of time distribution, the blue/green water resources are more in July and August, accounting for 29-32% of the whole year, and less in January and February, accounting for 5-6% of the whole year. In terms of temporal distribution, blue water resources show a distribution pattern of high in the southeast and low in the northwest, while green water resources show a trend of gradual increase from west to east, and the green water coefficient shows a relatively high in the northwest and a relatively low in the southeast.

(2) The spatial distribution of the normalized difference vegetation index (NDVI) and the ecological water demand in the YRB showed an overall increasing trend from northwest to southeast, with a lower NDVI in the source region of the Yangtze River and a generally higher NDVI in the middle and lower reaches of the Yangtze River plain. In terms of temporal distribution, the multi-year average ecological water demand in the YRB is about 368.89 mm, and the ecological water demand in the basin shows a gradual decrease, with a decrease of -0.81 mm/a. In terms of time distribution, the monthly average ecological water demand in the YRB is about 30.74 mm, of which the ecological water demand in the vegetation growing season (May to September) can account for about 68.52% of the total ecological water demand in the year. In the vegetation growing season (May to Sep), the ecological water demand can account for about 68.52% of the total ecological water demand of the year.

(3) The drought events monitored by the SSDI constructed in this paper are in great agreement with the historical drought events, and the SSDI and SPEI have relatively similar trends, and the correlation coefficient reaches 0.836, and the SSDI and SPEI in 38.1% of the areas show a strong correlation. the correlation coefficient between SSDI and NDVI (R²=0.94) is significantly higher than that between SPEI and NDVI (R²=0.88), and the correlation coefficient between SSDI and NDVI in 57.23% of the YRB is significantly higher than that between SSDI and SPEI. The correlation coefficient between SSDI and NDVI (R²=0.88) was significantly higher than that between SPEI and NDVI (R²=0.88), and the correlation coefficient between SSDI and Enhanced Vegetation Index (EVI) in 57.23% of the YRB could reach above 0.6. The above results indicate that SSDI has good applicability in ecological drought monitoring in the YRB, can better reflect the changes of vegetation condition, and has strong reliability, and can be used in ecological drought monitoring in the YRB.

(4) The vegetation ecosystem drought in the YRB mainly occurs in winter and spring, and the drought in the northern part of the basin mostly shows a short-term high-frequency development, while the southern part of the basin shows a long-term low-frequency development, and the areas with higher drought intensity are mostly distributed near the river and lake basins in the middle and lower reaches of the plain. In addition, the overall SSDI in the basin has shown an increasing trend in recent years, and the risk of drought disaster occurrence has a decreasing trend. There is a trend of wetting in the source region of the Yangtze River, the middle and lower reaches of the plain and most areas in southeastern Sichuan, which is closely related to the recovery of the vegetation ecosystem and the change of vegetation cover in the basin.