变化环境下极端干旱组合遭遇风险与脆弱性评估

Due to the strong influence of climate change and human activities, the water cycle has significant changes in the atmosphere, surface, soil, rivers and groundwater processes. The spatial and temporal evolution of the drought events is extremely complicated. In the past few years, the situation of drought occurring continuously, concurrently and alternately was even more pronounced, and a series of ecological, resource and environmental problems were induced. It is of great significance to study drought coincidence probability and drought vulnerability assessment under changing environment. Therefore, based on the hydrology multidisciplinary integrated cross-cutting research methods, mathematical description of the independent and dependent structure for meteorological hydrological elements is deduced and multiple abrupt changes are recognized, the key indicators of hydrological drought are identified combined with the multi-source hydrological information, the dynamic temporal and spatial evolution of multiple variables of drought under the influence of climate change and human activities is clarified, and the change pattern and driving mechanism of the drought coincidence risk are revealed. Based on this, a response model to clarify the mechanism of the impact of hydrological drought on plant greenness is established, the vulnerability of drought disasters considering the critical conditions of vegetation against drought disasters is assessed. Research can provide decision support for the rational development, optimal allocation, efficient use of water resources and improved farming systems and disaster preparedness and mitigation.

受气候变化和人类活动强烈影响，水循环大气、地表、土壤、河川和地下水等过程发生显著变化，干旱时空演变规律极为复杂，干旱连发、并发、交替发生态势更加显著并诱发一系列生态、资源、环境问题。变化环境下极端干旱组合遭遇风险及脆弱性评估成为亟待研究的重点问题。为此，以水文学为基础的多学科综合交叉研究方法，推导水循环过程气象水文要素独立结构和产汇流相依结构数学描述与变异识别方法，辨识能反映干旱形成、转化及整体现象的综合干旱关键指标，探求气候变化和人类活动影响下干旱多重特征变量动态时空演化规律，阐明极端干旱组合遭遇风险变化格局及驱动机制；以此为基础，建立多维响应模型揭示干旱与植被绿度指标复杂响应关系，确定区域植被生态抵御干旱灾害的临界条件，评估干旱灾害脆弱性。研究工作可为水资源合理开发、优化配置、高效利用以及改进耕作制度以及防灾减灾提供决策支持。

干旱已成为我国灾害防治的薄弱环节和减少因灾损失的关键与瓶颈，加快变化环境下干旱复杂演变机理研究是关系社会经济发展全局的一项重大而紧迫的任务。本项目采用以水文学为基础的多学科综合交叉研究方法，探求了气候变化和人类活动影响下干旱多重特征变量动态时空演化规律，阐明了极端干旱组合遭遇风险变化格局及驱动机制，确定了区域植被生态抵御干旱灾害的临界条件，评估了干旱灾害脆弱性。研究团队选取对环境变化较为敏感的长江上游区域，提出了基于copula方法的水文多变量依赖过程变异检测，研究了水文过程变异检测、干旱非平稳性演变以及人类活动和气候变化对极端干旱过程的影响；引入了隐马尔可夫过程推求了干旱转移概率及干旱组合遭遇概率，揭示变化环境下干旱事件连发、并发、交替 发生的规律；提出了基于干旱转移概率的干旱预测模型以及基于深度学习的干旱预测模型，在长江中游两个地区的实验结果表明基于干旱转移概率的模型预测精度比传统模型明显提高；在此研究成果基础上，团队开展了融合植物绿度指数多维响应的干旱脆弱性评估，包括土地利用/覆盖变化对水文过程影响研究以及植被遥感变化对干旱胁迫的多维响应评价。本研究取得一批具有科学创新和重要学术价值与应用前景的研究成果，为干旱灾害的防治提供重要的决策支持和科学依据。