NSF/EAR-BSF: Assessing controls on hydrologic connectivity, plant water availability and degradation risk in drylands with isotope tracers and Lagrangian modeling

由于缺乏与许多旱地生态系统相关的植被，降雨产生了大量的陆地流量，因此径流/径流过程对于向水可以渗透的植被场所提供流量变得重要，从而导致植物可用水资源的复杂模式。根据诸如植被和地形格局等容易观察到的特征预测这些流动模式，将有助于评价景观对沙漠化的脆弱性。此外，通过优化植被选择、坡面等高线和种植设计，为恢复活动提供设计策略。本研究旨在开发改进的技术，通过测试径流-径流过程的模型与同位素示踪剂的实验行为，并通过开发新技术来提高模拟植被和地形表面的真实性，从而实现这种预测。该研究将支持两名博士研究生（一名在以色列，一名在美国），并在加州大学伯克利分校和内盖夫本古里安大学的研究地点之间进行研究交流。它将与加州大学伯克利分校环境设计学院和以色列的Keren Kayemeth LeIsrael-Jewish National Fund的合作伙伴共同开展外展活动，利用研究成果为土地恢复规划和指导方针提供信息。该项目将根据植被的空间组织、山坡地形和风暴特征，推进旱地对径流-径流过程中断的脆弱性预测。这种预测的目的是确定为防止退化而进行干预的高优先级，并为恢复设计提供信息，重点是改变植被分布和山坡形式。以往基于空间格局推广植物水分有效性预测的建模尝试在两个方面受到限制：(i)用于进行预测的模型通常是根据径流过程的间接度量进行测试的，例如径流地块的净排放量，而不是根据风暴事件期间场址到场址的水源-汇输送的观测结果进行测试；因此，它们在径流场地内约束源汇行为的能力是值得怀疑的。用于将旱地地表特征与水再分配联系起来的模式实验依赖于理想化地表的个别实现。因此，对不确定性的描述很差，对真实景观的结果度量的意义也不清楚。该方案利用同位素示踪实验来揭示以色列Lehavim长期生态研究基地一个典型的旱地山坡上植被的连通性和水分可用性模式。这些模式将被用来测试径流-径流模型的预测，通过使用耦合欧拉-拉格朗日建模技术来调整这些模型来预测水的拉格朗日输运。然后，使用测试模型进行的模型实验将利用新兴的多点统计领域，以实现景观表面特征的现实实现，从而使预测的水可用性、退化风险或恢复质量的不确定性得到表征。该研究将根据径流在山坡上被捕获或通过从植被覆盖的斜坡出口而丢失的可能性，为沙漠化脆弱性的景观尺度控制提供见解。

Due to the lack of vegetation associated with many dryland ecosystems, rainfall generates significant overland flow so that runoff/runon processes become important for providing flows to vegetated sites where water can infiltrate, leading to complex patterns of water resources available to plants. Predicting these flow patterns from easily observable features such as vegetation and topographic patterns, would be instrumental in evaluating landscape vulnerability to desertification. In addition, it would provide design strategies for restoration activities by optimizing vegetation selection, hillslope contouring and planting design. This research aims to develop improved techniques to enable such predictions through models that test runon-runoff process against the experimental behavior of isotopic tracers and through the development of new techniques that improve the realism of modeled vegetation and topographic surfaces. The research will support two doctoral research students (one in Israel and one in the US), and research exchange between research sites at UC Berkeley and Ben Gurion University of the Negev. It will enable outreach activities that use the research findings to inform land restoration planning and guidelines, in conjunction with partners at the UC Berkeley College of Environmental Design and the Keren Kayemeth LeIsrael-Jewish National Fund in Israel.The project will advance predictions of dryland vulnerability to disruption of runon-runoff processes based on the spatial organization of vegetation, hillslope topography and storm characteristics. The aim of such predictions is to identify high priorities for intervention to prevent degradation, and to inform restoration designs, focusing on modification of the vegetation distribution and hillslope form. Previous modeling attempts to generalize the predictions of plant water availability based on spatial pattern were constrained in two ways: (i) The models used to make predictions are generally tested against indirect metrics of the runon-runoff process, such as net discharge from a runoff plot, rather than against observations of site-to-site source-sink transport of water during storm events; thus their ability to constrain the source-sink behavior within a runoff site is questionable. (ii) The model experiments used to connect dryland surface features to water redistribution rely on individual realizations of idealized land surfaces. Consequently, uncertainty is poorly characterized, and the meaning of the resulting metrics for real landscapes is unclear. This proposal exploits isotopic tracer experiments to reveal patterns of connectivity and water availability to vegetation on a well-characterized dryland hillslope at the Lehavim Long Term Ecological Research site in Israel. The patterns will be used to test the predictions of runoff-runon models by adapting these models to predict Lagrangian transport of water using Coupled Eulerian-Lagrangian Modeling techniques. Model experiments with the tested model will then draw on the emerging field of multi-point statistics to enable realistic realizations of landscape surface features, allowing uncertainty in predicted water availability, degradation risk or restoration quality to be characterized. The research will provide insights into landscape-scale controls on desertification vulnerability based on the likelihood that runoff is captured on a hillslope, or is lost via export from vegetated slopes.