CAREER: Dynamic connectivity: a research and educational frontier for sustainable environmental management under climate and land use uncertainty

动态景观代表了水文、环境和人为特征的网络，这些特征协同工作，赋予生态系统利益并提供社会需求。在土地利用改变和气候变化的压力下，景观正日益面临风险。如果我们要可持续地管理我们共享的水资源，了解景观如何动态地将水、沉积物和营养物质转移到河流中，以及人类在调节这种连通性方面发挥的作用至关重要。因此，这项工作背后的驱动问题是“人类如何使我们周围的景观变得更糟，我们如何能够更好地管理它们？”该项目将回答这些问题，并通过将农业、市政和雨水专业知识与高频水生传感、深度学习建模和大样本水质数据集相结合，推进可持续水管理的研究和教育前沿。这项研究将在确定气候、地貌和人为环境对河流系统的景观负荷的大小、持续时间和程度方面取得基础科学进展。今天的学生将成为未来的利益相关者，他们的教育将通过实践经验深深融入这个项目，这将使他们具备处理大数据和解决社会最大的水挑战的信心和沟通技巧。当代水文科学研究认识到水循环大多数方面连通性的重要性；然而，尽管它无处不在，连通性通常是定性的或在静态的结构背景下评估的。拟议的研究在向连通性动态评估的方向发展方面具有变革性的潜力。这个项目将量化美国在时间和空间上的动态连接。这将通过利用来自150多条河流的硝酸盐和浊度高频水生传感器来实现，这些传感器将作为深度学习模型的训练数据。此外，动态连接的数学描述将告知连接的主要途径。可解释的机器学习技术将动态景观属性如何导致河流水质影响联系起来。此后，将通过为从业者开发的web应用程序评估使用动态连接作为管理工具的潜力。这些成果将直接用于未来利益相关者的教育和培训，包括在高中环境中建立对大数据的信心，以及在大学生中培养科学传播技能。该项目由水文科学和促进竞争研究的既定计划（EPSCoR）共同资助。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Dynamic landscapes represent a network of hydrologic, environmental, and anthropogenic features that work in tandem to confer ecosystem benefits and provide for societal demands. Increasingly, landscapes are at risk under the growing pressures of land use alteration and climate change. Understanding how landscapes dynamically connect the transfer of water, sediment, and nutrients to rivers and the role humans play in modulating this connectivity is crucial if we are to sustainably manage our shared water resources. Thus, the driving questions behind this work are “how have humans changed the landscapes around us for the worse and how are we able to manage them for the better?” This project will answer these questions and advance the frontiers of research and education for sustainable water management by coupling agricultural, municipal, and stormwater expertise together with high-frequency aquatic sensing, deep learning modeling, and large-sample water quality datasets. This research will generate fundamental scientific advances to identify the magnitude, duration, and extent of landscape loading to river systems across climatological, geomorphic, and anthropogenic settings. The education of today’s students, who will become tomorrow’s stakeholders, is deeply embedded in this project through hands-on experiences that will equip them with the confidence and communication skills to handle big data and tackle society’s grandest water challenges. Contemporary research in hydrologic sciences recognizes the importance of connectivity in most aspects of the water cycle; however, despite its ubiquity, connectivity is often assessed either qualitatively or in a static, structural context. The proposed research has the potential to be transformative in moving toward a dynamic assessment of connectivity. This project will quantify dynamic connectivity through time and across space for the United States. This will be achieved by leveraging high-frequency aquatic sensors for nitrate and turbidity from over 150 rivers, which serve as training data for a deep learning model. Further, a mathematical description of dynamic connectivity will inform dominant pathways of connection. Explainable machine learning techniques will link how dynamic landscape attributes lead to riverine water quality impacts. Thereafter, the potential to use dynamic connectivity as a management tool will be assessed through a web application developed for practitioners. The outcomes will lead directly into the education and training of the stakeholders-of-tomorrow, including through building big data confidence in high school settings and science communication skills in college students.This project is jointly funded by Hydrologic Sciences and the Established Program to Stimulate Competitive Research (EPSCoR).This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.