ABSTRACT

A strong understanding of watershed function is necessary for responsible stewardship of water resources. However, fully characterizing or understanding all the complex processes that occur within watersheds is often not feasible and prohibitively costly. Many locations of concern simply do not have enough long-term data to predict how solutes will be transported through watersheds, and lack the time and money required to make such predictions. The purpose of this project is to develop a framework whereby better predictions of solute transport can be made, even in data scarce regions. The project leverages existing data and high-resolution models, constructed at sites that have already been characterized in great detail, to assess how flow and transport processes in similar watersheds are related. This information will lead to simple statistical models that can capture the complexity of real watersheds based on less detailed characterizations. The models are expected to allow the translation of knowledge from sites where great investments have been made to improve models of relatively data-poor sites. The project is also creating new educational tools, training undergraduate and graduate students, and reaching out to applied watershed managers to better understand their needs for real-world applications of solute transport models.

The approach used in this research focuses on using recent multi-domain correlated velocity models (MD-CVMs) to represent coupled subsurface and surface flow and transport in watersheds. Lagrangian particle-based numerical methods along streamtubes are the core of this approach, which couples interactions between particles to accurately represent crucial mixing and reaction processes. The water and solutes from each streamtube interact as they come together, simplifying the watershed geometry into a tree without sacrificing process-level realism. The streamtube approach will also enforce velocity correlations, which is a novel feature at watershed scales that is lacking in previous models despite evidence that persistent correlations exist. The advantage of using velocity correlations is that they are conceptually simple but yield robust models that show promise across different sites. The resulting dynamically coupled, yet realistic, representations of watersheds will expand the tools available for understanding and optimally managing real watersheds.

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.

Please report errors in award information by writing to: awardsearch@nsf.gov.