This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).
The Susquehanna River Basin (SRB) is the largest tributary to the Chesapeake Bay. Without this flow the estuary could not sustain its extraordinary diversity and productivity of aquatic life. Management of the SRB requires a balance between the competing societal and environmental demands placed on its freshwater resources. A long term goal of this research is to address the questions: ?How do humans and climate impact the sustainability of the water resources within large river basins? What role do large rivers play in the global climate system??. We seek to transform our ability to detect and/or predict the impacts of long-term changes on the hydrology of the SRB. We posit that our ability to understand human-climate impacts on the SRB?s environmental systems will require a paradigmatic shift towards community level evaluations of watershed models? predictive power and adaptive design of the basin-scale observation networks under uncertainty. This work will contribute time-evolving sensitivity maps for a suite of models that will allow us to explore how hydrologic flow controls change across a strong 3 year climatic gradient (i.e., drought in 2001 to wet conditions in 2003). The sensitivity maps will then be used to identify the key uncertainties to be incorporated into an ensemble-based data assimilation framework that explicitly considers systematic sources of bias in observations and model structures. The data assimilation framework will provide baselines for judging the forecasting skills of models (i.e., their mean prediction errors) as well as the value of observations (i.e., those observables that minimize the time evolving covariance of predictions). This proposed project will initiate the extension of WATERS cyberinfrastructure (CI) to facilitate the publication of the ?uncertainty? in hydrologic observations and web-services for sharing ensembles of model inputs/outputs to encourage scientific replication and extension of our prediction experiments. Overall the modeling and CI components of this work will help clarify the value of information provided by uncertain measurements of precipitation, evapotranspiration, river flow, soil moisture, and groundwater levels.
