Title: Functional Classification of Susquehnna River Basin Watersheds in an Uncertainty Framework
Project Type: Research
Start Date: 03/01/2006
End Date: 02/28/2007
Congressional District: 5th
Focus Categories: Hydrology, Surface Water, Models
Keywords: classification, watershed response, predictions in ungauged basins, ensemble predictions, uncertainty
Principal Investigator: Wagener, Thorsten
Federal Funds: $17,998
Non-Federal Matching Funds: $35,996
Abstract: Human activities and natural environmental change related to urbanization, agriculture, forestry, climate change etc. affect natural hydrological and ecological regimes, leading for example to increasing flood and drought risk, and to decreasing biodiversity due to negative impacts on riparian ecosystems in the Susquehanna River Basin as well as in many other regions of the US and beyond. Predictive tools are required that can assess impacts of land use and climate change on flow regimes and subsequently on aquatic ecosystems to support water resources management and land development decisions. Predictions of the response of ungauged or changed watersheds using current models are associated with large and often unknown uncertainties due to a lack of understanding regarding how watershed function (response) relates to watershed structure. The development of new tools to better understand and predict the response of watersheds to changes in land use is important for land use and development, water resources planning, and for investigations into the health of aquatic ecosystems.
A new approach to classification and predictions in ungauged basins is proposed here which involves the following steps:  Calculate streamflow indices that capture particular characteristics of the watershed response (e.g. runoff ratio, peak runoff, flow duration curve slope etc.).  Cluster the available watersheds based on their structural and based on their functional characteristics using both fuzzy and statistical clustering techniques.  Investigate correlations and develop regression relationships between watershed structure and watershed function based on clustering results.  Run a hydrologic model in a Monte Carlo framework and constrain the model using the regression results of step . Compare the predictions to currently used modeling approaches including use of the WinTR-55 and the Continuous API. We will utilize a dataset of about 60 watersheds mainly located in the upper Susquehanna River Basin and ranging from about 20 acres to 3848 km2.
The general objective of this research proposal is to improve the understanding of the relationship between watershed structure (e.g. topography, land use) and watershed function or response behavior (streamflow) for watershed classification and predictions in ungauged basins, through the development of a classification framework that relates streamflow indices to watershed characteristics. The classification framework will constrain the expected watershed response for ungauged locations and establish relationships between land use characteristics (e.g. percent forest cover) and streamflow behavior. This will support stormwater management (e.g. pre-development flows) and improve streamflow prediction in Pennsylvania.
Main deliverables of the project will be as follows:
Progress/Completion Report, PDF