Applicability of Hydrologic Landscapes for Model Calibration at the Watershed Scale in the Pacific Northwest
Citation:
Sawicz, K., S. Leibowitz, C. Jones, R. Comeleo, AND P. Wigington, Jr. Applicability of Hydrologic Landscapes for Model Calibration at the Watershed Scale in the Pacific Northwest. 2016 Pacific Northwest Water Research Symposium, Corvallis, OR, April 18 - 19, 2016.
Impact/Purpose:
EPA’s Western Ecology Division has published and is refining a framework for defining landscapes hypothesized to have similar hydrologic response to precipitation events. These hydrologic landscapes (HL) were developed to help characterize the change in hydrologic response of watersheds under climate change at a variety of different scales. The HL framework was created for the Pacific Northwest (PNW) states of Washington, Oregon and Idaho. The HLs were developed using the National Hydrography Dataset Plus (NHD+) catchments. Classification components describing climate, seasonality of water availability, aquifer permeability, terrain, and soil permeability are then applied. This approach has the potential to assist in management of water quality and quantity and ecosystem status because HLs represent hydrologic response across a variety of spatial scales. One hypothesis set forward from this study is that hydrologic response, as inferred from the aggregation of HL assessment units, will improve the accuracy of hydrologic model predictions and calibration. It also contributes to deliverables under ACE CIVA-2.3.
Description:
The Pacific Northwest Hydrologic Landscapes (PNW HL) at the assessment unit scale has provided a solid conceptual classification framework to relate and transfer hydrologically meaningful information between watersheds without access to streamflow time series. A collection of techniques were applied to the HL assessment unit composition in watersheds across the Pacific Northwest to aggregate the hydrologic behavior of the Hydrologic Landscapes from the assessment unit scale to the watershed scale. This non-trivial solution both emphasizes HL classifications within the watershed that provide that majority of moisture surplus/deficit and considers the relative position (upstream vs. downstream) of these HL classifications. A clustering algorithm was applied to the HL-based characterization of assessment units within 185 watersheds to help organize watersheds into nine classes hypothesized to have similar hydrologic behavior. The HL-based classes were used to organize and describe hydrologic behavior information about watershed classes and both predictions and validations were independently performed with regard to the general magnitude of six hydroclimatic signature values. A second cluster analysis was then performed using the independently calculated signature values as similarity metrics, and it was found that the six signature clusters showed substantial overlap in watershed class membership to those in the HL-based classes. One hypothesis set forward from this study is that hydrologic response, as inferred from the aggregation of HL assessment units, will improve the accuracy of predictions set by hydrologic models. A framework is presented to use watershed-scale HL information for use in hydrologic model calibration.