Premium
Development and application of a hydroclimatological stream temperature model within the Soil and Water Assessment Tool
Author(s) -
Ficklin Darren L.,
Luo Yuzhou,
Stewart Iris T.,
Maurer Edwin P.
Publication year - 2012
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1029/2011wr011256
Subject(s) - environmental science , soil and water assessment tool , snowmelt , hydrology (agriculture) , streamflow , surface runoff , watershed , swat model , calibration , streams , discharge , stream bed , drainage basin , geology , geography , ecology , computer network , statistics , cartography , geotechnical engineering , mathematics , machine learning , computer science , biology
We develop a stream temperature model within the Soil and Water Assessment Tool (SWAT) that reflects the combined influence of meteorological (air temperature) and hydrological conditions (streamflow, snowmelt, groundwater, surface runoff, and lateral soil flow) on water temperature within a watershed. SWAT currently uses a linear air‐stream temperature relationship to determine stream temperature, without consideration of watershed hydrology. As SWAT uses stream temperature to model various in‐stream biological and water quality processes, an improvement of the stream temperature model will result in improved accuracy in modeling these processes. The new stream temperature model is tested on seven coastal and mountainous streams throughout the western United States for which high quality flow and water temperature data were available. The new routine does not require input data beyond that already supplied to the model, can be calibrated with a limited number of calibration parameters, and achieves improved representation of observed daily stream temperature. For the watersheds modeled, the Nash‐Sutcliffe (NS) coefficient and mean error (ME) for the new stream temperature model averaged 0.81 and −0.69°C, respectively, for the calibration period and 0.82 and −0.63°C for the validation period. The original SWAT stream temperature model averaged a NS of −0.27 and ME of 3.21°C for the calibration period and a NS of −0.26 and ME of 3.02°C for the validation period. Sensitivity analyses suggest that the new stream temperature model calibration parameters are physically reasonable and the model is better able to capture stream temperature changes resulting from changes in hydroclimatological conditions.