A spatially explicit network-based model for estimating stream temperature distribution

approved: John P. Bolte The WET-Temp (Watershed Evaluation Tool Temperature) model is designed to take advantage of spatially explicit datasets to predict stream temperature distribution. Datasets describing vegetation cover, stream network locations, elevation and stream discharge are utilized by WET-Temp to quantify geometric relationships between the sun, stream channel and riparian areas. These relationships are used to estimate the energy gained or lost by the stream via various heat flux processes (solar and longwave radiation, evaporation, convection and advection). The sum of these processes is expressed as a differential energy balance equation applied at discrete locations across the stream network. The model describes diurnal temperature dynamics at each of these locations and thus temperature distribution across the entire network. WET-Temp is calibrated to a tributary of the South Santiam River in western Oregon, McDowell Creek. The mean differences between measured and modeled values in McDowell Creek were 0.6°C for daily maximum temperature and 1.3°C for daily minimum temperature. The model was then used to predict maximum and minimum temperatures in an adjacent tributary, Hamilton Creek. The mean differences between modeled and measured values in this paired basin were 1.8°C for daily maximum temperatures and 1.4°C for daily minimum temperatures. Influences of model parameters on modeled temperature distributions are explored in a sensitivity analysis. The ability of WET-Temp to utilize spatially explicit datasets in estimating temperature distributions across stream networks advances the state of the art in modeling stream temperature. Redacted for privacy