Brad Barnhart

Stream temperature is one of the most important factors for regulating fish behavior and habitat. Therefore, models that seek to characterize stream temperatures, and predict their changes due to landscape and climatic changes, are extremely important. In this study, we extend a mechanistic stream temperature model within the Soil and Water Assessment Tool (SWAT) by explicitly incorporating radiative flux components to more realistically account for radiative heat exchange. The extended stream temperature model is particularly useful for simulating the impacts of landscape and land use change on stream temperatures using SWAT. The extended model is tested for the Marys River, a western tributary of the Willamette River in Oregon. The results are compared with observed stream temperatures, as well as previous model estimates (without radiative components), for different spatial locations within the Marys River watershed. The results show that the radiative stream temperature model is able to simulate increased stream temperatures in agricultural sub-basins compared with forested sub-basins, reflecting observed data. However, the effect is overestimated, and more noise is generated in the radiative model due to the inclusion of highly variable radiative forcing components. The model works at a daily time step, and further research should investigate modeling at hourly timesteps to further improve the temporal resolution of the model. In addition, other watersheds should be tested to improve and validate the model in different climates, landscapes, and land use regimes.