Empirical Stream Thermal Sensitivities May Underestimate Stream Temperature Response to Climate Warming

Stream temperature has been increasing in tandem with air temperature, with potentially negative impacts on cold-water fish such as salmon. Assessing future stream temperature change is critical for developing effective management responses. Empirical models of stream thermal sensitivity generally predict less future warming compared to physically based models. Here we reconcile these discrepancies by using a process-based hydrology and temperature model to simulate daily flow and water temperature for forested headwater catchments in a maritime region under both historic and projected future climatic conditions. The primary reason that the empirical approach underestimates thermal response to climate change is that it does not account for thermal memory in the catchment, especially related to the effect of snow cover. Empirical thermal sensitivities thus may underestimate stream temperature response to future climate warming. In addition, groundwater-fed streams may only resist warming in the short-medium term, due to lagged response of groundwater temperature. More process-based understanding and modeling of stream thermal regimes is needed to effectively manage aquatic ecosystems.