MODELING THERMAL EFFECTS OF CLIMATE CHANGE FROM LANDFORM PREDICTORS OF GROUNDWATER INFLUENCE IN CHESAPEAKE BAY HEADWATER STREAMS

Understanding groundwater contribution to stream temperature is critical for predicting effects of climate change on thermal habitat in streams. We evaluated landform features characterizing stream geomorphology, geological structure, and stream network topology to identify landscape attributes that predict groundwater influence on temperatures of headwater streams in the Chesapeake Bay Watershed, USA. We modeled groundwater influence at 81 sites within nine watersheds in Shenandoah National Park in 2012-2013 and evaluated landform predictors using linear modeling and machine-learning techniques. Results indicated that groundwater thermal influence was spatially structured within watersheds and that spatial variation in groundwater contribution is influenced by complex interactions among landscape attributes. Measures of stream network topology and geology provided the strongest predictors of groundwater influence. Spatially structured groundwater inputs influenced projected thermal habitat suitability for brook trout (Salvelinus fontinalis) in climate forecasting models (mean annual air temperature increases of 1.5, 3.0, and 5.0 degrees C). We discuss the importance of spatially structured groundwater influences on climate change in contrast to models assuming spatially uniform changes in stream temperature.