With over 55% forest cover, the southern Appalachians (SA) are a main water resource for the surrounding areas. These water resources are at risk due to changing climate and precipitation regimes as well as changes in forest cover. Understanding the implications of these risks will help to develop management strategies for an increasingly valuable resource. Evapotranspiration (ET), the combination of plant transpiration and surface evaporation, can vary across space and time, and is a significant component of the hydrological cycle in densely forested regions. Quantifying ET is critical to understanding the available water resource, especially in the SA. In the SA, ET averages 50% of annual precipitation in forested watersheds and can climb to 85%. However, ET is among the most difficult and complex component of the water cycle to measure and model. This dissertation addresses these complexities by investigating the ability of sap flow models to estimate ET and examining the impact of potential temperature and compositional shifts on water yield. We also examined sap flow input variables to determine the best methods for the SA including the spatial estimation of climatological variables, phenological dates, and leaf area index (LAI) estimates all of which would particularly enhance the development of our hydrological models.