In this era of global environmental change, forests and the ecosystem services that they provide are changing. Particularly in semi-arid regions, warmer temperatures and more frequent and extensive droughts have led to declines in forest productivity, increased mortality and fire severity. Forest responses, however, are remarkably heterogeneous, even within the same watershed. To help explain this varied forest response and to support the management of this resource, we need better integration of the best-available science tools. An effective integration strategy is to synthesize and extend field-observations and experiments through process-based models. Here, we illustrate how this model-data integration can lead to insights into climate responses of mixed versus homogenous forest stands and stands of differing densities. We do this by combining field-based observations of forest water use with RHESSys (Regional Eco-hydrologic simulation system) for case studies in Israel and California. Our work highlights the importance of field study for identifying forest mechanisms that can lead to resilience or vulnerability to drought and the use of models to combine and extend what we learn from these studies, and finally, the potential for model visualization to make this knowledge more accessible.