Climate change is advancing at an unprecedented rate, creating a need for rapid adaptation in sectors such as agriculture, which are potentially subject to disastrous impacts. Successful adaptation, which can help ensure a stable food supply for a growing population, will require the ability to reliably assess which agricultural practices are most effective in dealing with adverse weather. To make these assessments, researchers can currently turn either to datasets from randomized field experiments, which can be small and limited in scope, or to large observational datasets from satellites and other sources, which can lead to conclusions biased by confounding variables. In my research, I seek to develop new statistical methods for combining insights from both types of datasets. Using both experimental and observational datasets in the same statistical analysis would lead to better, more reliable estimates of the effectiveness of certain agricultural practices in dealing with adverse weather.