The sustainability of California agriculture depends on its ability to adapt to climate change and to mitigate it. Though field and laboratory experiments provide important data to this end, they require time and resources that limit the pace of progress. Coupling these data with modeling tools can accelerate agroecological research and highlight important directions for more comprehensive in vivo studies. Here, we use a combination of field data and modeling tools improve our understanding of stress tolerance and to rapidly assess the environmental impacts of various agricultural products. Chapter one evaluates water use dynamics in Sorghum bicolor cv. (Photoperiod LS) grown in California's Imperial Valley (IV) and the implications for stress tolerance. Coupling field data with an optimization model, we found sorghum to exhibit high heat tolerance at the expense of water use efficiency (WUE). However, the extent to which WUE was compromised varied in response to soil water status, suggesting sorghum adjusts its water use patterns depending on the type and severity of stress present. In chapter two, we assessed the environmental impacts associated with biofuel production using IV sorghum as a feedstock. We used field data to parameterize a crop model, allowing us to evaluate a wide range of nitrogen (N) management scenarios for feedstock production. Crop model output was coupled with life cycle assessment models to quantify the well-to-wheel environmental impacts of each production scenario. Overall, biofuels from IV sorghum had about 1/3 the global warming potential (GWP) of gasoline, but had much greater impacts to local air and water pollution. Efficient use of N was an important pathway for mitigating adverse impacts. Using a similar experimental framework, chapter three evaluates the GWP of California rice under a range of irrigation schedules varying in the severity and frequency of drainage events during the growing season. Severe and frequent dry-downs provided the greatest greenhouse gas mitigation potential, but showed potential to reduce grain yields, compromising the advantages of this practice from a global perspective. Our work highlights the diverse and valuable contributions of data-model fusion in preparing California for the conditions and demands of twenty-first century agriculture.