Modeling the Impacts of Climate Change on Cocoa

Understanding the future of cocoa production under climate change requires robust modeling approaches. This paper reviews the existing literature on cocoa suitability and impact assessment, focusing on both statistical and process-based models. While statistical models have been widely used to predict changes in cocoa-growing regions, process-based models offer a mechanistic understanding of crop-climate interactions. We highlight key findings from suitability studies and discuss the strengths and limitations of the very few process-based models such as JULES, CASEJ, and ALMANAC. To deepen this comparison, we offer an analysis replicating the approach of Asante et al. (2025) using ALMANAC, for the first time assessing whether and how results differ between process-based cocoa models. By synthesizing past research and conducting a targeted model comparison, this work aims to clarify gaps and future directions in cocoa-climate impact modeling. Our ALMANAC simulations produced plausible yield estimates without requiring post hoc scaling, compared to CASEJ, which heavily overestimated yields before applying yield gap adjustments. Both models predicted yield increases under climate change, but, in ALMANAC, only when including the CO₂ fertilization effect and improved management. The magnitude of the CO₂ effect also differed substantially between models. These findings underscore the need for further model development, long-term field validation, and careful interpretation of predictions that depend heavily on assumptions about CO₂ response, management intensity, or tree aging. We hope to see more attention and resources dedicated to developing process-based cocoa models to help narrow projection ranges and better understand future uncertainties.