gent-based modeling – in which system-level outcomes emerge from interactions among individuals and their environment – can provide unique insights in the study of livelihoods decision-making and adaptation responses to climate hazards and shocks. However, finding empirical support for the underlying behavioral assumptions in ABM can be a challenge. Data on decision-making is typically expensive to collect, or in some cases simply not possible – it can be hard to report on things (such as sea level rise, or other compound floods) that one hasn’t experienced yet. In this talk I present results and work in progress on a stakeholder-engaged approach to modeling environmental forcings on migration (applications to Bangladesh and Senegal), with a focus on data gaps in adaptation modeling and the frontier of opportunities (expert elicitation, through participatory methods and high-frequency data collection) to resolve parts of them.