Understanding COVID-19 spatiotemporal dynamics requires a detailed understanding of population interaction patterns, heterogeneities in risk, and temporal changes in interventions and behaviors. To better understand what has happened with COVID-19 in Florida, and to predict future dynamics and intervention impact, we have developed a spatially-explicit synthetic population representing the state’s population, including 21m people residing in 11m households and 3.8k long term care facilities, who go to work in 2.3m workplaces and attend 7.6k schools. Individuals in the model have attributes including age, sex, and presence of comorbidities. This approach requires sourcing, reconciling, and integrating datasets from the U.S. Census (household composition), National Corporation Directory (workplace type and location), CDC (comorbidities), and SafeGraph (behavioral changes), among others. We then use this synthetic population as an input to the agent-based model we developed of SARS-CoV-2 transmission and COVID-19 detection and reporting. Using this model, we have predicted the impact of a range of interventions the state of Florida might have employed thus far, as well as the impact of a range of vaccination scenarios that may become available in the coming months.