An AI Approach to Integrating Climate, Hydrology, and Agriculture Models

Abstract

Understanding the interactions between natural processes and human activities poses major challenges as it requires the integration of models and data across disparate disciplines. It typically takes many months and even years create valid end-to-end simulations as the different models need to be configured in consistent ways so their results can be meaningfully interpreted. MINT is a novel framework that uses AI for model integration. MINT captures extensive knowledge about models and data, including their requirements and constraints. MINT guides a user to pose a well-formed modeling question, select and configure models, find appropriate datasets, set up scenarios and parameters, run the simulations, and visualize the results. MINT currently includes climate, hydrology, and agriculture models for different areas of Ethiopia, Kenya, and South Sudan. Our goal is to understand droughts through integrating meteorological, hydrological, and agricultural analyses.