Healthy land use choices depend on our ability to understand how soils, weather, management history, and future practices interact over time. But turning that complexity into something useful for planners, land managers, researchers, and producers requires more than a model — it requires an infrastructure that can connect science, spatial data, and real-world decision-making.

This presentation shares the development of a cloud-based modeling backend designed to support regenerative land-use planning through scalable DayCent biogeochemical modeling. The system brings together SSURGO soils, ERA5 weather, user-drawn polygons, management histories, treatment scenarios, and calibrated model parameters, then transforms them into the inputs needed to simulate soil carbon and greenhouse gas exchange.

Behind a single user scenario, the system may need to coordinate dozens or hundreds of model runs across soils, weather grids, historical land-use phases, and future management options. To make this possible, the architecture uses Postgres/PostGIS, Docker, AWS Lambda, S3, and automated orchestration to prepare inputs, run the model in parallel, manage outputs, and synthesize results.

Rather than focusing only on infrastructure for its own sake, this session highlights how geospatial data engineering and serverless cloud design can help make complex environmental modeling more accessible, responsive, and actionable. Attendees will see how spatial data pipelines, legacy scientific models, and modern cloud architecture can work together to support healthier soils, more resilient landscapes, and more informed transitions toward climate-conscious land management.