SEALS Home

SEALS, the Spatial Economic Allocation Landscape Simulator, is a land-use change model that downscales projections of land-use change from aggregate (regional or coarse-gridded) inputs to a much finer resolution, typically 10–300 meters.

Coarse economic and integrated-assessment models tell you how much cropland, pasture or forest a region gains or loses. They cannot tell you where. SEALS answers that question, which is what makes it possible to run spatially explicit ecosystem service models on economic scenarios.

What makes it distinctive

  • Speed. SEALS downscales to 300 meters globally — roughly 8.4 billion grid cells — in about an hour on a laptop. Performance-critical functions are written in C++ via Cython; everything else is Python.

  • Scalability. It is designed for parallel execution, so it scales from a laptop to high-performance, cloud, or distributed computing.

  • Source-agnostic inputs. It downscales from coarse gridded projections (LUH2, MAgPIE) or from regional totals (a CGE model such as GTAP-AEZ), or from both at once.

  • Reproducibility. Runs are driven by a scenarios CSV rather than by edited code, and every step is cached, so an interrupted run resumes instead of restarting.

SEALS is a part of the Earth-Economy Devstack, a software platform supported by NatCap TEEMs (The Earth-Economy Modellers).

Where it fits

SEALS is the land-use step inside GTAP-InVEST: GTAP-AEZ produces regional land-use change, SEALS downscales it to 300 m, and InVEST computes the resulting ecosystem service changes. It is equally usable on its own, or coupled to any other model that produces coarse land-use projections.

It was introduced alongside GTAP-InVEST in Johnson et al. (2023), PNAS.

Documentation and code

SEALS is under active development and changes rapidly. Pull the latest seals_dev before reporting a problem.