What problem does it solve? Running the USGS TRIGRS Fortran model for rainfall-induced shallow landslide analysis requires manual GIS grid preparation, error-prone unit conversions, and command-line execution where silent unit mistakes produce plausible but wrong factor-of-safety results. ## Core Features & Use Cases - Validated Python pipeline: Five tools convert rainfall forcing and soil properties into TRIGRS inputs, generate the tr_in.txt initialization file, run the compiled trg binary with TopoIndex, and parse factor-of-safety grids to CSV. - Unit-trap defense: Explicit conversion tables and physical-bounds validation catch the silent 1000x and 3.6-million-fold unit errors (Pa vs kPa, m/s vs mm/hr) that dominate TRIGRS failures. - Diagnostic recovery: Eighteen symptom-diagnosis-remedy triplets map known failure modes to fixes, plus a preflight check that verifies the binary, environment, and data before execution. - Use Case: Given a DEM, slope grid, soil texture table, and a rainfall time series for a 48-hour storm, generate all TRIGRS inputs, run the model, and obtain a minimum factor-of-safety grid identifying cells where Fs drops below 1. ## Quick Start Run the preflight check in this directory, then ask the agent to prepare TRIGRS inputs from your DEM, soil, and rainfall data and execute the model to produce the minimum factor-of-safety grid.
