What problem does it solve? Running the MONICA process-based crop model correctly requires deep operational knowledge: converting climate and soil data into exact formats and units, configuring crop rotations, and diagnosing silent failures where the model runs but produces physically wrong results. This Skill packages that expertise so an AI agent can execute, validate, and debug MONICA simulations reliably. ## Core Features & Use Cases - End-to-end pipeline tools: Convert CMFD/MSWX/ERA5 climate forcing into MONICA climate.csv, transform HWSD/SoilGrids soil data into site.json profiles, execute the monica-run binary, and parse output CSVs into clean time series and metrics. - Diagnostic recovery: 32 documented symptom-to-remedy triplets (e.g., radiation unit traps inflating yield 10-100x, soil layer thickness in cm causing crashes) plus a preflight check that verifies the binary, parameters, and tools before any run. - Validation against literature: Ranked output variables (Yield is rank-1) with cited pass-bands for PBIAS, NMAE, and PMARE, plus a documented China winter-wheat multi-site workflow with photoperiod adaptation for non-European latitudes. - Use Case: Simulate winter wheat yield across Chinese provinces from 1991-2000 using CMFD forcing and HWSD soils, with automatic irrigation, then compare provincial PBIAS against published validation bands. ## Quick Start Run python preflight_check.py in this directory to verify the MONICA binary and parameters, then ask the agent to simulate winter wheat yield at your site by converting your climate and soil data with the provided tools.
