What problem does it solve? Running the GLM (General Lake Model) for a lake or reservoir requires scattered operational expertise: converting meteorological forcing into GLM's strict CSV units (m/day rain, percent humidity), building depth-area morphometry, writing Fortran namelists, and diagnosing silent failures like a disabled ice model. This Skill packages that expertise into a validated 11-stage Python pipeline so an agent can run GLM v3.3.3 with AED2 water quality on any lake without manual data preparation. ## Core Features & Use Cases - End-to-end pipeline: 17 tools covering lake identification in HydroLAKES, morphometry construction, CMFD/MSWX/VIC forcing conversion, inflow/outflow setup, namelist generation, AED2 water quality configuration, execution, output parsing, calibration, and coupling to CaMa-Flood. - Diagnostic recovery: 30+ diagnostic triplets map symptoms to root causes and remedies, including silent unit errors (rain in mm/day vs m/day) and the undocumented dt_iceon_avg parameter required for ice simulation. - Validated results: Production-validated on Miyun Reservoir (2001-2010) and Lake Catoma with documented NSE, RMSE, and KGE metrics against published and observed data. - Use Case: Ask the agent to simulate thermal stratification and ice cover for a reservoir using CMFD forcing; it builds morphometry, converts forcing, generates glm3.nml, runs the GLM binary, and returns temperature heatmaps and validation metrics. ## Quick Start Run python preflight_check.py in this directory, then ask the agent to simulate your target lake by providing its name or coordinates, simulation period, and forcing dataset.
