topoflow-ki

Run preflight_check.py first to verify the environment, then prepare forcing with convert_forcing.py, soil parameters with convert_soil_params.py, and execute via run_topoflow.py or python -m topoflow with a cfg_prefix and cfg_directory. Success ends with a Finished message and 0D-Q output files.

Use tools/convert_forcing.py with --source era5, providing the NetCDF directory, target lat/lon, and date range. It converts precipitation from kg/m²/s to mm/hr, temperature from Kelvin to °C, and writes text time-series files the meteorology component reads.

Precipitation must be mm/hr in the .cfg file; TopoFlow converts to m/s internally, so never pre-convert. Saturated hydraulic conductivity Ks must be m/s, requiring a ÷3.6e6 conversion from the mm/hr values common in soil databases.

Silent unit errors are the usual cause: precipitation left in mm/day causes 24x overestimation, while pre-converting to m/s produces near-zero rain. Check diagnostics/triplets.yaml for the matching symptom and its validated remedy before changing anything else.

Validation targets the rank-1 output outlet_discharge (Q_outlet) against observed streamflow using NSE and PBIAS. Cited bands mark NSE ≥ 0.5 satisfactory, ≥ 0.65 good, and ≥ 0.75 very good, with PBIAS within 10-25 of zero.

Infiltration options are Green-Ampt, Smith-Parlange, Richards 1-D, and Beven. Channel routing supports kinematic, diffusive, and dynamic wave closures with Manning or Law-of-Wall friction, selected via the case-sensitive provider file.