glm-evaluator

name	glm-evaluator
description	Evaluate GLM simulation results using field observations and RMSE metrics. Use this skill when you need to merge simulation results with field observations and calculate RMSE values for different conditions (overall, annual_deep, summer_deep).

GLM Evaluator Skill

Data Requirements

• Observations: field_temp_oxy.csv (contains depth, temp, oxygen, and datetime).
• Simulation Results: output/output.nc (contains water temperature profiles).

Evaluation Methodology

1. Preprocessing Observations:
   • Parse datetime.
   • Extract depth and temp.
2. Preprocessing Simulation:
   • Extract temp profiles and time.
   • Map simulation depths to observation depths.
3. Merging:
   • Perform an exact datetime + rounded-depth merge.
   • Do not use nearest-time matching or interpolation.
4. Metric Definitions:
   • overall_rmse: RMSE of all matched pairs.
   • annual_deep_rmse: RMSE for depths >= 13 m.
   • summer_deep_rmse: RMSE for depths >= 13 m and months June to September.
5. Reporting:
   • Save results to metrics.json with keys: overall_rmse, annual_deep_rmse, summer_deep_rmse, overall_n_pairs, annual_deep_n_pairs, and summer_deep_n_pairs.

Python Libraries

Use pandas for data manipulation, netCDF4 or xarray to read .nc files, and numpy for calculations.