Model Calibration

Note

This section is under development.

Overview

Model calibration is the process of adjusting model parameters to minimize the discrepancy between simulated and observed results. For FLUXOS, the primary calibration parameter is the roughness height (Manning’s roughness), which controls the friction forces in the shallow water equations.

Calibration Parameters

The following parameters can be adjusted during calibration:

Parameter	Config Key	Description
Roughness height	ROUGNESS_HEIGHT	Manning’s roughness coefficient controlling bed friction [m]
Minimum water depth	H_MIN_TO_PRINT	Threshold for wet/dry classification and output filtering [m]

Calibration Workflow

A typical calibration workflow for FLUXOS involves:

1. Define objective function: Select metrics to compare simulated vs. observed data (e.g., Nash-Sutcliffe Efficiency, RMSE, KGE)

2. Select observation data: Identify available field measurements (water levels, discharge at cross-sections, inundation extent)

3. Parameter sampling: Define parameter ranges and sampling strategy (manual, grid search, or automated optimization)

4. Batch simulation: Run FLUXOS with different parameter sets

5. Post-processing: Extract simulated values at observation locations from the .vtu / .txt output (see supporting_scripts/2_Read_Outputs/ for the template-driven flood-statistics report; supporting readers in output_supporting_lib/)

6. Evaluation: Compare simulated vs. observed data using the objective function

7. Iteration: Refine parameter ranges and repeat