Introduction to Calibration Plots

Model calibration is important to ensure that model predictions represent the actual hydraulic and water quality conditions in the system. An attempt to calibrate a model should always be made when model results are used in decisions concerning possible remedial actions, augmentation works, forecasting etc.

Calibration is primarily focused on reproducing the observed hydraulics and water quality behaviour of the system in terms of flow depth/pressure, flow discharges, and velocities. The model calibration may include comparisons between model simulation results and field measurements for, but not limited to, the following data:

• Flow
• Water level / Pressure / Hydraulic head
• Velocity
• Water mass balance
• Contaminant concentrations
• Contaminant migration rates
• Degradation rates

These comparisons may be presented as maps, tables, or graphs. The calibration results need to be evaluated by the modeller using engineering professional judgement. There are no universally accepted "goodness-of-fit" criteria that can be applied in all cases. However it is important that the user makes every attempt to minimize the difference between model simulations and measured field data.

The model calibration tool in MIKE+ provides comparisons of measured and computed time series, scatter plot comparisons as well as a number of numerical calibration criteria, all provided at point locations were measurements are available.