Multi-Criterial Calibration

Pareto-optimal calibration for multiple hydrological signatures and spatio-temporal patterns
  • Project P3 aims at improving the simulation of continental water cycle dynamics by a multi-objective calibration scheme for WaterGAP following the principle of Pareto optimality.
  • The Pareto-optimal calibration (POC) approach should consider more than two objective functions for calibration, considering the various observational data sets developed in this Research Unit.
  • Adequate hydrological signatures and quantitative performance criteria to consider the dominant spatial patterns of total water storage variations shall be developed and used for calibration.
  • Adequate or deficient WaterGAP model structures and model uncertainties can be explored by the trade-offs among different objective functions within the POC framework.

 

Example of a Pareto plot for the calibration of WaterGAP with four observables (river discharge, total water storage, surface water storage, evapotranspiration) Seasonal dynamics of total water storage, simulated with different WaterGAP runs that were calibrated against different observables