SEARCH

SEARCH BY CITATION

Keywords:

  • paramterization;
  • MPR;
  • HRU;
  • ungauged basin;
  • scale;
  • mHM

[1] Distributed hydrologic models are increasingly used to describe the spatiotemporal dynamics of water fluxes and state variables occurring within a basin. Reliable model simulations at multiple scales and locations, however, can be obtained only if a robust parameterization method is employed. In this study, we evaluated the performance of three parameterization methods: two based on the hydrological response units (HRUs) and one on the multiscale parameter regionalization (MPR) technique. They were implemented into the distributed mesoscale hydrologic model and then applied to 45 southern German basins. A set of numerical experiments were conducted to assess the effectiveness of the transferability of free parameters to scales and locations other than those used during calibration. The performance of all three methods was comparable for daily discharge simulations when their free parameters were calibrated independently in each basin at a given scale. A significant deterioration in performance of HRU was, however, noticed when free parameters calibrated at coarser scales were shifted to finer ones (up to 60%), whereas MPR exhibited quasi scale-invariant performance with losses in modeling efficiency of less than 2%. Moreover, the latter preserves the spatiotemporal pattern of the simulated variables at a given scale, when transferred from other scales. Evaluation of the transferability of free parameters to ungauged locations further indicated the higher effectiveness and reliability of MPR compared with those of HRU. This study emphasizes the importance of a robust parameterization method in distributed hydrologic modeling.