Automatic Calibration of a Physically-based Distributed Hydrologic Model  Fulltext 

[ Reza Entezarolmahdi ] - Ph.D.
[ Riccardo Rigon ] - Associate professor
[ Giacomo Bertoldi ] - Post-doc fellowship

خلاصه مقاله:

The identification problem entails estimation of system parameters. The latter essentially becomes a problem of optimization, wherein some prior knowledge of system would be useful. In fact, the general problem of model identification involves the selection of appropriate structures for the various model components and specification of values for the parameters of those model components, such that the resulting model provides a sufficiently accurate simulation and/or prediction of watershed response. The present study deals with the estimation of four important arameters of the physicallybased distributed hydrologic model GEOtop 0.75, using the automatic optimization algorithm SCEM-UA, and applied to the basin Mandola, Caldonazzo, Trento, Italy. The process by which parameters are estimated (using a finite set of observations on input and output) has come to be called model calibration. Using automatic optimization approach, the model parameter calibration has been done so as to find the most probable parameter set. In order to assess whether the calibrated model can be considered valid for subsequent use, it was validated against data different from those used for calibration.

Physically-based hydrologic model, Automatic calibration, GEOtop 0.75, SCEM-UA method, Optimization.