Presenter/Author Information

David Maréchal
Ian P. Holman

Keywords

rainfall-runoff, assessment, ungauged, catchment scale

Start Date

1-7-2004 12:00 AM

Abstract

The objective of this study is to assess the performance of a regional hydrological model in catchments treated as ungauged. The Catchment Resources and Soil Hydrology (CRASH) model is a daily, catchment-scale, rainfall-runoff model that has been previously regionalised for England and Wales. In this paper, the regional CRASH is evaluated in three catchments located in East Anglia – eastern England - and it is compared to the catchment-specifically calibrated CRASH. The results demonstrate that the performance criteria are met in the three catchments for both the Nash-Sutcliffe (R2) and the percent bias efficiency indexes. The R2 results of the regional CRASH in the three catchments (0.70, 0.56 and 0.48) compare well with another study in one of the catchments using another hydrological model specifically calibrated and are within the range of results from other simulation studies in ungauged catchments in England, Australia, Canada and Norway. The degradation between the regional and the catchment specific models is only limited for all the efficiency indexes. Finally, the uncertainty analysis on the model parameters showed that there is a reasonable confidence in the regional model.

COinS
 
Jul 1st, 12:00 AM

Comparison of Hydrologic Simulations using Regionalised and Catchment-Calibrated Parameter Sets for three Catchments in England

The objective of this study is to assess the performance of a regional hydrological model in catchments treated as ungauged. The Catchment Resources and Soil Hydrology (CRASH) model is a daily, catchment-scale, rainfall-runoff model that has been previously regionalised for England and Wales. In this paper, the regional CRASH is evaluated in three catchments located in East Anglia – eastern England - and it is compared to the catchment-specifically calibrated CRASH. The results demonstrate that the performance criteria are met in the three catchments for both the Nash-Sutcliffe (R2) and the percent bias efficiency indexes. The R2 results of the regional CRASH in the three catchments (0.70, 0.56 and 0.48) compare well with another study in one of the catchments using another hydrological model specifically calibrated and are within the range of results from other simulation studies in ungauged catchments in England, Australia, Canada and Norway. The degradation between the regional and the catchment specific models is only limited for all the efficiency indexes. Finally, the uncertainty analysis on the model parameters showed that there is a reasonable confidence in the regional model.