Compilation of the MOPEX 2004 results

As part of the MOPEX 2004 workshop, the participants were asked to submit simulations using the common database provided for the workshop (see Chahinian et al this issue). The simulations were then analysed and the evaluation criteria computed to compare the models' performance for the gauged and ungauged modes using six criteria describing the model's performance in both high and low flow conditions. The comparisons were undertaken for all three participation levels (i.e. 3, 12 and 40 catchments). The results indicate that on the 3 catchment level model ranking may vary according to the tested criterion and catchment. Hence a larger number of catchments are necessary to evaluate the models' performance. Among the 10 models tested on the 12 and 40 catchment samples in gauged mode, GR5H, Mordor and SAC-SMA rank as the top three models. When analysing model results in the ungauged mode, SAC-SMA ranks as the best of the four tested models. The analysis of the submitted files highlights the need for continuing efforts to develop model parameterisation strategies for ungagued catchments in order to improve prediction in ungauged basins (PUB).