COMPARISON OF MODEL RESULTS OBTAINED WITH SEVERAL EUROPEAN REGIONAL AIR-QUALITY MODELS

(A)n intercomparison study has been performed with four photo-oxidant dispersion models (EMEP, EURAD, LOTOS and REM3) which are currently ca pable of performing photo-oxidant formation calculations over larger p arts of Europe. The models, in principle, were run in the mode in whic h they are normally used, with their own input data and process descri ptions. No attempt has been-made to harmonize the input data. The stud y focused on the simulation of a six day episode (31 July-5 August, 19 90) with relatively high observed ozone levels over northwest and cent ral Europe(upto 140 ppbv) and which ended by a frontal passage. The st udy was mainly restricted to an analysis of the behaviour at four rece ptor sites across the centre of each modelling domain. Differences in yearly averaged anthropogenic emissions are in the order of 10% only. However, episode specific hourly emissions, both anthropogenic and bio genic, show much larger differences. Here up to 40% differences for an thropogenic emissions are found for central Europe while biogenic emis sions differ even up to a factor 2-3. The meteorological data base for all models is created from the synoptic measurement network by differ ent techniques. This can result in quite different model inputs as e.g . the value of the mixing height or trajectories. Other model inputs, e.g. dry deposition velocities or photolysis frequencies, show much cl oser agreements than expected from their derivation through different parameterizations. Model performance for the regional scale could only be observed for ozone. In central Europe, the large-scale ozone patte rns are quite similar for all models. Here the temporal evolution at s ome receptor sites shows also a reasonable agreement with observations . However, in case a multi-component evaluation is possible it can be expected that differences between models can be revealed to a greater detail than was possible here. The model responses to emission reducti ons show substantial differences although the direction of the respons es generally agree. The most complex model (EURAD) has the most pronou nced response to anthropogenic emisssion reductions. The treatment of the biogenic emissions can have a large impact on scenario simulations . (C) 1997 Elsevier Science Ltd.