THE APPLICATION OF A CONTINENTAL-SCALE RADAR DATABASE TO HYDROLOGICAL PROCESS PARAMETRIZATION WITHIN ATMOSPHERIC GENERAL-CIRCULATION MODELS

Atmospheric General Circulation Models (GCMs) used for climate research operate on a resolution of about 250 km and provide outputs averaged on a daily and monthly basis. The formation and evaporation of precipitation within cloud systems produces large latent heating effects. These effects constitute an important source of uncertainty in climate and, to a lesser extent, weather-forecasting models. There remains a large impact from this uncertainty on the performance of these models, produced by inadequate parametrization on scales of around 100-300 km. Likewise, variations of soil properties and of infiltration are not adequately defined at these scales, and therefore it is necessary to parametrize these processes. Unfortunately, the relationship between rainfall and runoff is highly non-linear, and the rate of infiltration will vary throughout a catchment area, making parametrization very difficult. Since 1986 the Commission of the European Communities (CEC) COST-73 Weather Radar Networking Project has encouraged the exchange of weather radar data between the countries of Europe. As part of the work undertaken in this project, instantaneous radar composite images over NW Europe were generated every hour with a grid resolution of about 5 km. Data from radars located in the UK, Ireland, France, The Netherlands, Switzerland, and Belgium have been used. In the paper we discuss the nature of this database, and consider its reliability for estimating daily and monthly precipitation. We explore how this database might be used to develop a more effective parametrization scheme for surface hydrological processes within GCMs. Finally, the adequacy of such a database for continental-scale hydrological studies is assessed. It is noted that there is considerable scope for further studies of this type using radar data from most countries in both Western and Eastern Europe.