Diagnosing Z-R relations using NWP

The choice of appropriate Z-R constants is of lesser importance compared to effects due to range dependency and the vertical reflectivity profile when measuring homogeneous precipitation, such as snow, with radar. Such conditions are common in winter over the Baltic Sea and its catchment area. Melting layers do occur, however, throughout the year although they are best developed during summer, where they can reach heights of over three km. Radar reflectivities from such bright bands can be more than 10 dBZ higher than reflectivities above and below them. In such cases, the radar will receive echoes from snow, rain, and mixed-phase precipitation, all within the same pulse volume, which makes the use of just one set of Z-R constants inappropriate for quantitative applications. A simple model has been developed for diagnosing precipitation phase, in order to dynamically select phase-dependent Z-R coefficients for application to radar data. An atmospheric model state is provided hourly by a mesoscale hydrostatic limited area model (HIRLAM); a 3-D wet-bulb temperature analysis is performed and a statistical relation between it and proportion of snow in precipitation is used to determine the altitude of the top of the bright band. Given a fixed bright band depth, and Z-R coefficients for rain, aggregate snow, and dry snow, dynamic Z-R relations are derived and applied according to a given location within a given radar measurement volume. Presence of hail is also diagnosed and Z-R relations for it applied. The model will be presented and results using it illustrated with data from the BALTEX Radar Data Centre. Benefits and shortcomings using this approach will be highlighted.