Benefits and problems of integrated operational use of radar and other observational data in thunderstorm situations

The operational suite of the modern nowcaster has changed dramatically in the last 10 years. Besides conventional data diverse remote-sensing observations provide a lot of real time information on actual weather. These data are displayed on advanced presentation systems. Especially thunderstorm situations can be closely monitored by different remote-sensing data, highly resolved in space and time. But the potential available to the forecaster represents not only benefits but also problems while fitting the data together and interpreting them.

Remote-sensing systems detect specific characteristics of scale, structures and development of thunderstorms. Satellite data cover large-scale and mesoscale features of thunder-storms in terms of clouds and cloud parameters. Geostationary satellites have special preference on a good time-scale, polar-orbiting satellites have advantages in horizontal resolution. Radar data are playing the main role in monitoring thunderstorms in that they are sensing from ground the key parameter precipitation. Radar data display the internal structures of thunderstorm reflectivity and wind-field. Spatial resolu-tion of radar data is as a rule 1 km, time intervals are 5-15 minutes. In radar networks radar data are operationally combined to regional composites. Lightning data are the only remote sensing data evidencing directly thunder-storms. They are best resolved in space and time with horizontal resolu-tion down to some hundred meters and operational time intervals of some minutes.

To interpret all this information together with conventional data the nowcaster can make use of conceptual models of thunderstorms and thunderstorm systems comprising typical structures, life-cycles and related weather phenomena. He/she is supported by NWP-products, special tools and in some services by automated nowcasting systems. To interactively investigate the actual weather situation the presenta-tion systems offer procedures like overlaying of different data and temporal anima-tions (loops). Loops may show single data with best possible temporal resolu-tion or a combina-tion of different data types with an appropriate selection of time-interval. In this way the forecaster gets insight into all aspects of thunderstorm structures and development.

Precise application of conceptual knowledge of thunderstorms requires accurate mutual spacing and timing of the data displayed on the screen. This is often not the case. The real data time is differing from the nominal data time in synoptic, satellite and radar data. You have to take into account scanning length and operational regulations for nominal time allocation. This may lead to time discrepancies of more than 1/4h, critical in situations of rapid thunderstorm development. Further problems exist with spatial allocation. Difficulties in exact navigation of satellite images result in horizontal displace-ments of sometimes 20 or more kilometers. A similar shift occurs by the parallax of high clouds in satellite data, not corrected in operational images. Also radar data can reveal systema-tic problems in exact spacing of structures. Therefore satellite, radar and lightning data don't fit always precisely together. With further progress in conceptual applications these problems have to be tackled more deeply.