Ways of Using, and Correcting for Errors in Conventional Radar Reflectivity Data

Abstract

Over the years much research has been directed towards exploring the potential of radar as an instrument for estimating rain. It is shown that with the present reflectivity measuring radar, acceptable quantitative information is already obtained from radar networks in many places and in a wide range of applications. It is unlikely, however, that radar will ever replace the raingauge, since gauges are vital as ground truth for adjusting and/or checking the radar data. On the other hand, as pointed out by many workers, we would need an extremely dense and costly network of gauges to obtain a spatial resolution easily attainable with radar.

But radar will only estimate the precipitation it can “see”. Even the most ingenious procedure using sophisticated equipment will not allow us to make measurements in parts lost because of reduced visibility. Reduced visibility will occur behind mountains. But also at longer ranges, even in flat country, we will find increased errors due to losses caused by the earth curvature and the reduced resolution of the radar beam. Thanks to the availability of inexpensive, high-speed data processing equipment, it is possible today to determine the echo distribution in the whole radar coverage area in three dimensions. This knowledge, together with knowledge about the position of the radar and the orography around it, allows one to correct in real time for a large fraction of - or at least to estimate the magnitude of - the vertical profile problem. This correction allows us to extend the region in which an accuracy acceptable for many hydrological applications is obtained. This paper is in part based on the reviews by Joss and Waldvogel (1989) and Joss and Smith (1989), where the reader may find more details.