Static and dynamic calibration of radar data for hydrological use

Abstract

The HYREX dense raingauge network over the Brue catchment in Somerset, England is used to explore the accuracy of calibrated (raingauge-adjusted) weather radar data. Calibration is restricted to the use of any single gauge within the catchment so as to simulate the conditions in a typical rainfall monitoring network. Combination of a single gauge and a radar estimate is used to obtain calibrated radar estimates, with the "calibration factor" varying dynamically from one time-frame to the next. Comparing this dynamic calibration with a static (long-term) calibration factor indicates the distance from a gauge over which the dynamic calibration is useful. A tapered calibration factor is implemented which behaves in the same way as the raw dynamic calibration at short distances, tending towards the static calibration factor at larger distances. This hybrid approach outperforms raingauge, uncalibrated radar, and statically-calibrated radar estimates of rainfall for the majority of raingauges in the catchment. The results provide valuable guidance on the density of raingauge network to employ in combination with a weather radar for flood estimation and forecasting.