Use of a dual Kalman filter for real-time correction of mean field bias of radar rain rate

- Use of a dual Kalman filter for the prediction of the G/R ratio.
- Comparison of prediction accuracy of the Kalman filter and dual Kalman filter.
- Superior prediction ability of the dual Kalman filter for short-time ranges and complex storm events.

SummaryThis study applied a dual Kalman filter (DKF) for real-time correction of the mean-field bias of radar rain rate. The DKF is a dual estimation system, and is different from the conventional Kalman filter (KF). The DKF is composed of a state estimation system and a model estimation system, thus it operates two KFs. The state estimate system is the same as the conventional KF, but the model estimation system continuously updates the model parameters. The DKF was applied to four radar stations in Korea; the Kwanaksan Radar, Osungsan Radar, Jindo Radar and Kudeoksan Radar. As a result, it was found that the DKF can be superior to the KF application when the temporal variability of the G/R (rain gauge rain rate/radar rain rate) ratio is very high. Additionally, the application of the DKF for a short-duration severe storm event should be emphasized. In particular, for a flash flood warning system, the DKF application can have a special meaning; that of improving the quality of the radar rain rate data.