Separation and estimation of reflectivity errors according to distance and altitude using Specific Differential Phase

In order to estimate quantitative precipitation estimation (QPE) with high accuracy for flood forecasting, the quantitative uncertainty of heavy rainfall observations of radar must be identified in the spatiotemporal aspect. Considering the beam attenuation and the height of precipitation detected by radio waves, the accuracy of observations tends to be higher in short-range areas where the degree of beam attenuation is less and the observation height is low in order to estimate accurate precipitation used for ground flood forecasting. However, there have not been many cases where the error of precipitation estimation according to distance and altitude has been individually quantified and evaluated. Against this background, this study analyzed 22 major heavy rainfall events observed by five S-band dual-polarization radars in 2016 to quantify the reflectivity error according to observation distance and altitude, and derived the reflectivity error according to distance and altitude separately using Specific Differential Phase (Kdp). The analysis results showed that the average distance error of rainfall radar was approximately 10% or less up to 100 km and exceeded 30% above 150 km. The radar average elevation error was found to be approximately 10% or less for the second elevation angle from the ground among the six operating elevation angles, 20% for the third and above, and over 50% for the fourth and above. And the changes in observation accuracy during the heavy rainfall according to the observation range of 300km and 150km were compared through experiments. The experimental results showed that the cumulative reflectivity of the 150km observation was large when the distance from the radar was less than 75km, and the cumulative reflectivity of the 150km observation was large when the distance was more than 75km. This study is expected to contribute to establishing an appropriate rainfall radar observation strategy when operating a rainfall radar for the purpose of accurate quantitative rainfall observation for flood forecasting.

 

Acknowledgments

This research was supported by a grant(2022-MOIS61-003(RS-2022-ND634022)) of Development Risk Prediction Technology of Storm and Flood for Climate Change based on Artificial Intelligence funded by Ministry of Interior and Safety(MOIS, Korea).