Short-Term Predictability of Extreme Rainfall Using Dual-Polarization Radar Measurements

抄録

Dual-polarization radar often detects columnar regions of enhanced differential reflectivity (Z_DR) extending vertically above the environmental 0 °C level. Indicative of supercooled liquid drops and wet ice particles lofted by strong updrafts, these Z_DR columns are increasingly understood to be of use in predicting extreme rainfall. With the aim of achieving practical application of Z_DR column measurements, this paper focuses on the relationship between the height of Z_DR columns and rainfall intensity near the ground.

All the data on Z_DR columns analyzed in this study was collected from weather radar stations in Japan. The height of each column and rainfall rates at low levels were analyzed using an automated algorithm. A regression analysis result reveals peak column height to be positively correlated with maximum rainfall rate near ground level, and that rainfall intensity on the ground is likely to exceed 50 mm h⁻¹ when radar identifies a Z_DR column. Furthermore, extreme rainfall with an intensity of 180 mm h⁻¹ or more is likely associated with a column over 3 km tall from the 0 °C level. These findings suggest that surveillance of Z_DR columns can contribute to the reliability of very short-range forecasts or nowcasts as well as assist with the issue of early warnings of extreme rainfall and flash floods.