Applying Data Assimilation to Three-Dimensional Precipitation Nowcasting with Phased-Array Weather Radar

The phased-array weather radar (PAWR) is a new generation rapid-scanning weather radar. The PAWR at Osaka University can perform three-dimensional volume scans at about 100 elevation angles within 60 km every 30 seconds. To take advantage of dense and frequent observations, Otsuka et al. (2016b) performed three-dimensional precipitation nowcasting experiments with space-time extrapolation of PAWR data, and demonstrated that the three-dimensional space-time extrapolation outperformed the conventional two-dimensional space-time extrapolation.

However, erroneous motion vectors sometimes degrade the forecast performance. Here, data assimilation is employed to improve the motion vector field. Otsuka et al. (2016a) implemented the Local Ensemble Transform Kalman Filter (LETKF) with the two-dimensional space-time extrapolation system, and the system was successfully applied to the Global Satellite Mapping of Precipitation (GSMaP) data. In this study, the nowcasting system with LETKF is extended to the three-dimensional motion vector field, and applied to the PAWR precipitation nowcasting. A case study on an isolated convective system showed that the three-dimensional space-time extrapolation with LETKF outperformed that without data assimilation in terms of precipitation threat scores.

References:

Otsuka, S., S. Kotsuki, and T. Miyoshi, 2016a: Wea. Forecasting, in print, doi:10.1175/WAF-D-16-0039.1.

Otsuka, S., G. Tuerhong, R. Kikuchi, Y. Kitano, Y. Taniguchi, J. Ruiz, S. Satoh, T. Ushio, and T. Miyoshi, 2016b: Wea. Forecasting, 31, 329-340, doi:10.1175/WAF-D-15-0063.1.