Short term forecasting for precipitation by utilizing a new three-dimensional super-rapid phased array weather radar data

GULAMBAER TUERHONG; SHIGENORI OTSUKA; JUAN RUIZ; R. KIKUCHI; Y. KITANO; Y. TANIGUCHI; TAKEMASA MIYOSHI

Kobe

Simposio; The Fifth AICS International Symposium; 2014

RIKEN Advanced Center for Computational Science

Short-term precipitation forecasting (a.k.a. nowcasting) is important for navigating meteorological disasters such as typhoons and severe storms. Conventional nowcasting approaches are based on two-dimensional five-minute-scan radar data which may limit their forecasting accuracies due to the difficulties in capturing a rapid development of convections. In this study, a 3-D precipitation extrapolation system is developed using the latest phased array weather radar (PAWR) which is capable of a 100-m resolution 3-D volume scan every 30 seconds. First, motion vectors are estimated based on the COTREC algorithm using two consecutive radar scans. Second, the precipitation field is extrapolated with the motion vectors based on a weighted essentially non-oscillatory (WENO) scheme. The extrapolation system showed improved forecasting results with the super rapid 3-D PAWR data compared to its 2D counterparts