Vertical Motion and Growth of Precipitation Measured by Dual Phased Array Weather Radar Every 30 Seconds

The first X-band phased array weather radar (PAWR) developed in Japan was installed at Osaka University, Suita in May 2012. The second and third PAWRs were subsequently installed at NICT advanced ICT Research Institute, Kobe, and NICT Okinawa electromagnetic technology center, Onna in March 2014, respectively. In the common observation area between Suita PAWR and Kobe PAWR (Fig. 1), we can get dual-Doppler measurement in a three-dimensional (3D) space every 30 seconds. In this study, the vertical motion and growth of a precipitation core in a cumulonimbus cloud are investigated using both 3D radar echo images every 30 seconds and 3D wind vectors derived from dual-Doppler analysis. The PAWR is one-dimensional phased-array radar system with an electronic scan in an elevation direction and a mechanical rotation in an azimuth direction. To reduce 3D measurement time, we use fan-beam transmitting and Digital Beam Forming (DBF) receiving using 128 antenna elements. Although it is possible to select several observation sequences, we usually operate a dense 3D observation with about 100 elevation angles from 0 to 90 degree in 30 seconds. The 100 m range resolution data covers over the observation area of 60 km radius. The observation data are translated into a Cartesian grid, which size is 250 x 250 x 250 m, using Cressman scheme. This gird size is equivalent of 8.3 m/s (=250m/30sec), that is close to a maximum of motion velocity or terminal falling velocity of rain droplets. In 3D visualization of radar reflectivity every 30 seconds, we found that a first echo appears in a middle layer, and it develops and spreads in an isolated cumulonimbus cloud. The echo reaches to the ground in 10 to 15 minutes after the appearance. In a case of an organized precipitation system (e.g. rain band), strong precipitation core appears in a middle or upper level in a cloud, and falls to the ground with growing. To evaluate the motion of the precipitation core quantitatively, we apply a modified TREC (Tracking Radar Echoes by Correlation) algorithm to compute the 3D motion vectors of precipitation from every 30 seconds data. On the other hand, we can calculate 3D wind vectors from dual-Doppler analysis every 30 seconds. In principle, Doppler velocity reflects the motion of precipitation particles. Because the particles are swept by wind, the particle horizontal motion indicates the horizontal wind velocity. However, looking as if the precipitation is an aggregate like a cloud system, the measured Doppler velocity represents in/out-flow from the system. In the dual-Doppler analysis, the vertical wind vectors are calculated from horizontal wind divergence using a continuity equation. Consequently, the vertical winds are entirely independent from the vertical motion derived from TREC algorithm. In fact, the result of dual-Doppler analysis showed convective circulation in a vertical slice is almost nothing on changing within a short period of 30 seconds (Fig. 2). The feature of wind vectors of convective circulation was maintained for at least several minutes. The vertical motion of a precipitation core is a result from both vertical winds (up/down-draft) and terminal falling velocity of precipitation particles. At the same time, the precipitation core is growing in the convection. Although the explanation of the vertical motion is complicated, the every 30 seconds 3D measurement has an advantage in investigating the precipitation development.