Yamato TAKEDA

Study on Estimation of Ground Precipitation during Snowfall Using Two Types of Weather Radar Data

Toshiro KUMAKURA

Solid precipitation particles such as snowflakes fall slowly and are advected by the wind, so they do not fall directly below the observation surface of a weather radar. Previous studies using backward trajectory analysis using a three-dimensional wind field that takes into account the advection of solid precipitation particles have used a weather radar (NUT-Radar) installed at Nagaoka University of Technology, but the problem was that it could not be performed outside the observation range of NUT-Radar. Therefore, this study examined the possibility of estimating ground precipitation when performing backward trajectory analysis using the Japan Meteorological Agency's nationwide synthetic radar data (JMA-Radar) for the period from January to February 2022. The analysis target site was the AMeDAS Nagaoka site, and cases where the AMeDAS temperature was below 0℃ were targeted. A backward trajectory analysis was performed using five elevation angle data from NUT-Radar, horizontal wind speed predicted by the Japan Meteorological Agency's non-hydrostatic model (NHM), and a fall speed of 1.0m/s, and the distribution pattern of vertical precipitation intensity was analyzed. Additionally, precipitation intensity from JMA-Radar was obtained using backward trajectory analysis, and the results were evaluated based on the precipitation intensity distribution from NUT-Radar. Analysis of the precipitation intensity distribution from NUT-Radar showed that the advection of precipitation particles was accurately estimated in approximately 70% of cases where precipitation was present, which is thought to enable precipitation amount estimation using JMA-Radar with high accuracy. A simple regression analysis of the precipitation obtained from AMeDAS and JMA-Radar after converting it to 1-hour precipitation amounts showed that it is possible to estimate ground precipitation amount using backward trajectory analysis using JMA-Radar as well.