|
From June 1-2, 2017, the Meiyu front triggered an extreme rainfall event in northern Taiwan, with accumulations exceeding 400 mm in eight hours. This study applies a k-means clustering approach to partition heavy rainfall distributions from 128 ensemble simulations of the WRF model into five groups and compares them with three-dimensional wind fields retrieved by the Wind Synthesis System using Doppler Measurements (WISSDOM). We examine dynamical differences between clusters whose rainfall locations align with observations and those that differ, identifying the key factors that govern heavy rainfall placement in northern Taiwan: : the strength and orientation of marine boundary layer jets (MBLJs) and barrier jets (BJs), together with the movement speed of the low-pressure center of the northern Meiyu front. These findings offer a robust basis for improving forecasts of the extreme rainfall events and their likely locations.
|