The National Center for Atmospheric Research (NCAR) has been collaborating with the Taiwan Central Weather Bureau (CWB) to transfer and tune the AutoNowCaster (ANC) system for the purpose of nowcasting heavy rainfall-producing convection. The ANC is an expert system that produces rapidly updating (approximately every 5 minutes), 1-hour nowcasts of thunderstorm initiation, growth, and decay. An important component of the system is the 4-D Variational Doppler Radar Assimilation System (VDRAS), a four-dimensional variational assimilation system that produces frequently updating (on the order of 10 minutes) analyses using Doppler radar, surface observations, and a mesoscale model background. This study focuses on the three-dimensional wind field analyses produced by the system and how they may be used to aid in nowcasting convective initiation driven by local forcing.

Using data from the Southwest Monsoon Experiment/Terrain-influenced Monsoon Rainfall Experiment (SoWMEX/TiMREX), 17 weak-synoptic forcing days were chosen and the time and location of initiations within the SoWMEX/TiMREX network in southern Taiwan identified. VDRAS was run using output from the Weather Research and Forecasting (WRF) model as the mesoscale background, and radar data from the NCAR S-Pol research radar and the Taiwanese Chi-Gu (RCCG) operational radar and surface observations were assimilated to produce the final analyses. This paper will present the results of analysis of the VDRAS winds to determine how well VDRAS identifies the local circulations within the complex terrain of Taiwan that influence when and where initiation occurs and how the VDRAS wind analyses may be used within the ANC system to nowcast thunderstorm initiation.