總計畫暨子計畫:CReSS-NHOES高解析大氣海洋偶合模式之建置與梅雨防災應用(III)

The purpose of this project is two fold. The first is to introduce and impliment the Cloud-Resolving Storm simulator-Non-Hydrostatic Ocean model for Earth Simulator (CReSS-NHOES) coupled model to perform mei-yu quantitative precipitation forecast (QPF), in a hope to further improve the heavy-rainfall QPF in Taiwan mei-yu season. This year, the implementation of the CReSS-NHOES model is completed, using the JAMSTEC JCOPE2 data as inputs to NHOES, and five cases in 2017-2019 seasons are selected for evaluation and comparison with the QPFs produced by the CReSS model alone for day-1 (0-24 h), day-2 (24-48 h), and day-3 (48-72 h) QPFs. For this part, our overall results suggest that the inclusion of the NHOES model and air-sea interaction and feedback have positive impact to QPFs: The typical overforecast of rainfall by CReSS is reduced, and for most of the forecasts evaluated, the rainfall areas are also better placed in Taiwan. Thus, even for short-term forecasts within 72 h, positive impacts can be seen and this is very encouraging. The second purpose is to analyze and perform an ensemble-based analysis on the “super heavy-rainfall event” that occurred on 2 June 2017, both preliminary works. Synoptic analysis indicates that this case was associated with strong and deep southwesterly flow that transported warm and moist air from upstream toward Taiwan. The speed of the low-level jet (LLJ) could reach 30 kts at 850 hPa. The pre-storm environment was consitionally unstable, and winds veered with height to indicate warm air advection. Taiwan was located under the rear-right quadrant of the upper-level jet (ULJ), which provided divergence and uplifting, and was beneficial to triggering of deep convection. The rainfall in Taiwan on 2 June could be classified into two types: One that was associated with the Mei-yu front (in northern Taiwan) and the other from the southwesterly flow (in the mountains). While 1-km runs show importance of high resolution in producing the localized extreme rainfall in the north shore of Taiwan, results from our ensemble-based analysis indicate that the southwesterly flow (to the south of the front) was an important factor to more rainfall production in the model. The location, timing, and intensity of a frontal low that developed to the north of Taiwan also significantly influence the accumulative amount of rainfall. The formation of this meso--scale low increased local convergence just upstream of northern Taiwan, and helped the frontogenesis there, and led to more rainfall subsequently. The method adopted here can also used in real-time operation, and can provide useful information to forecasters to improve QPFs. In the appendices, we list some selected and relevant SCI papers and conference papers in this three-year project for reference.