Global Navigation Satellite System (GNSS) radio occultation (RO) provides plentiful sounding profiles over regions lacking conventional observations. The Gridpoint Statistical Interpolation (GSI) hybrid system for assimilating RO data is integrated in this study with the Model for Prediction Across Scales–Atmosphere (MPAS) to improve tropical cyclone forecasts. After the MPAS-GSI assimilation cycles, dynamical vortex initialization (DVI) that may effectively spin up the initial inner typhoon vortex through cycled model integration is implemented to improve the initial analysis fit to the best track position as well as maximum wind or pressure intensity for Typhoon Nepartak (2016) that moved northwestward toward southern Taiwan. During the cycling assimilation, assimilation with RO data improves the temperature and moisture analysis, and largely reduces the forecast errors compared to those without RO data assimilation. The two RO operators that assimilate local bending angle or refractivity produce similar analyses, but the temperature and moisture increments from bending angle assimilation are slightly larger than those from refractivity assimilation. The MPAS forecasts at 60-15 km resolution show that the typhoon track prediction is improved with RO data, especially using bending angle data. The reduction in track deviations is explained by the wavenumber-one potential vorticity budget for several forecasts associated with the track deflection near southern Taiwan. Assimilation of RO data has fewer impacts on the typhoon intensity forecast compared to the DVI that largely improves the initial and thus forecasted intensity of the typhoon but at the cost of a slightly degraded track. Use of the enhanced 3 km resolution in the typhoon path also further improved the forecasts with and without the DVI. The feasible performance of the MPAS-GSI system with the RO data impact is also illustrated for Typhoon Mitag (2019), that passed around northern Taiwan.
|出版狀態||已出版 - 9月 2022|