5:15 PM - 6:45 PM
[AAS06-P05] Regional climate simulation of tropical cyclone at gray-zone and convection-permitting resolution over western North Pacific in 2013
Keywords:Tropical cyclone, Regional climate model, Convection-permitting resolution, Western North Pacific
Tropical cyclones (TCs) simulated at gray-zone (9km) and convection-permitting (4.5km) separately in the Advanced Research version of the Weather Research and Forecasting Model (WRF-ARW) are compared over the Western North Pacific (WNP) during 2013 TC seasons (June–November). The effects of the horizontal resolution and cumulus parameterization (CPS) are tested with three sets of experiments: 4.5km not involving CPS (NOCU4.5), 9km not involving CPS (NOCU9) and 9km involving CPS (CU9). Unlike NOCU4.5, although CU9 simulated closed total TCs, it yielded different TC tracks compared with the observation.
NOCU9 simulated much fewer TCs than observed consistent with Bian et al. (2023) and it is found that simulated TC number is sensitive to the horizontal resolution for NOCU experiments. Specially, we found that NOCU9 hardly captured TCs which were oriented from south of Japan to the east of China (SJEC). Overall, under the mesh grid of 9km, the simulated number of TCs is sensitive to parameterized convection which showed that the model produced dramatically fewer TCs once the parametrization was turned off. However, under the convection-permitting resolution, the model could simulate appropriate TC number and more similar TC tracks. Compared with NOCU9, NOCU4.5 could simulate more strong TCs closer to the observations of IBTrACs. Then CU9 could reproduce the observed relationship well, it could not capture strong TCs (Vmax> 50 m/s). NOCU4.5 has a tendency to produce a relationship that drift away from the observed with TC intensity growing. Last, It seems that NOCU4.5 simulates smaller TC size which might be related with some environment factors such as larger vertical wind shear and smaller eady growth rate.
NOCU9 simulated much fewer TCs than observed consistent with Bian et al. (2023) and it is found that simulated TC number is sensitive to the horizontal resolution for NOCU experiments. Specially, we found that NOCU9 hardly captured TCs which were oriented from south of Japan to the east of China (SJEC). Overall, under the mesh grid of 9km, the simulated number of TCs is sensitive to parameterized convection which showed that the model produced dramatically fewer TCs once the parametrization was turned off. However, under the convection-permitting resolution, the model could simulate appropriate TC number and more similar TC tracks. Compared with NOCU9, NOCU4.5 could simulate more strong TCs closer to the observations of IBTrACs. Then CU9 could reproduce the observed relationship well, it could not capture strong TCs (Vmax> 50 m/s). NOCU4.5 has a tendency to produce a relationship that drift away from the observed with TC intensity growing. Last, It seems that NOCU4.5 simulates smaller TC size which might be related with some environment factors such as larger vertical wind shear and smaller eady growth rate.