Author(s): Nguyen Van Trung; Han Soo Lee
Linked Author(s): Han Soo Lee
Keywords: Sensitivity analysis; WRF; Typhoon Sinlaku (2014); The south China sea
Abstract: On November 26,2014 Tropical Storm Sinlaku formed over the southeastern Philippines. As a tropical depression Sinlaku caused flooding in areas of the Visayas and Mindanao. Due to low vertical wind shear, Sinlaku gathered strength while it was on the South China Sea. The storm then moved west and crossed the South China Sea where it headed to Vietnam for a final landfall. It made landfall in east-central Vietnam bringing some moderate to heavy rainfall with it and consequential local flooding. In the present study, we have used the Weather Research and Forecasting (WRF) model to simulate typhoon Sinlaku (2014) and examined its sensitivity via 12 alternative model configurations, including four different model physics options: vertical layers, microphysical schemes, cumulus schemes and planetary boundary layer (PBL) schemes. WRF modellings are carried out with three nested domains with the innermost domain of 3 km grid spacing with explicit convection. Comparisons between estimated and observed data were made through standard statistical measures such as correlation coefficient, root mean square error and bias. The results illustrate that air temperature at 2 m above surface and surface pressure seem to be less sensitive to number of vertical layers because the less vertical layers show the better agreements with the observed data. Meanwhile, the combination between Lin et al. scheme and Kessler scheme for microphysics has been found to produce better estimates of air temperature, surface pressure and wind speed for this region. For cumulus schemes, both Kain-Fritsch (new Eta) scheme and Grell 3D ensemble scheme showed their effects to the results. In terms of temperature, Kain-Fritsch (new Eta) scheme show the better agreements with the observed, while Grell 3D ensemble scheme is suitable for surface pressure and wind speed. The PBL scheme showed that the Mellor-Yamada-Janjic TKE scheme have the best result compared with the observed and those from other schemes. Overall, this analysis is useful for further optimization of WRF model parameters to improve typhoon simulation in the South China Sea region.