Accurate forecasting of wind speeds on a bridge during a tropical cyclone is essential for safeguarding traffic and public safety. By predicting when wind speeds will exceed critical thresholds, the traffic operator can proactively close the bridge and prevent accidents. The present paper is a follow-up study of the simulation of wind speeds recorded on a bridge for two tropical cyclone cases in Hong Kong in 2023. The major change is the use of two additional sets of model outputs as boundary conditions, namely, an hourly re-analysis and a regional model, with the assimilation of meteorological observations in the vicinity of Hong Kong with higher spatial resolution. It turns out that, for a relatively larger tropical cyclone, the use of hourly re-analysis could improve the wind speed forecasts overall, yet there is no significant improvement in the prediction of the maximum wind speed at the top of the bridge arc and its occurrence time. On the other hand, the regional model does not improve the wind forecast. For a more compact tropical cyclone, no improvement in the wind forecast is observed, and significant discrepancies remain when compared with actual observations, even with the use of the two new sets of boundary conditions. Further efforts to improve the wind forecast would still be required.
Pak Wai CHAN (Mon,) studied this question.