In this study, a storm surge model of the semi-enclosed Tokyo Bay was constructed to investigate its hydrodynamic response to major typhoon parameters, such as the point of landfall, approach angle, forward speed, size, and intensity. The typhoon simulation was validated for Typhoon Lan in 2017, and 31 hypothetical storm surge scenarios were generated to establish the sensitivity of peak surge height to the variation in typhoon parameters. The maximum storm surge height in the upper bay adjacent to the Tokyo Metropolitan Area was found to be highly sensitive to the forward speed and size of the passing typhoon. However, the importance of these parameters in disaster risk reduction has been largely overlooked by researchers and disaster managers. It was also determined that of the many hypothetical typhoon tracks evaluated, the slow passage of a large and intense typhoon transiting parallel to the longitudinal axis of Tokyo Bay, making landfall 25 km southwest, is most likely to cause a hazardous storm surge scenario in the upper-bay area. The results of this study are expected to be useful to disaster managers for advanced preparation against destructive storm surges.

In this study, 42 years of tidal records and landfall TC best tracks in Japan were used to demonstrate that TC pre-landfall forward speed is significantly correlated with maximum storm surge height. Coastal morphology was the determining factor for the correlation between storm surge and TC forward speed. Fast-moving TCs tended to amplify the storm surge along open coastlines (Pearson correlation coefficient, R = 0.62), but reduce it in semi-enclosed bays (R = -0.52). The negative correlation contrasts with the general perception that the coincidence of TC wind speed and forward speed vectors generates a larger storm surge. The influence of coastal morphology was most prominent for TCs with a central pressure lower than 956 hPa. Tropical cyclone (TC) operational forecasts are continuously improving; however, there is still scope to improve the precision of storm surge predictions. These findings could contribute to the improvement of storm surge forecasting and provide emergency management personnel with more precise early warnings of dangerous storm surges.

Over the past several decades, operational forecasts of typhoon tracks have improved steadily. However, storm surge forecast skills have experienced rather modest improvements as it has been assumed to be primarily a function of maximum typhoon wind speed. In this study, numerical sensitivity experiments have been conducted for the semi-enclosed Tokyo Bay to investigate the existence of any connection between typhoon size and peak storm surge height. The radius of the maximum wind (Rmax) derived based on the 50-kt wind radius (R50) is used to define the size of a typhoon. The results show that peak storm surge height tends to increase as the size of typhoon becomes larger, which may also be supported by historical observations. Storm size plays a significant role in surge generation, particularly for very large typhoons making landfall in the upper bay. Analyses show that for a given hypothetical typhoon, the water level in the inner bay is increased by 1 m, changing Rmax from 13 km to 89 km. The findings of this study will be beneficial for the storm surge modeling community as it gives insight into the role of typhoon size, which is essential to forecast peak surge height precisely.