Extreme wind return periods from tropical cyclones in Bangladesh: insights from a high-resolution convection-permitting numerical model

We use high-resolution (4.4 km) numerical simulations of tropical cyclones to produce exceedance probability estimates for extreme wind (gust) speeds over Bangladesh. For the first time, we estimate equivalent return periods up to and including a 1-in-200 year event, in a spatially coherent manner over all of Bangladesh, by using generalised additive models. We show that some northern provinces, up to 200 km inland, may experience conditions equal to or exceeding a very severe cyclonic storm event (maximum wind speeds in >= 64 kn) with a likelihood equal to coastal regions less than 50 km inland. For the most severe super cyclonic storm events (>= 120 kn), event exceedance probabilities of 1-in-100 to 1-in-200 events remain limited to the coastlines of southern provinces only. We demonstrate how the Bayesian interpretation of the generalised additive model can facilitate a transparent decision-making framework for tropical cyclone warnings.