Bangladesh is one of the most disaster-prone countries in the world, ranking seventh in the 1999–2018 Long-Term Climate Risk Index (Eckstein et al., 2019). Large portions of the population are exposed to the multiple natural hazards, including those derived from tropical cyclones (TCs), such as high winds, storm surge and flooding (e.g. Dilley et al., 2005). In the last 30 years, TCs impacting Bangladesh, from the Bay of Bengal (BoB), have been responsible for damages of ca. USD 8.9 billion and affected 45 million people (EM-DAT, 2021), with average annual extreme-weather-event-related losses amounting to 1.8 % of GDP between 1990 and 2008 (International Monetary Fund, 2019b). The wider north Indian Ocean basin averages five cyclones per year (accounting for ca. 7 % of global tropical cyclone activity) (Sahoo and Bhaskaran, 2016); however, there is some indication of a decrease in TC frequency (Alam et al., 2003; Mohapatra et al., 2017; Rao, 2004; Singh et al., 2019) and an increase in cyclone intensity (Balaguru et al., 2014) that is projected to continue under a warming climate (Knutson et al., 2020).

Recently, the International Monetary Fund (2019b) highlighted the early response Bangladesh is taking to the challenges posed by climate change; however, they also emphasise the importance of insurance mechanisms to enhance financial cover against impacts of natural disasters (International Monetary Fund, 2019a). Insurance facilitates disaster risk resilience and adaptation by transferring residual risk away from individuals and communities. Cost-effective and risk-informed sustainable development is based on the comprehensive understanding of hazards; the vulnerability of economies, societies and governments; and the exposure of society, people and belongings (UNDRR, 2019), but the lack of understanding of one or more of these components frequently limits the use of insurance mechanisms in many regions of the world most at risk from weather and climate hazards. This leaves significant populations around the world more vulnerable to the economic consequences of events that are otherwise manageable in countries with well-developed insurance markets (von Peter et al., 2012).

Detailed understanding of hazards is an essential part of understanding risk, but a relatively sparse meteorological observational network and interrupted non-continuous data records impose fundamental constraints on the description of TC hazards. Simulations of tropical cyclones in the BoB remain challenging for the current generation of seasonal forecasting systems (Camp et al., 2015), global climate models (Shaevitz et al., 2014) and reanalyses (Hodges et al., 2017), partly due the relatively coarse spatial and temporal resolution of the numerical simulations. It is well understood that large-scale thermodynamics and vertical wind shear have a significant impact on TC intensity, but there are also numerous vortex, convective, turbulent and frictional dissipative processes (e.g. Bryan and Rotunno, 2009; Nolan et al., 2007; Tang et al., 2015 amongst others) that occur on much smaller scales and also influence TC intensity, the impacts of which are not captured in low-resolution modelling. For example, extreme gusts associated with vigorous (deep) convection will generally be underestimated without kilometre-scale grid spacing that can explicitly resolve deep convection (e.g. Leutwyler et al., 2017; Weisman et al., 1997). More generally, as summarised by Leutwyler et al. (2017, and references therein), grid spacings of O(1km) are comparable to the size of the particularly energetic eddies in the planetary boundary layer.

Previous insights into TC hazards affecting Bangladesh focus on compiling catalogues of events (see Alam and Dominey-Howes, 2015, and references therein), or apply statistical analysis to event catalogues (e.g. Bandyopadhyay et al., 2018; Bhardwaj et al., 2020), and can only provide limited insight into the spatial extent, variability and magnitude of events based on first-hand eyewitness reports and limited observational records. Other authors take a parametric wind-field approach, combing the geostrophic (gradient) wind with a planetary boundary layer model to produce hazard maps at kilometre-scale resolution (e.g. Done et al., 2020; Krien et al., 2018; Tan and Fang, 2018); although this is a relatively computationally inexpensive approach, the quality of the result appears highly variable between global TC basins. Additionally, there are several holistic risk assessment views that combine multiple sources of hazard data, recognising that there are multiple hazards associated with TCs and that a combined risk assessment is non-trivial. However, these techniques are often limited to particular events (e.g. Hoque et al., 2016, 2019) or particular areas (e.g. Alam et al., 2020). In both cases, the quality of hazard and/or risk assessment is limited by available observational and track data.

In this study we seek to improve our understanding of the historical extreme gust speed hazard associated with recent TCs. To address the lack of observation data in this region, we use the latest-generation Met Office regional model over the BoB to simulate nine versions of 12 historical tropical cyclone cases representing 1979–2019. This generates spatially and temporally consistent counterfactual simulations (relative to observed TC cases), albeit limited by the constraints of the model configuration and computational resources. This ensemble configuration enhances our understanding of how each cyclone may evolve if a similar event were to happen again. We combine the ensemble information in a spatially coherent manner to produce hazard maps at 4.4 km resolution over Bangladesh for extreme wind (gust) hazards. Using Bayesian inference, we estimate gust speed exceedance intervals (return periods) across all of Bangladesh and demonstrate how this information can be directly integrated into a decision-making framework.

Tropical cyclone simulations are derived from a nine-member ensemble for 12 historical events, using the latest-generation Met Office Unified Model (Brown et al., 2012) convection-permitting regional atmosphere configuration RAL2-T, based on Bush et al. (2020) – hereafter referred to as RAL2. The RAL2 4.4 km domain avoids placing model boundaries over the Himalayas and covers Nepal, Bhutan, Myanmar, most of India, and parts of the Tibetan Plateau. To ensure model stability over this mountainous terrain, the RAL2 model was run with a 30 s time step. Each ensemble member requires a 24 h spin-up period as the RAL2 model adjusts from weak initial conditions taken from the ERA5 driving global model (of Hersbach et al., 2020). This initial 24 h period of model data is discarded in subsequent analysis and data files. Thereafter, each ensemble member is free running for a further 48 h, with hourly boundary conditions provided by ERA5. Collectively, the ensembles members sample a range of lead times before landfall from 12–36 h.

The parameterised RAL2 gust diagnostic represents a prediction of the 3 s average wind speed at every time step. The maximum of this 3 s average speed over an hour is then taken to give the hourly maximum 3 s gust speed. While not truly resolving deep convection, RAL2 is able to explicitly represent deep convective processes within the resolved dynamics. At these kilometre-scale resolutions the lower horizontal size limit of convective cells is still set by the effective resolution of 5 to 10 times the grid length (Boutle et al., 2014; Skamarock, 2004), and consequentially we expect that turbulent processes, as well as the dominant turbulent length scale, will still be under resolved in this 4.4 km dataset. The RAL2 model uses a gust parameterisation based on 10 m wind speed with scaling proportional to the standard deviation of the horizontal wind that also accounts for friction velocity, atmospheric stability and roughness length (Lock et al., 2019).

We use the ensemble output to first derive event “footprints” – a common method within the catastrophe modelling community to define peak hazard relating to a given event. In this case, footprints are based on the maximum wind gust speed achieved within each model run of 48 h, which implicitly collapses the time dimension to leave a 2D gust field in a longitude–latitude frame of reference. Although the original regional model data in Steptoe et al. (2021) covers a significant portion of the BoB, we crop the data to approximately 87.5 to 93.0∘ E and 20.5 to 27.5∘ N.

In general, median peak gust speeds from the RAL2 model ensemble are found to be 22 to 43 ms-1 faster compared to ERA5 reanalysis, but it is known that extreme gusts associated with vigorous convection in ERA5 are generally underestimated, sometimes by a factor of 2 (Owens and Hewson, 2018). For wind speed, the RAL2 median difference is 18 ms-1 faster compared to ERA5 and 5 and -3 ms-1 compared to the International Best Track Archive for Climate Stewardship data (IBTrACS, of Knapp et al., 2010, 2018) for the India Meteorological Department and Central Pacific Hurricane Center, Honolulu, regional forecast centres respectively. Further details of the regional modelling process and validation against IBTrACS and ERA5 reanalysis can be found in Steptoe et al. (2021).

Aggregating the 12 historical tropical cyclones ensembles, Fig. 3 shows the 50th, 95th and 99th percentiles of the posterior predictive maximum gust speed distribution across Bangladesh. Based on historical cases, the provinces of Chittagong, Barisal and Khulna are most exposed to high wind speed associated with tropical cyclone gusts, whilst Sylhet and Rajshahi are least exposed. The cities of Chittagong and Cox's Bazar are particularly at risk of maximum tropical cyclone gust speeds exceeding 45 ms-1 (87 kn) and 60 ms-1 (116 kn) respectively, in 5 % of events making landfall. Maximum gust speeds in Dhaka are likely to reach 35 ms-1 (68 kn) in 1 % of events, 25 ms-1 (48 kn) in 5 % to 50 % of events. We note that despite the northern provinces of Rajshahi, Rangpur and Mymensingh being over 200 km inland, they experience 95th and 99th percentile gust speeds greater than those observed in the populated provincial capitals of Dhaka, Barisal and Khulna. These extreme percentiles reflect the influence of cyclones Fani (May 2019) and Aila (May 2009) which had strong persistent in-land tracks.

The gust speed hazard can also be considered in terms of the probability of exceeding a threshold. Using WMO thresholds for tropical cyclone wind speeds (WMO, 2018), Fig. 4 shows that significant areas of southern provinces (Khulna, Barisal and Chittagong) will experience maximum wind speed in excess of “severe” cyclonic storm condition ≥25 ms-1 (48 kn) with a probability of 20 %–50 % per tropical cyclone event. At higher wind speeds, only areas within 30 km of the coastline are predicted to experience gust speeds in excess of “very severe” cyclonic storm conditions ≥33 ms-1 (64 kn) with the same likelihood (20 %–50 % per event). Wind speeds in excess of “super cyclonic” conditions ≥62 ms-1 (120 kn) are predicted to be exceeded with a likelihood of 0.5 %–5 % per event in limited areas south of Chittagong, with a small area in the vicinity of Cox's Bazar seeing exceedances of 5 %–10 % per event.

In addition to specific thresholds, exceedance probability curves (Fig. 5) summarise information for gust speeds up to 80 ms-1 (155 kn) for 18 of the most populated towns and cities in Bangladesh (grey lines) with four key cities highlighted. The coastal cities of Cox's Bazar and Chittagong are unsurprisingly the population centres most exposed to high gust speeds. Chittagong and Cox's Bazar are roughly 2.5 and 4.8 times more likely to experience tropical cyclones exceeding very severe cyclonic storm conditions than Dhaka for a landfalling cyclone.

Generalised additive models (GAMs) provide a useful framework for condensing spatial hazard information in an interpretable way, from multiple numerical model simulations, into a single spatially coherent hazard map. Using a restricted maximum likelihood approach to fit the GAM allows us to interpret model predictions in a Bayesian fashion that logically provides credible exceedance estimates. High-resolution convection-permitting numerical predictions of 12 historical cyclone events, in an ensemble model set-up, give an improved sense of the plausibility and likelihood of possible extreme events without being constrained by the lack of observational history in this region. Combining ensemble simulations with a GAM then allows us to robustly quantify the likelihood of maximum gust speed exceedances in a spatially coherent manner.

Our new maps of exceedance intervals show that north-western provinces of Bangladesh are relatively exposed to high-wind-speed hazards – in some areas the exceedance probabilities are equal to those experienced along the coast. Our hazard-to-decision-making framework suggests that these areas may need to be considered in an equivalent manner to coastal regions from a disaster risk reduction perspective. In coastal areas of Cox's Bazar and Chittagong we show super cyclonic conditions may occur as frequently as 1-in-20 to 1-in-100 years. We hope that these kilometre-scale hazard maps facilitate one part of the risk assessment chain to improve local ability to make effective risk management and risk transfer decisions. Future work to co-produce a proper loss function, given wind speed thresholds, would facilitate a method of transparent operational decision-making that could be used as the basis of an operational warning system.