Cyclone-Induced Climate Risk and Adaptation
Cyclone-induced risks in Bangladesh commonly inclunde, storm-tides, and rainfall-driven inundation. In our work, we study these hazards including cascading effects that couple cyclones with heat-stress and seasonal rainfall extremes. ESSG’s extensive work within CREWSnet thus includes novel approaches to downscale cyclones, extreme rainfall, and to drive coastal, fluival, and pluvial processes considering tides and sea level rise.
ESSG’s work advances cyclone-induced climate risk and adaptation in various ways:
Science — physics-based Tropical Cyclone downscaling, climate controls on intensity (severity), frequency, and size, rapid intensification, paleotempestology, interactions between monsoon and cyclones, heat stress and cyclones.
Modeling: Downscaling using physics and machine learning, integrated flood models, and compound and cascade risk to support cyclone-induced risk assessment.
Observatories: ESSG has developed Bangladesh’s first Co-Active Soil Observatory (CASO) to support cyclone-induced hazard studies.
Assessments: The first cyclone-induced wind, rain, and flood climate hazards in Southwest Bangladesh.
Engagement: New community engagement models through ESSG’s Computational Sustainability Stack (CS2) are changing the way agricultural communities utilize risk assessments for climate adaptation under cyclone-induced hazards in a changing climate
Uneven future storm tide risk at the regional scale
Climate change, particularly tropical cyclone climatology changes and sea level rise, impacts different coastal regions of Bangladesh unevenly due to variations in topography and shoreline shape. We assess Bangladesh’s future storm tide risk at the regional scale, dividing its 300 km coastline into three distinct regions: the Ganges Tidal Plain (west), the Meghna Deltaic Plain (center), and the Chattogram Coastal Plain (east). Using a large-ensemble, multi-model approach, we quantified the amplification of storm tides across these regions under different climate scenarios. Our findings indicate that storm tide risks will increase across all regions by the end of the century, with greater amplification under higher emission scenarios. Notably, the Meghna estuary is projected to experience the highest increase in storm tide risk, emphasizing the need for stakeholders and planners to prioritize vulnerable areas when designing resilient infrastructure.
Left: Projections for the Ganges (A, D, G, J). Projections for the Meghna (B, E, H, K). Projections for the Chattogram (C, F, I, L). Storm tides are the total water levels (combined components of astronomic tide, storm surge, and mean sea-level state) relative to the mean sea level of the 1995-2014 baseline. Dashed lines indicate the ensemble median (0.5 quantiles), while shaded areas indicate each estimate’s confidence interval (CI, 0.1-0.9 quantiles). The current climate period (gray) spans from 1981 to 2000, while the future climate period (yellow for CMIP6 SSP2-4.5, light red for SSP3-7.0, deep red for SSP5-8.5) spans from 2081 to 2100. The confidence intervals account for variability in Tide, SLR, TCs, climate models, multi-station data, and Kernel-General Pareto Distribution fitting parameters.
Future risk of historically deadly storm tides
Historical storms and the storm tides they generate play a hugerole in understanding the relative risk in future climates. Because the damages from these storms are vividly remembered, communicating how likely they are in future scenarios is a robustway of generating early warnings for the oncoming rise in hazards. Our approach to quantify the distribution of cyclones in future climates allows assessments of the relative probability of observed cyclones in those regimes. We perform such assessments for the most damaging storms, here Bhola and Gorky, respectively. The increased hazard is alarming.
Above: TC Bhola (A) at Northern Bhola Island and TC Gorky (B) at station Anwara. The red pentagram icon in the top-right corner of each subplot marks the location of the maximum measured storm tide during the two landfall TCs. The median storm tide return period corresponding to the observed peak storm tide in present climate (gray), with whiskers indicating the estimated confidence interval (CI, 0.1–0.9 quantiles). The confidence intervals account for variability in Tide, Sea Level Rise (SLR), TCs, climate models, and return period fitting parameters. Incorporating SLR yields the largest reductions in return periods, making these historical storms 10-20 times more likely.
Shifted seasonal regimes of storm tide severity
Climate change will lead to seasonal disparities in future storm tide risk across Bangladesh due to the strong seasonality of cyclones in the Bay of Bengal. Cyclones are most active before and after the monsoon, while strong vertical wind shear during the monsoon inhibits their formation. Using a physics-based TC downscaling method, we capture the seasonal patterns under both current and future climate scenarios, which then translate into storm tide variations. Our results show that storm tide risk increases the most during the late monsoon and late post-monsoon seasons. The rise in late monsoon heightens the risk of cascading floods due to potential overlaps between cyclones and extreme monsoon rainfall, while the increase in late post-monsoon storm tides suggests that winter cyclones, currently rare, will become more frequent under warming scenarios. Understanding the seasonal changes is crucial for stakeholders and planners to better prepare for periods of heightened risk.
Left: Seasonal distribution of storm tide severity (greater than 2 m water level height)) under the current climate (A) and CMIP6 SSP2-4.5 (B), SSP3-7.0 (C), and SSP5-8.5 (D) climates, respectively. The increasing ratio (E) of extreme storm tide intensities across the year, with a 3-point sliding average applied to obtain monthly ratios. Two relatively inactive storm tide seasons are identified based on their seasonal behavior under the current climate (A): one during the late summer monsoon (±7 days around August 01, August 15, and September 01) and the other during the late post-monsoon (±7 days around December 01 and December 15). The light square located at the top of each swarm in subplots (A, B, C, D) indicates the mean value of the top ten extremes, while the area between two vertical dashed lines in subplot (E) indicates the monsoon season of the BoB.