Probabilistic-deterministic storm surge return level dataset for the Bengal delta

Bengal delta shoreline, spanning Bangladesh and India, gets hit every 3 years on average by a major tropical cyclone. Although their occurrence is relatively moderate compared to other tropical regions (accounting for only 5% of global cyclones), the impact of these events is major, accounting for 50% of the victims recorded worldwide. This is due to the very low topography of the delta above sea level (less than 5 meters), high storm surge induced water level and flooding, combined with the high density of the vulnerable population. On one hand, the unavailability of long-term reliable water level data on a sparse tide-gauge network along the coastline has hindered the assessment of storm surge hazards. The application of hydrodynamic modelling to fill the data gap also suffers from the unavailability of a reliable long-term storm dataset over the region. The complex topography of the Bengal delta, with defence structures, and a dense network of rivers presents another modelling challenge. Finally, the interaction of tide, surge and wave further complicate the numerical complexity, needing a coupled modelling framework. Thanks to advancements made to acquire high-quality regional nearshore bathymetry and topography (Krien et al. 2016, Khan et al. 2019), as well as coupled storm surge modelling (Krien et al. 2017, Khan et al. 2021), the tidal and storm surge dynamics over the Bengal delta is now well captured by recent high-resolution coupled SCHISM-WWM Bay of Bengal model (Khan et al. 2021). To estimate the risk of storm surge and associated flooding across the Bengal delta, we have integrated the wave-coupled hydrodynamic model of Khan et al. (2021) for a large ensemble (~3600 cyclones, ~5000 years of storm activity) of synthetic cyclones generated through the statistical-deterministic method of Emanuel (2006). Our storm and surge ensemble covers the whole range of natural variability of storm frequency, size, intensity and track location, with a dense spatial distribution. The interactions among the tide, surge, and waves are modelled explicitly at high spatial resolution. The storm surge-induced water level at various return periods, up to 500 years, is then determined at high spatial resolution (250m at the coast) using a ranking-based technique. The dataset distributed here represents the storm surge water level estimate (e.g. total water level from the tide, surge, and wave computed dynamically through the model) at 25 to 500 year return period (25-year step). The corresponding variable in the self-describing netCDF data file is 'maxelev'. The estimated storm surge water level values are interpolated in a 30" (~1km at the equator) structured grid over the Bengal delta from the original unstructured-grid model outputs (250m resolution at the coast). This dataset is a part of a manuscript, currently being submitted to Natural Hazards and Earth System Sciences (https://nhess.copernicus.org/). Please cite the original paper, along with the dataset if used in your work as - Khan, M. J. U., Durand, F., Emanuel, K., Krien, Y., Testut, L., and Islam, A. K. M. S.: Storm surge hazard over Bengal delta: A probabilistic-deterministic modelling approach, Nat. Hazards Earth Syst. Sci. Discuss. [preprint], https://doi.org/10.5194/nhess-2021-329, in review, 2021.横跨孟加拉国与印度的孟加拉三角洲（Bengal Delta）岸线，平均每3年就会遭受一次强热带气旋（tropical cyclone）袭击。尽管相较于其他热带区域，该区域热带气旋的发生频次相对温和（仅占全球热带气旋总数的5%），但其灾害影响却极为严重，全球范围内有50%的热带气旋遇难者来自此地。这一现象源于该三角洲极低的海拔（不足5米）、风暴潮（storm surge）引发的高水位与洪涝灾害，加之沿岸脆弱人口密度极高。 一方面，沿岸稀疏的验潮站（tide gauge）网络缺乏长期可靠的水位观测数据，极大阻碍了风暴潮灾害风险的评估工作；而为填补数据缺口而开展的水动力模拟（hydrodynamic modelling）应用，也因该区域缺乏长期可靠的风暴数据集而受限。孟加拉三角洲复杂的地形地貌、沿岸防护工程以及密布的河网，又为数值模拟带来了另一重挑战。最终，潮汐、风暴潮与海浪之间的相互作用进一步提升了数值模拟的复杂度，因此需要采用耦合模拟框架（coupled modelling framework）。 得益于区域近岸水深与地形（bathymetry and topography）高精度观测技术的进步（Krien等，2016；Khan等，2019），以及风暴潮耦合模拟方法的发展（Krien等，2017；Khan等，2021），近期基于高分辨率的孟加拉湾耦合SCHISM-WWM模型（Khan等，2021）已能够较好地复现孟加拉三角洲的潮汐与风暴潮动力过程。 为评估孟加拉三角洲全域的风暴潮及关联洪涝风险，本研究采用Emanuel（2006）提出的统计-确定性方法生成了大规模合成气旋集合（约3600个气旋，覆盖约5000年的风暴活动时长），并集成了Khan等（2021）的海浪耦合水动力模型开展模拟。本研究构建的风暴与风暴潮集合样本涵盖了风暴频次、尺度、强度及移动路径的全部自然变率范围，且空间分布极为密集；潮汐、风暴潮与海浪之间的相互作用以高空间分辨率显式模拟。 随后，研究采用基于排序的技术，以高空间分辨率（沿岸分辨率达250米）计算了重现期最高达500年的各重现期风暴潮增水水位。本次发布的数据集包含了25年至500年重现期（间隔25年）的风暴潮增水水位估算结果（即通过模型动态计算得到的潮汐、风暴潮与海浪共同作用下的总水位）。在自描述netCDF数据文件中，对应的变量名为"maxelev"。 研究将原始非结构网格模型输出结果（沿岸分辨率250米）插值至覆盖孟加拉三角洲的30角秒（赤道处约1公里）结构化网格中，得到最终的风暴潮增水水位估算值。本数据集为一篇已投稿至《Natural Hazards and Earth System Sciences》（https://nhess.copernicus.org/）的论文的配套数据。若您的研究中使用了本数据集，请引用如下原文及数据集信息：Khan, M. J. U., Durand, F., Emanuel, K., Krien, Y., Testut, L., and Islam, A. K. M. S.: Storm surge hazard over Bengal delta: A probabilistic-deterministic modelling approach, Nat. Hazards Earth Syst. Sci. Discuss. [预印本], https://doi.org/10.5194/nhess-2021-329, 已投稿待审, 2021.