Increasing sequential tropical cyclone hazards along the US East and Gulf coasts

Sequential tropical cyclone (TC) landfalls, where two or more TCs make landfall in close proximity, pose significant challenges to coastal communities. These sequential hazards exacerbate damage due to the weakened resilience of infrastructure and communities following the initial impact. This study focuses on the temporal compounding of TC hazards, investigating how the probability of sequential landfalls and their associated hazards has changed in the past and is projected to change in the future. The increase in TC intensity and frequency, along with the effects of sea-level rise (SLR), are considered as potential drivers of this change. Understanding these temporal compounding events is crucial for developing effective mitigation and adaptation strategies for coastal regions. This is particularly relevant to the US East and Gulf Coasts, which are highly vulnerable to TC impacts. The research aims to quantify the historical trend of sequential TC hazards and provide robust projections of future events while incorporating uncertainties in TC frequency predictions. This will involve analyzing both historical observations and advanced climate simulations to pinpoint the underlying physical mechanisms behind this increase in sequential TC hazards.

Previous research has examined individual and joint TC hazards, focusing on the combined effects of wind, surge, and rainfall. However, studies on sequential TC hazards – where the impacts of multiple storms compound over time – remain limited. Existing literature suggests an increasing likelihood of sequential TC landfalls and a general trend towards more hazardous TCs. However, the analysis of historical trends in sequential hazard occurrence, the robustness of future projections considering uncertainties in TC frequency, and the precise mechanisms driving changes in sequential hazards are still unclear. This research builds upon previous findings by explicitly investigating the historical trend and developing robust projections of these temporally compounding TC hazard events, offering a more comprehensive understanding of coastal risk.

The study employed a multi-faceted approach combining historical observations and climate simulations. For the historical analysis (1949-2018), a probabilistic model describing sequential TC landfalls was extended to incorporate hazard-producing capabilities and durations of TCs. This model was fit to historical TC hazard data from nine locations along the US East and Gulf coasts, obtained from the International Best Track Archive for Climate Stewardship and other sources. The yearly minimal impact interval (MII), the minimum time between hazard-producing TCs, served as a key metric. The 95th percentile of daily maximum water level, total rainfall, and maximum wind speed were used as thresholds to define hazard-producing events. For future projections (2070-2100), a physics-based TC hazard analysis method was applied. Synthetic TCs were generated using a statistical-deterministic TC model forced by six Coupled Model Intercomparison Project Phase 6 (CMIP6) climate models under both high (SSP58.5) and moderate (SSP24.5) emission scenarios. Storm tides were simulated with ADCIRC, rainfall with the TCR model, and wind using the complete wind profile model. Probabilistic SLR projections from the IPCC were incorporated. The probabilistic model was then fit to the simulated TC hazards to assess changes in the return period of various MII levels. The study also analyzed marginal and joint TC wind, surge, and rainfall hazards to better understand the causes of changes in sequential TC hazards. A sensitivity test was conducted using future storms with control frequency to isolate the impact of increased hazard severity from changes in landfall frequency.

The analysis revealed a clear increasing trend in the yearly probability of experiencing sequential TC hazards at most locations, with some locations doubling the probability over the past seven decades. The return period of MII (minimal impact interval) for sequential TC hazards is projected to substantially decrease by 2100 along the US East and Gulf coasts. For example, a 15-day MII, a 40-year event in the control climate for Texas, would become a 2-year event under the high emission scenario (SSP58.5) when SLR is considered. Across the US East and Gulf coasts, under SSP58.5, the return period of a 15-day MII would decrease from 10-92 years to 1-2 years (considering SLR). This decrease is driven by both increased landfall frequency and increased storm hazard-producing ability. Sensitivity tests showed that increased hazard severity, mainly due to increased TC intensity and SLR, significantly contributes to the increased frequency of sequential TC hazards, even without an increase in the overall frequency of TCs. The analysis of joint hazard probabilities indicated a substantial decrease in the return period of joint hazards across all US coastal regions, with a significant increase in the proportion of landfalling TCs producing joint hazards. SLR significantly increases the proportion of landfalling TCs producing surge hazards, even for weaker storms, reaching around 80% under both high and moderate emission scenarios. SLR also increased the hazard duration. The study also investigated the probability of sequential extreme events, such as a Katrina-like storm followed by a Harvey-like storm within 15 days, finding that such events, nonexistent in the control simulation, could have an annual occurrence probability exceeding 1% by 2100 under the high emission scenario. The differences between the SSP2 4.5 and SSP5 8.5 scenarios in the change in hazard-producing storm ratios were relatively small, particularly for the Gulf Coast and surge-producing storm ratios (without considering SLR).

The findings demonstrate a significant increase in the probability of sequential TC hazards along the US East and Gulf coasts, primarily driven by increased TC hazard severity, particularly rainfall and surge. SLR plays a crucial role, making even weaker TCs capable of producing significant surge hazards. The increased TC rainfall rates, projected to increase by 10–32% by 2100, contribute to this increased hazard potential. While TC climatology change is a dominant factor, SLR has a stronger impact on the increase in moderate events. The similarity in the return period of sequential TC hazards under both high and moderate climate change scenarios highlights the importance of mitigation efforts, regardless of the specific emission trajectory. The higher frequency of sequential extreme events under the SSP58.5 scenario underlines the potential for catastrophic cascading failures. This necessitates prioritizing investments in infrastructure resilience and robust emergency response systems.

This study highlights the increasing risk of sequential tropical cyclone hazards along the US East and Gulf coasts due to increased TC hazard severity and SLR. The projected increase necessitates urgent adaptation and mitigation strategies. Further research should investigate the specific vulnerabilities of different coastal communities and regions to sequential TC hazards, and develop tailored adaptation measures to enhance resilience. More detailed investigations into the uncertainties in SLR projections and their role in transforming weak TCs into hazard-producing events are also needed.

The study uses the 95th percentile as a threshold for defining hazard-producing storms. While sensitivity analyses using different percentiles were performed, the choice of percentile remains somewhat arbitrary. The study also assumes conditional independence of storms within a year, which might not perfectly capture the complex interactions between sequential events. Future research could investigate more sophisticated models to address these limitations. The use of synthetic TCs, while offering valuable insights, involves inherent uncertainties associated with the underlying climate models and statistical downscaling techniques.