Figure: The 19 estuaries used to develop the ML model
Figure: Developed ML model to detect disturbances in water quality in estuaries
Figure: Consecutive cyclones (Dolly, Fay) caused disturbances at Weeks Bay, AL ( The red-shaded area prominently highlights the potential impact period of the cyclone. The values of all four features are normalized between 0 and 1)
This study develops a novel approach to assessing the impact of tropical cyclones on estuarine ecosystems. It is motivated by the significant disruptions these storms cause, including deteriorating water quality, changes in phytoplankton productivity, and physical damage to vegetation. Given the increasing frequency and intensity of tropical cyclones due to climate change, it is crucial to develop robust methods for detecting and quantifying their effects on estuaries. Traditional statistical methods have been used to analyze water quality variations, but they often struggle to capture the complex and dynamic nature of estuarine systems.
The study employs a Long Short-Term Memory (LSTM)-based deep learning model to detect disturbances in estuarine water quality, quantify their severity, and estimate recovery times. The methodology involves training the model using data from NOAA’s National Estuarine Research Reserve System, allowing it to distinguish between cyclone-induced anomalies and natural fluctuations in environmental parameters such as salinity, turbidity, temperature, and dissolved oxygen. Additionally, a Gaussian filter-based algorithm is developed to estimate the time required for estuaries to return to pre-cyclone conditions. The findings demonstrate the model’s ability to capture the impact of tropical cyclones with high accuracy, marking a significant advancement over conventional statistical methods.
This research is particularly important as it bridges the gap between machine learning and ecological monitoring, providing a scalable and adaptable tool for environmental resilience assessment. By quantifying disturbances and estimating recovery times, the study offers valuable insights for policymakers, conservationists, and resource managers involved in coastal protection and disaster mitigation. Moreover, the developed model has broader applicability beyond estuaries, as it can be used to detect and analyze anomalies in other time-series datasets related to climate and environmental science.
Ibrar, M.A., Usama, M. & Salman, A.M. A machine learning model for detecting and quantifying tropical cyclone related disturbance and recovery in estuaries. Sci Rep 15, 5230 (2025). https://doi.org/10.1038/s41598-025-89196-6
Figure: Average changes in SST from 2020 to 2100 for RCPs 2.6, 4.5, 6.0, and 8.5 in ℃
Figure: Changes in 3-sec gust wind speed from 2020 to 2100 for different Mean Recurrence Intervals (MRI)
Figure: Changes in storm surge hazard from 2020 to 2100 for different Mean Recurrence Intervals (MRI)
Figure: Percentage increase due to storm surge and sea level rise by the end of the century compared to the estimated surge height for the year 2020 corresponding to 100-year MRI
This study investigates how rising sea surface temperatures (SST) driven by climate change influence hurricane intensity and storm surge hazards. The study is motivated by the increasing threat of hurricanes, which cause extensive economic damage and loss of life, particularly in coastal communities. With climate change projected to increase SST, understanding how this warming affects hurricane wind speeds and storm surge is critical for infrastructure planning, risk assessment, and disaster preparedness.
The study employs an Empirical Track Model (ETM) to simulate hurricane tracks and intensities under different climate scenarios based on the Intergovernmental Panel on Climate Change (IPCC) Representative Concentration Pathways (RCPs 2.6, 4.5, 6.0, and 8.5). The storm surge hazard is modeled using the Sea, Lake, and Overland Surges from Hurricanes (SLOSH) model, a widely used computational tool for estimating hurricane-induced flooding. The study considers various SST projections and examines their impact on hurricane wind speeds, frequency, and storm surge risk in selected locations across the US Atlantic and Gulf Coasts. Findings reveal that rising SST will result in stronger hurricanes, particularly in the northeast Atlantic coast. The Gulf Coast is expected to face the highest increases in storm surge heights, exacerbated by sea-level rise (SLR).
This work underscores the growing risks posed by climate change-induced hurricanes and storm surges. The study highlights the urgent need for updated design standards, improved flood risk mapping, and enhanced resilience planning for coastal infrastructure. By integrating climate projections with hurricane modeling, the research provides valuable insights for policymakers, engineers, and emergency planners to mitigate future hurricane impacts on vulnerable communities.
Salarieh, B., Ugwu, I.A. & Salman, A.M. Impact of changes in sea surface temperature due to climate change on hurricane wind and storm surge hazards across US Atlantic and Gulf coast regions. SN Appl. Sci. 5, 205 (2023). https://doi.org/10.1007/s42452-023-05423-7