May 19, 2026•2 min read•from Frontiers in Marine Science | New and Recent Articles

Machine learning based accurate storm surge peak and timing forecast in Pearl River Estuary

Our take

As climate warming intensifies storm surge threats to coastal cities, accurate and timely forecasts of peak surge height and timing are essential. This study presents a novel single-station Gated Recurrent Unit (GRU) model developed using data from 42 tropical cyclones in the Pearl River Estuary. By integrating tropical cyclone parameters and employing Support Vector Regression (SVR) for peak timing correction, the model achieved significant reductions in peak underestimation and timing errors.

Machine learning based accurate storm surge peak and timing forecast in Pearl River Estuary

In the face of climate change, storm surges pose an increasing threat to coastal cities, necessitating accurate forecasts for peak surge height and timing. The recent study on a machine learning-based approach to storm surge forecasting in the Pearl River Estuary exemplifies the kind of innovative solutions that are essential to address these challenges. Utilizing a Gated Recurrent Unit (GRU) model and refining it with an iterative forecasting scheme, the researchers have made significant strides in improving both the magnitude and timing forecasts of storm surges. This research is particularly relevant when considering other studies, such as Deep learning-based sea level anomaly forecasting around Taiwan Island integrating ConvLSTM and attention mechanism, which highlight the complexities of forecasting in dynamic coastal environments.

The study leverages storm surge data from 42 tropical cyclones over a 23-year period, demonstrating the importance of historical data in enhancing forecast accuracy. By addressing the limitations of traditional GRU models, which tend to underestimate peak magnitudes and struggle with timing errors, this research provides a critical advancement in predictive capabilities. The integration of tropical cyclone parameters through methods such as Accumulated Local Effects (ALE) significantly enhances the model's reliability. Moreover, the use of a Support Vector Regression (SVR) model to correct peak timing errors represents a thoughtful approach to refining predictions, marking a step forward in the application of machine learning in meteorological forecasting. Such advancements are vital for coastal resilience, especially as we observe the impacts of severe weather events becoming more frequent and intense.

Understanding the implications of this research extends beyond immediate forecasting improvements. It speaks to a broader need for integrating advanced technological solutions in environmental monitoring and disaster preparedness. As coastal communities grapple with rising sea levels and increasing storm activity, the ability to provide precise and timely forecasts becomes essential for effective response planning. This is echoed in other relevant discussions, such as Geopolitical conflicts and the restructuring of maritime transport networks: the causal effect of the Red Sea crisis on port throughput, which underscores the interconnectedness of environmental factors and global logistics. Accurate forecasting can mitigate risks, enabling communities and policymakers to make informed decisions that protect lives and infrastructure.

Looking forward, the advancements presented in this study raise important questions about the future of storm surge forecasting and its role in climate resilience. As machine learning techniques continue to evolve, how can we further refine these models to incorporate real-time data and improve predictive accuracy? Additionally, what collaborative efforts can be fostered among researchers, policymakers, and technology developers to share knowledge and best practices? As we navigate these challenges, it is crucial to remain focused on fostering an integrated data ecosystem that enhances our understanding of ocean dynamics and contributes to global ocean stewardship. The future of storm surge forecasting holds promise, and with continued innovation and collaboration, we can better prepare for the uncertainties that lie ahead.

With climate warming, storm surge increasingly threatens coastal cities, underscoring the necessity for accurate and timely forecasts of peak surge height and timing. Among widely used neural network-based approaches in storm surge forecasting, the Gated Recurrent Unit (GRU), while strong in sequential prediction, tends to underestimate peak magnitudes and exhibits timing errors at extended lead times. To address the issue, this study developed a single-station GRU prediction model using storm surge data from 42 tropical cyclones (2000-2023) at nine stations in the Pearl River Estuary. To reduce peak underestimation, an iterative forecasting scheme was designed, incorporating tropical cyclone (TC) parameters selected via Accumulated Local Effects (ALE) and correlation coefficient. To mitigate peak timing errors, a Support Vector Regression (SVR) model based on TC peak intensity elements was proposed to predict and correct peak timing errors. Additionally, multi-station joint forecasting was explored by embedding spatial information between stations and TC tracks. Results based on leave−one−out cross−validation (LOOCV) indicated that integrating TC location, distance, and 34−kt wind radii in the key quadrants yielded a 27% reduction in Peak Error (PE) at 12–18 h and a 13% reduction in RMSE over 1–18 h for the 35 test cases. After applying the SVR−based phase correction, the forecast for all cases achieved an 55% reduction in Timing Error (TE) and a 2.81% reduction in RMSE relative to the uncorrected forecast. Compared to numerical model outputs, the RMSE of water level sequence during Typhoon Hato was decreased by 0.22 m. Multi-station joint forecast results showed that embedding such spatial information not only enhances forecast skill across multiple gauges but also provides benefits for stations with sparse historical data. This study offers a promising method that targets both the magnitude and timing of storm surge peaks, thereby improving the accuracy and reliability of storm surge forecasting.

Read on the original site

Open the publisher's page for the full experience

View original article →

Tagged with

#ocean data#data visualization#climate monitoring#climate change impact#storm surge#Pearl River Estuary#machine learning#forecasting#tropical cyclone (TC)#Gated Recurrent Unit (GRU)#Support Vector Regression (SVR)#peak surge height#timing errors#iterative forecasting scheme#multi-station joint forecasting#cross-validation (LOOCV)#Peak Error (PE)#root mean square error (RMSE)#peak intensity elements#Accumulative Local Effects (ALE)