Storm surge modeling in the AI era: Using LSTM-based machine learning for enhancing forecasting accuracy

Coastal Engineering, Journal Year: 2024, Volume and Issue: 191, P. 104532 - 104532

Published: April 20, 2024

Language: Английский

---

Real-time forecasting of coastal flood inundations under regulated reservoir and storm-tide influences

Advances in Water Resources, Journal Year: 2025, Volume and Issue: unknown, P. 104920 - 104920

Published: Feb. 1, 2025

Language: Английский

---

Advancing storm surge forecasting from scarce observation data: A causal-inference based Spatio-Temporal Graph Neural Network approach

Coastal Engineering, Journal Year: 2024, Volume and Issue: 190, P. 104512 - 104512

Published: March 23, 2024

Language: Английский

---

Surge-NF: Neural Fields inspired peak storm surge surrogate modeling with multi-task learning and positional encoding

Coastal Engineering, Journal Year: 2024, Volume and Issue: 193, P. 104573 - 104573

Published: July 9, 2024

Language: Английский

---

Sea level forecasting using deep recurrent neural networks with high-resolution hydrodynamic model

Applied Ocean Research, Journal Year: 2025, Volume and Issue: 157, P. 104496 - 104496

Published: March 10, 2025

Language: Английский

---

Machine Learning Techniques for Predicting Typhoon‐Induced Storm Surge Using a Hybrid Wind Field

Journal of Geophysical Research Machine Learning and Computation, Journal Year: 2025, Volume and Issue: 2(2)

Published: April 10, 2025

Abstract Accurate and timely storm surge prediction is critical information in coastal zone management risk reduction strategies. The Bohai Sea, a semi‐enclosed bay the Northwest Pacific that used to be less prone typhoon disasters, has been witnessing paradigm shift activities recent past. Since there have limited typhoon‐induced surges an innovative system warranted address frequent intense impacts. Four Machine Learning (ML) models (Long Short‐Term Memory (LSTM), Convolutional Neural Networks (CNN), CNN‐LSTM, ConvLSTM) were built predict significantly improve when combined with three‐dimensional Finite Volume Community Ocean Model (FVCOM), is, FVCOM‐ML. In this study, FVCOM‐ML model was driven by hybrid wind field superimposed Holland reanalysis field. ML trained via Advanced Circulation simulations compensate for in‐situ observations. performances analyzed both spatial (e.g., single multiple sites) temporal steps) scale variability. overcome residual error of FVCOM, effectively reducing inherent uncertainty traditional methods. offers significant advantage over standalone FVCOM or while better incorporating realistic physical constraints improving accuracy forecasts.

Language: Английский

---

Artificial neural network-based multi-input multi-output model for short-term storm surge prediction on the southeast coast of China

Ocean Engineering, Journal Year: 2024, Volume and Issue: 300, P. 116915 - 116915

Published: March 14, 2024

Language: Английский

---

The Impact of Climate Change and Urbanization on Compound Flood Risks in Coastal Areas: A Comprehensive Review of Methods

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(21), P. 10019 - 10019

Published: Nov. 2, 2024

Many cities worldwide are increasingly threatened by compound floods resulting from the interaction of multiple flood drivers. Simultaneously, rapid urbanization in coastal areas, which increases proportion impervious surfaces, has made mechanisms and simulation methods disasters more complex. This study employs a comprehensive literature review to analyze 64 articles on risk under climate change Web Science Core Collection 2014 2024. The identifies for quantifying impact factors such as sea level rise, storm surges, extreme rainfall, well like land subsidence, drainage systems floods. Four commonly used quantitative studying discussed: statistical models, numerical machine learning coupled models. Due complex structure high computational demand three-dimensional joint probability along with increasing number drivers complicating grid interfaces frameworks coupling different most current research focuses superposition two disaster-causing factors. three or change-driving is emerging significant future trend. Furthermore, often overlooked studies should be considered when establishing Future focus exploring models better simulate, predict, understand mechanisms, evolution processes, disaster ranges change.

Language: Английский

---

Enhanced prediction of cement raw meal oxides by near-infrared spectroscopy using machine learning combined with chemometric techniques

Frontiers in Chemistry, Journal Year: 2024, Volume and Issue: 12

Published: June 3, 2024

The component analysis of raw meal is critical to the quality cement. In recent years, near-infrared (NIR) has been emerged as an innovative and efficient analytical method determine oxide content cement meal. This study aims utilize NIR spectroscopy combined with machine learning chemometrics improve prediction in Savitzky-Golay convolution smoothing applied eliminate noise interference for calcium carbonate ( CaCO3 ), silicon dioxide id="m2">SiO2 aluminum id="m3">Al2O3 ferric id="m4">Fe2O3 ) materials. Different wavelength selection techniques are used perform a comprehensive model, comparing performance several techniques. back-propagation neural network regression model based on particle swarm optimization algorithm was also optimize extracted screened feature wavelengths, checked evaluated using id="m5">Rp RMSE. conclusion, results indicate that combination ML great potential effectively materials highlight importance modeling By enabling more accurate determination materials, coupled meta-modeling revolutionize assurance practices manufacturing.

Language: Английский

---

Ensemble Neural Networks for the Development of Storm Surge Flood Modeling: A Comprehensive Review

Journal of Marine Science and Engineering, Journal Year: 2023, Volume and Issue: 11(11), P. 2154 - 2154

Published: Nov. 11, 2023

This review paper focuses on the use of ensemble neural networks (ENN) in development storm surge flood models. Storm surges are a major concern coastal regions, and accurate modeling is essential for effective disaster management. Neural network (NN) ensembles have shown great potential improving accuracy reliability such presents an overview latest research application NNs covers principles concepts ENNs, various architectures, main challenges associated with NN algorithms, their benefits forecasting accuracy. The part this pertains to techniques used combine mixed set predictions from multiple combination these models can lead improved accuracy, robustness, generalization performance compared using single model. However, generating also requires careful consideration trade-offs between model diversity, complexity, computational resources. must balance factors achieve best performance. insights presented particularly relevant researchers practitioners working regions where critical.

Language: Английский

---