GreenSurge: an efficient additive model for predicting storm surge induced by tropical cyclones DOI
Beatriz Pérez-Díaz, Laura Cagigal, Sonia Castanedo

et al.

Coastal Engineering, Journal Year: 2024, Volume and Issue: unknown, P. 104691 - 104691

Published: Dec. 1, 2024

Language: Английский

Physics Informed Neural Network Modelling for Storm Surge Forecasting — A Case Study in the Bohai Sea, China DOI
Zhicheng Zhu, Zhifeng Wang, Changming Dong

et al.

Coastal Engineering, Journal Year: 2024, Volume and Issue: unknown, P. 104686 - 104686

Published: Dec. 1, 2024

Language: Английский

Citations

1

Model of Storm Surge Maximum Water Level Increase in a Coastal Area Using Ensemble Machine Learning and Explicable Algorithm DOI Creative Commons
Kun Sun, Jiayi Pan

Earth and Space Science, Journal Year: 2023, Volume and Issue: 10(12)

Published: Dec. 1, 2023

Abstract This study proposes a novel, new ensemble model (NEM) designed to simulate the maximum water level increases caused by storm surges in frequently cyclone‐affected coastal of Hong Kong, China. The relies on and data spanning 1978–2022. NEM amalgamates three machine learning algorithms: Random Forest (RF), Gradient Boosting Decision Tree (GBDT), XGBoost (XGB), employing stacking technique for integration. Six parameters, determined using Recursive Feature Elimination algorithms (RF‐RFE), are used as input features NEM. These parameters nearest wind speed, gale distance, air pressure, minimum pressure drop within 24 hr, large radius. Model assessment results suggest that exhibits superior performance over RF, GBDT, XGB, delivering high stability precision. It reaches coefficient determination ( R 2 ) up 0.95 mean absolute error (MAE) fluctuates between 0.08 0.20 m test set. An interpretability analysis conducted SHapley Additive exPlanations (SHAP) method shows distance speed most significant predicting peak during surges. this could provide practical implications predictive models concerning findings present essential tools mitigation disasters improvement marine disaster warning systems.

Language: Английский

Citations

2

High-Accuracy and Fast Calculation Framework for Berthing Collision Force of Docks Based on Surrogate Models DOI Creative Commons
Haikun Zeng, Ruihu Zhu, Qiming Wang

et al.

Journal of Marine Science and Engineering, Journal Year: 2024, Volume and Issue: 12(6), P. 898 - 898

Published: May 28, 2024

The accurate prediction of the collision force magnitude resulting from ship berthing on docks is crucial for rationality and safety dock structural design. This paper presents a novel framework calculation ships (CBCF), which integrates field data, finite element models, surrogate models. Based data analysis, constructs compares four models with low sample requirements, ultimately selecting optimal model predicting force. Furthermore, sensitivity analysis parameters conducted based selected model, followed by comparison various methods used prediction. results illustrate effectiveness proposed in replacing rapid Comparison existing also underscores advantages framework, including high accuracy, exceptional efficiency. In summary, this study not only introduces precise swift force, but it offers valuable insights into prevention wharf accidents facilitates rational safe design structures.

Language: Английский

Citations

0

Influence of Grid Resolution and Assimilation Window Size on Simulating Storm Surge Levels DOI Creative Commons
Xin Bi,

Wenqi Shi,

Junli Xu

et al.

Journal of Marine Science and Engineering, Journal Year: 2024, Volume and Issue: 12(7), P. 1233 - 1233

Published: July 22, 2024

Grid resolution and assimilation window size play significant roles in storm surge models. In the Bohai Sea, Yellow East China influence of grid on simulating levels was investigated during Typhoon 7203. order to employ a more realistic wind stress drag coefficient that varies with time space, we corrected model using spatial distribution coefficient, which inverted data method based linear expression Cd = (a + b × U10) 10−3. Initially, two resolutions 5′ 10′ were applied numerical adjoint model. It found different is almost negligible. But model, root mean square (RMS) errors between simulated observed under 11.6 cm 15.6 cm, average PCC WSS values for 10 tidal stations changed from 89% 92% E3 93% 96% E4, respectively. The results indicate finer can yield closer consistency simulation observations. Subsequently, effects sizes 6 h, 3 2 1 h evaluated an resolution. show RMS 10.6 9.6 9.3 four sizes. particular, at RuShan station 3.9 10.2 reduction 61.76%. E4 E7 separately showed increases, 85% 98% 99%. These demonstrate when smaller, level observation. Further, are observation smaller temporal

Language: Английский

Citations

0

Ensemble Tidal Prediction Scheme by Combining Harmonic Analysis and Meteorological Predictive Module DOI
Rui Wang, Jianchuan Yin, Dongxing Xu

et al.

Communications in computer and information science, Journal Year: 2024, Volume and Issue: unknown, P. 299 - 313

Published: Sept. 21, 2024

Language: Английский

Citations

0

Cyclone Intensity Estimation Using INSAT-3D IR Imagery DOI

D. Vasanthi,

Adityan Jothi,

Sivasakthi Thanigainathan

et al.

Published: Aug. 22, 2024

Language: Английский

Citations

0

Regional modelling of extreme sea levels induced by hurricanes DOI Creative Commons
Alisée A. Chaigneau, Melisa Menéndez, Marta Ramírez-Pérez

et al.

Natural hazards and earth system sciences, Journal Year: 2024, Volume and Issue: 24(11), P. 4109 - 4131

Published: Nov. 27, 2024

Abstract. Coastal zones are increasingly threatened by extreme sea level events, with storm surges being among the most hazardous components, especially in regions prone to tropical cyclones. This study aims explore factors influencing performance of numerical models simulating Atlantic region. The maxima, durations, and time evolutions surge events evaluated for four historical hurricanes against tide gauge records. Advanced Circulation (ADCIRC) Nucleus European Modelling Ocean (NEMO) ocean compared using similar configurations terms domain, bathymetry, spatial resolution. These then used perform sensitivity experiments on oceanic atmospheric forcings, physical parameterizations wind stress, baroclinic/barotropic modes. NEMO ADCIRC demonstrate abilities induced hurricanes. Storm simulated ERA5 reanalysis forcing generally more accurate than those parametric inclusion baroclinic processes improves amplitudes at some coastal locations, such as along southeastern Florida peninsula (USA). However, exploring different implementations stress interactions surges, tides, mean have shown minimal impacts hurricane-induced surges.

Language: Английский

Citations

0

GreenSurge: an efficient additive model for predicting storm surge induced by tropical cyclones DOI
Beatriz Pérez-Díaz, Laura Cagigal, Sonia Castanedo

et al.

Coastal Engineering, Journal Year: 2024, Volume and Issue: unknown, P. 104691 - 104691

Published: Dec. 1, 2024

Language: Английский

Citations

0