Analysis and machine-learning-based prediction of beach accidents on a recreational beach in China DOI Open Access
Yuan Li, Jiqiang Tang, Chi Zhang

et al.

Anthropocene Coasts, Journal Year: 2024, Volume and Issue: 7(1)

Published: Dec. 30, 2024

Abstract Beachgoers are sometimes exposed to coastal hazards, yet comprehensive analyses of characteristics and potential factors for beach accidents rarely reported in China. In this study, information on was collected a recreational from 2004 2022 by searching the web or apps. The were therefore analysed terms age, gender, activity beachgoers. resolved environmental aspects meteorology, waves, tides, morphology. Results show that mainly occur summer, with highest occurrence afternoon evening. number male beachgoers is five times higher than females. 90% when at high-risk level rip currents, providing evidence accuracy risk map built previous study. Three machine learning models, i.e., Support Vector Machine, Random Forest, BP Neural Networks, trained predict accidents. performances these three algorithms evaluated precision, recall, F1 score. Machine Networks significantly outperform Forest prediction. predicting "safe" "dangerous" classes approximately 80% model. This paper provides preliminary study based accident prediction specific tourist beach. future, will be applied throughout mainland

Language: Английский

Analysis of Tidal Cycle Wave Breaking Distribution Characteristics on a Low-Tide Terrace Beach Using Video Imagery Segmentation DOI Creative Commons
Hang Yin, Feng Cai,

Hongshuai Qi

et al.

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(24), P. 4616 - 4616

Published: Dec. 10, 2024

Wave breaking is a fundamental process in ocean energy dissipation and plays crucial role the exchange between nearshore sediments. Foam, primary visible feature of wave areas, serves as direct indicator processes. Monitoring distribution foam via remote sensing can reveal spatiotemporal patterns breaking. Existing studies on processes primarily focus individual events or short timescales, limiting their effectiveness for regions where hydrodynamic are often represented at tidal cycles. In this study, video imagery from typical low-tide terrace (LTT) beach was segmented into four categories, including foam, using DeepLabv3+ architecture, convolutional neural networks (CNNs)-based model suitable semantic segmentation complex visual scenes. After training testing manually labelled dataset, which divided training, validation, sets based different time periods, overall classification accuracy 96.4%, with an 96.2% detecting foam. Subsequently, heatmap over cycle LTT generated. During cycle, density exhibited both alongshore variability, pronounced bimodal structure cross-shore direction. Analysis morphodynamical data collected field indicated that driven by variations. The key factor shaping profile morphology beaches. High-frequency monitoring further showed vary significantly levels, leading to diverse geomorphological features various locations.

Language: Английский

Citations

1
The combined effects of tide and storm waves on beach profile evolution DOI
Xiangming Cao, Jian Shi, Chi Zhang

et al.

Ocean Engineering, Journal Year: 2024, Volume and Issue: 299, P. 117416 - 117416

Published: March 11, 2024

Language: Английский

Citations

0
Analysis and machine-learning-based prediction of beach accidents on a recreational beach in China DOI Open Access
Yuan Li, Jiqiang Tang, Chi Zhang

et al.

Anthropocene Coasts, Journal Year: 2024, Volume and Issue: 7(1)

Published: Dec. 30, 2024

Abstract Beachgoers are sometimes exposed to coastal hazards, yet comprehensive analyses of characteristics and potential factors for beach accidents rarely reported in China. In this study, information on was collected a recreational from 2004 2022 by searching the web or apps. The were therefore analysed terms age, gender, activity beachgoers. resolved environmental aspects meteorology, waves, tides, morphology. Results show that mainly occur summer, with highest occurrence afternoon evening. number male beachgoers is five times higher than females. 90% when at high-risk level rip currents, providing evidence accuracy risk map built previous study. Three machine learning models, i.e., Support Vector Machine, Random Forest, BP Neural Networks, trained predict accidents. performances these three algorithms evaluated precision, recall, F1 score. Machine Networks significantly outperform Forest prediction. predicting "safe" "dangerous" classes approximately 80% model. This paper provides preliminary study based accident prediction specific tourist beach. future, will be applied throughout mainland

Language: Английский

Citations

0