Abstract :
Sea level rise poses increasing risks to coastal regions, highlighting the need for accurate and reliable forecasting methods. This study proposes a probabilistic sea level forecasting framework by integrating a Long Short-Term Memory (LSTM) model with the Moving Block Bootstrap (MBB) technique. The LSTM model is used to capture nonlinear temporal dependencies in sea level time-series data, while the bootstrap approach is employed to quantify prediction uncertainty through probabilistic forecasting. The LSTM model achieved high deterministic prediction accuracy with an MSE of 2.11 × 10!”, RMSE of 0.00459, MAE of 0.00356, and MAPE of 0.34%. The proposed hybrid MBB–LSTM model generates probabilistic forecasts with a 95% confidence interval, resulting in an MSE of 0.01155, RMSE of 0.10749, MAE of 0.08370, and MAPE of 8.99%. Forecast results indicate relatively stable sea level variability until 2026 with an estimated rising trend of approximately 7.44 mm per year. The proposed hybrid framework provides a more informative prediction approach by combining deep learning with bootstrap-based uncertainty estimation, which is valuable for coastal risk assessment and climate adaptation planning.
Keywords :
Long Short-Term Memory, Moving Block Bootstrap, Probabilistic forecasting, Sea level prediction, Sea level rise., Time series analysisReferences :
- Griggs and B. Reguero, “Coastal flooding and sea-level rise under climate change,” Ocean & Coastal Management, 2021.
- Yanqiong et al., “Nonlinear climate dynamics and ocean variability,” Climate Dynamics, 2024.
- Bernier et al., “Extreme sea-level events under climate change scenarios,” Ocean & Coastal Management, 2024.
- Yin et al., “Climate change impacts on extreme sea-level variability,” Water, vol. 17, no. 3, 2025.
- Omid et al., “Deep learning approaches for nonlinear environmental time-series prediction,” Environmental Modelling & Software, 2023.
- Granata and F. Di Nunno, “Importance of tidal prediction for coastal infrastructure,” Coastal Engineering Journal, 2021.
- Kartawijaya et al., “LSTM-based tidal forecasting compared with ARIMA models,” Marine Data Science Journal, 2024.
- Yadav and Thakkar, “LSTM applications in oceanographic time-series forecasting,” Ocean Engineering, 2024.
- Almaliki et al., “Hybrid TCN–LSTM model for short- and long-term tidal prediction,” Environmental Modelling & Software, 2025.
- Liu et al., “LSTM–DNN model for sea-level estimation using GNSS-R data,” Remote Sensing, 2024.
- Masri, Saepudin, and Adytia, “LSTM and BiLSTM for non-tidal sea-level fluctuation prediction,” Marine Informatics Journal, 2024.
- Ambadar et al., “LSTM-based tidal prediction in Indonesian coastal regions,” Journal of Applied Informatics and Computing, 2025.
- Kang et al., “Oversampling techniques for imbalanced data learning,” Pattern Recognition Letters, 2024.
- Zhu et al., “Oversampling strategies for LSTM time-series classification,” IEEE Access, 2020.
- Nguinabé et al., “Bootstrap-based predictive distributions for LSTM time-series models,” Machine Learning for Environmental Data, 2025.
- Chawishborwornworng et al., “Bootstrap methods for epistemic uncertainty estimation in time series,” Applied Statistics, 2024.
- R. Künsch, “The jackknife and the bootstrap for general stationary observations,” Annals of Statistics, 1989.
- N. Lahiri, Resampling Methods for Dependent Data. Springer, 2003.
- Bühlmann, “Bootstraps for time series,” Statistical Science, 2002.
- D. N. Politis and H. White, “Automatic block-length selection for dependent bootstrap,” Econometric Reviews, 2004.

