An improved sparrow search algorithm and CNN-BiLSTM neural network for predicting sea level height

Xiao Li,Shijian Zhou,Fengwei Wang,Laiying Fu

DOI: https://doi.org/10.1038/s41598-024-55266-4

IF: 4.6

2024-02-25

Scientific Reports

Abstract:Accurate prediction of sea level height is critically important for the government in assessing sea level risk in coastal areas. However, due to the nonlinear, time-varying and highly uncertain characteristics of sea level change data, sea level prediction is challenging. To improve the accuracy of sea level prediction, this paper uses a new swarm intelligence algorithm named the sparrow search algorithm (SSA), which can imitate the foraging behavior and antipredation behavior of sparrows, to determine optimal solutions. To avoid the algorithm falling into a local optimal situation, this paper integrates the sine–cosine algorithm and the Cauchy variation strategy into the SSA to obtain an algorithm named the SCSSA. The SCSSA is used to optimize the parameter values of the CNN-BiLSTM (convolutional neural network combined with bidirectional long short-term memory neural network) model; finally, a combined neural network model (named SCSSA-CNN-BiLSTM) is proposed. In this paper, the time series data of seven tidal stations located in coastal China are used for experimental analysis. First, the SCSSA-CNN-BiLSTM model is compared with the CNN-BiLSTM model to predict the time series data of SHANWEI Station. With respect to the training and test sets of data, the SCSSA-CNN-BiLSTM model outperforms the other models on all the evaluation metrics. In addition, the remaining six tide station datasets and five neural network models, including the SCSSA-CNN-BiLSTM model, are used to further study the performance of the proposed prediction model. Four evaluation indices including the root mean squared error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE) and coefficient of determination (R 2 ) are adopted. For six stations, the RMSE, MAE, MAPE and R 2 of SCSSA-CNN-BiLSTM model are ranged from 20.9217 ~ 27.8427 mm, 9.4770 ~ 17.8603 mm, 0.1322% ~ 0.2482% and 0.9119 ~ 0.9759, respectively. The experimental analysis results show that the SCSSA-CNN-BiLSTM model makes effective predictions at all stations, and the prediction performance is better than that of the other models. Even though the combination of SCSSA algorithm may increase the complexity of the model, indeed the proposed model is a new prediction method with good accuracy and robustness for predicting sea level change.

multidisciplinary sciences

What problem does this paper attempt to address?

The paper mainly addresses the following issues: ### Research Background and Objectives - **Research Background**: With global warming, sea level rise has become a major environmental issue worldwide. Accurate prediction of sea level changes is crucial for assessing the risks to coastal areas and addressing the socio-economic impacts of climate change. - **Main Challenges**: Due to the nonlinear, time-varying, and highly uncertain nature of sea level change data, predicting sea level poses significant challenges. ### Solution - **Algorithm Innovation**: An improved sparrow search algorithm (SCSSA) is proposed, which combines the sine cosine algorithm and Cauchy mutation strategy to avoid the problem of local optima. - **Model Construction**: SCSSA is used to optimize the parameter values of a combined model of convolutional neural network (CNN) and bidirectional long short-term memory neural network (BiLSTM), resulting in the SCSSA-CNN-BiLSTM joint neural network model. - **Experimental Validation**: Time series data from seven tide gauge stations along the coast of China were used for experimental analysis, comparing the prediction performance of the SCSSA-CNN-BiLSTM model with other models. ### Main Contributions 1. **Model Application**: The CNN-BiLSTM joint neural network model is applied to sea level time series prediction, and its performance is validated through experiments with data from multiple tide gauge stations, showing superiority over individual models. 2. **Parameter Optimization**: SCSSA is used to optimize parameters such as the learning rate, number of hidden layer nodes, and regularization coefficients of the CNN-BiLSTM model, effectively avoiding the irrationality brought by experience-based or heuristic parameter settings. 3. **Performance Validation**: The accuracy and robustness of the SCSSA-CNN-BiLSTM model in sea level prediction are validated by predicting sea level time series data measured at six tide gauge stations and comparing it with LSTM, CNN, BiLSTM, and CNN-BiLSTM models. ### Conclusion In summary, this paper addresses the important yet challenging task of sea level prediction by proposing a method that combines an improved sparrow search algorithm with a CNN-BiLSTM model. This approach effectively improves prediction accuracy and model robustness, providing new insights and technical support for research in related fields.

An improved sparrow search algorithm and CNN-BiLSTM neural network for predicting sea level height

Salinity Prediction Based on Improved LSTM Model in the Qiantang Estuary, China

Research on Short-Term Prediction of Typical Trial Sea Environment in China Based on LSTM Neural Network

Short- to Medium-Term Sea Surface Height Prediction in the Bohai Sea Using an Optimized Simple Recurrent Unit Deep Network

Estimating sea surface swell height using a hybrid model combining CNN, ConvLSTM, and FCN based on spaceborne GNSS-R data from the CYGNSS mission

Prediction of Salinity in Qiantang Estuary Based on Wavelet Neural Network Optimized by Particle Swarm Optimization

Predicting significant wave height in the South China Sea using the SAC-ConvLSTM model

Research on SVR Water Quality Prediction Model Based on Improved Sparrow Search Algorithm

Mid-Term Simultaneous Spatiotemporal Prediction of Sea Surface Height Anomaly and Sea Surface Temperature Using Satellite Data in the South China Sea

A Prediction Model of Significant Wave Height in the South China Sea Based on Attention Mechanism

Water quality prediction research based on SSA-LSTM model

Sea-water-level Prediction Via Combined Wavelet Decomposition, Neuro-Fuzzy and Neural Networks Using SLA and Wind Information

Improving wave height prediction accuracy with deep learning

A deep-learning model for rapid spatiotemporal prediction of coastal water levels

Multi-step ahead short-term predictions of storm surge level using CNN and LSTM network

Sea Level Prediction Based On The Long Time Series Satellite Observations

Improved tide level prediction model combined GA-BP neural networks and GNSS SNR data

Long-term prediction of sea surface chlorophyll-a concentration based on the combination of spatio-temporal features

Advancing Accuracy in Sea Level Estimation with GNSS-R: A Fusion of LSTM-DNN-Based Deep Learning and SNR Residual Sequences

A training strategy to improve the generalization capability of deep learning-based significant wave height prediction models in offshore China

Optimizing Back-Propagation Neural Network to Retrieve Sea Surface Temperature Based on Improved Sparrow Search Algorithm