Long-Term Prediction of Significant Wave Height Based on SARIMA Model in the South China Sea and Adjacent Waters

Shaobo Yang,Tianliang Xia,Chongwei Zheng,Shanhua Duan,Linlin Fan,Zhenquan Zhang,Xingfei Li,Hongyu Li

doi:10.1109/access.2019.2925107

Abstract

Nowadays, the world is facing the dual crisis of the energy and environment, and renewable energy, such as wave energy, can contribute to the improvement of the energy structure of the world, enhance energy supply and improve the environment in the framework of sustainable development. Long-term prediction of the significant wave height (SWH) is indispensable in SWH-related engineering studies and is exceedingly important in the assessment of wave energy in the future. In this paper, the spatial and temporal characteristics of wave energy in the South China Sea (SCS), and adjacent waters are analyzed. The results show that there are abundant wave energy resources in the waters around the Taiwan Strait, the Luzon Strait, and the north part of the SCS with annual average SWH (SWH) of over 1.4 m and obvious increasing trend. Then, the SARIMA approach considers the relationship between the current time and the values, residuals at some previous time and the periodicity of the SWH series are proposed to forecast the SWH in the SCS and adjacent waters. The results obtained are promising, showing good performance of the prediction of monthly average SWH in the SCS and adjacent waters.

Highlights

As the second decade of the 21st century passing away, the two problems, crisis of resource and environmental pollution, are approaching gradually
We can see that the significant wave height (SWH) in the South China Sea (SCS) and adjacent waters reaches its peak in winter (January, February and December), and most of the areas are above 1.2 m
Due to the complicated and stochastic behavior of ocean waves, long-term prediction of SWH is full of challenge

Summary

INTRODUCTION

As the second decade of the 21st century passing away, the two problems, crisis of resource and environmental pollution, are approaching gradually. Kamranzad et al employed ANN as a robust data learning method to forecast the wave height for the few hours in 2011 They evaluated the effects of different parameters using different models with various combinations of wind and wave parameters [14]. L. Cornejo-Bueno et al presented a hybrid Grouping Genetic Algorithm-Extreme Learning Machine approach (GGA-ELM) for SWH and wave energy flux prediction and obtained good results. Cornejo-Bueno et al presented a hybrid Grouping Genetic Algorithm-Extreme Learning Machine approach (GGA-ELM) for SWH and wave energy flux prediction and obtained good results This approach can solve feature selection problems and may be applied to alternative regression approaches [18]. The model for the long-term and regional prediction of wave energy resources is proposed, which will contribute to the construction of future WECs network and provide a significant tool for optimizing operating costs and improving reliability.

STUDY AREA AND METJODLOGY

PREDICTION METHOD

PREDICTED RESULTS AND DISCUSSION

CONCLUSIONS