Session: 12-04-02 Coastal Hazards - Tsunamis and Storm Surges II

Paper Number: 130308

130308 - Typhoon Wave Prediction Using Long Short-Term Memory Networks for Offshore Windfarm on Western Coast of Taiwan 

There are abundant potential wind energy resources in the Taiwan Strait. However, Taiwan would be invaded by an average of 3-4 typhoons every summer and autumn. The gusts of typhoons may bring severe threats to wind turbines. Moreover, the extreme wave conditions also impact the scheduling of offshore wind farm maritime operations and maintenance. Therefore, developing an accurate and efficient typhoon wave prediction model is important for improving the efficiency of offshore windfarm management.

In the earlier studies, short-lead-time (i.e., 1 to 3 hours) typhoon wave prediction models were developed for the Taiwan coastal area. These models were constructed by Backward Propagation Neural Network (BPNN) with local meteorological information. Sufficient prediction lead-time is essential for early warning and response to offshore windfarm during typhoon events. Furthermore, past research on typhoon waves along the western coast of Taiwan often presented an underestimated tendency due to the structure of the typhoon being destroyed by the Central Mountain Range.

The purpose of this study is to establish a novel long-lead-time typhoon wave prediction model using deep learning Long Shor-Term Memory (LSTM) networks while carefully considering typhoon parameters. The basic concept of LSTM is to utilize the memory cell to capture the features or vectors of time-related data, significantly enhancing prediction accuracy. Similar to previous research, the results of LSTM demonstrate high consistency with in-situ data for 1-hour lead time (i.e., the correlation coefficient is up to 0.95). For longer lead time (e.g., 6 hours), the LSTM networks exhibit a prominent improvement in learning and generalizing capability than shallow learning methods (e.g., BPNN, SVR). The correlation coefficients for training and validation reach 0.85 and 0.80, respectively.

Presenting Author: Tai-Wen Hsu Department of Harbor and River Engineering, National Taiwan Ocean University

Presenting Author Biography: Educational Background：
National Cheng Kung University, Ph.D. in Hydraulic and Ocean Engineering
Social Service：
President, National Taiwan Ocean University
(2020.08.01~Present)
Chair Professor, Department of Harbor and River Engineering, National Taiwan Ocean University (2016.08.01~Present)
Area of Specialization：
Nearshore Hydrodynamics、Wind wave modeling forecasting for deep and shallow water regions、Coastal development and conservation、Ocean Energy and Strategy、Wind wave modeling、Coastal and topographic changes

Authors:

Wei-Ting Chao Center of Excellence for Ocean Engineering, National Taiwan Ocean University
Chih-Chieh Young Department of Marine Environmental Informatics, National Taiwan Ocean University
Tai-Wen Hsu Department of Harbor and River Engineering, National Taiwan Ocean University