Abstract
Significant wave height (SWH) is a fundamental concept in marine-related applications, activities, and renewable wave energy. The sea state is characterized by the SWH, and real-time ocean operations suffer from missing data. Further, expensive deployment and maintenance operations, physical constraints, or both hamper the design of dense SWH buoy networks. In this study, the identification of optimal buoy locations is performed for a specific number of total buoy stations. These optimal locations are then extrapolated to obtain the complete state SWH network data. This extrapolation process is accomplished using the QR decomposition with a column pivoting algorithm, which is executed based on a data-driven approach that utilizes empirical orthogonal function (EOF) analysis. The monitoring network is composed of 15 buoys on the West Coast of the US in the Pacific Ocean. The Nash–Sutcliffe coefficient of efficiency (CE) and mean square error (MSE) diagnostic metrics are utilized for the model performance assessment. Based on the diagnostic metrics, the EOF–QRP model performance is at an acceptable level when two best QR algorithm identified buoys are used. The performance level increases with the total number of stations used. The model performed very well with six buoys’ data according to error metrics. The EOF–QRP model advocated in this study has proved successful when identifying the minimum number of buoys and their locations and provided a general promising framework for optimal station network design.