Enhanced methods of ocean wave spectra and sea state parameter estimation from X-band marine radar data

Al-Habashneh, Al-Abbass Y. (2018) Enhanced methods of ocean wave spectra and sea state parameter estimation from X-band marine radar data. Doctoral (PhD) thesis, Memorial University of Newfoundland.

[English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (34MB)

Abstract

Due to the ocean's importance in human lives, researchers have been studying the ocean and developing systems to estimate its state since the 19th century. During the last three decades, remote sensing of the ocean surface using X-band marine radars has emerged as a reliable tool to estimate ocean wave spectra and sea state parameters such as mean wave period and direction and significant wave height. The purpose of this thesis is to develop methods that produce accurate and reliable estimates of ocean wave spectra using X-band marine radar data. The approach taken in this thesis is to determine the sources of ocean wave spectra estimation error in existing methods and then to develop new methods that minimize those errors. In this thesis, four sources of error are addressed: the dependency of spectra estimation on the orientation of the analysis windows; the effect of the radar sampling process; the effect of the scan conversion process; and the accuracy of surface current estimation. The azimuthal location of the X-band radar data analysis window affects the estimation of ocean wave spectra. it has been reported in the literature, and supported by our results, that using the up-wave directions for analysis windows produces higher signal to noise ratios and hence more accurate ocean wave spectra estimates. In order to minimize error due to dependency on the orientation of the analysis windows, a new method referred to as the Adaptive Recursive Positioning Method (ARPM) is proposed. The ARPM is a recursive approach that dynamically determines the optimal number of analysis windows and their corresponding orientation toward the up-wave directions. Second, in this thesis, it has been demonstrated that the sampling process of the ocean surface by X-band marine radar during data collection significantly affects the estimation of ocean save spectra from X-band marine radar data. Therefore, a method referred to as the Inverse Sampling Averaging Filter (ISAF) is proposed to mitigate the effect of the radar sampling process of the ocean surface on the ocean wave spectra estimation using X-band marine radars. ISAF was designed based on a novel understanding of the radar sampling process to involve an averaging process or low pass filtering of the ocean wave spectra. Third, in this thesis, a method referred to as the Polar Fourier Transform (PFT) is proposed to eliminate the distortion presented by the scan conversion process to the estimated wave spectra. Unlike the existing methods which use the Cartesian Fourier Transform (CFT) to acquire the ocean wave spectra, the PFT method is designed to apply a Fourier-type transformation on the radar data in its native format, which is sampled in the polar coordinates, without the need for the intermediate stage of scan conversion used to map the data into Cartesian coordinates. The performance of the proposed methods, the ARPM,ISAF and PFT, are individually validated by comparing their ocean wave spectra estimates to those acquired using the existing methods with respect to ground truth wave spectra acquired using a wave rider buoy. Furthermore, the proposed methods were also combined together to seek further enhancement. The wave spectra estimation results from different combinations of the proposed methods were validated in comparison to the ground truth data. Finally, a new method to estimate surface current using X-band marine radar is proposed. This method is referred to as the Hybrid Least Squares (HLS) method. The HLS combines two existing approaches: the Iterative Least Squares (ILS) method and the Normalized Scalar Product (NSP). The HLS is designed to inherit the short computational time of ILS and the high reliability of NSP. To validate its accuracy and reliability, the proposed HLS method was applied on a number of simulated X-band marine radar image sets and the results were compared to the estimates acquired using the ILS and the NSP.

Actions (login required)

View Item