High-Resolution Spectral Estimation of HF Radar Data for Current Measurement Applications

Abstract

This paper presents a comparative study of high-resolution methods for high-frequency radar current mapping. A z-domain transformation and auxiliary z-domain manipulation of the autoregressive method is proposed for this comparison. A Weibull distribution test is recommended to justify the Rayleigh distribution of the sea clutter for quality control. Upon the power spectrum estimation, a conventional centroid method and a new symmetric-peak-sum method for the identification of current Doppler shift are proposed as another comparison. HF radar data were collected over the period from November 2012 to August 2013 at Placentia Bay, Newfoundland, Canada, and were compared with measurements from an acoustic Doppler current meter. This comparison is used to study the utility of high-resolution spectrum estimation and Bragg identification methods for surface current mapping. Results show promising use of these methods in different current scenarios and suggest combined applications to improve accuracy.

High-frequency surface wave radar performance analysis for CA-CFAR algorithm in Weibull-distributed clutter

“…Sea clutter is asymmetrically spread across Doppler zero, peaking at specific resonance frequencies with some smaller peaks at adjacent frequencies. Due to this pattern with frequent peaks irregularly scattered around the Doppler dimension, the overall distribution of sea echo is better represented by heavy-tailed distributions than the traditionally used Rayleigh, 4,5 such as Weibull, log-normal, Pareto, or K-distribution. [6][7][8] In particular, the Weibull distribution is a suitable choice for modeling of general radar clutter signals, and Bucciarelli 9 showed that a good performance can be obtained in this scenario.…”

High-Frequency Radar Ocean Current Mapping at Rapid Scale with Autoregressive Modeling

“…In particular, the use of Autoregressive (AR) modeling was applied by [16] within the framework of iceberg detection in HFR coastal measurements. Later on, [17] first proposed an AR approach in the context of oceanographic HFR, whose interest has been confirmed through subsequent studies [18,19,20]. To date, however, these methods have been confined to specific applications and have not yet earned their letters of nobility in the common HFR signal processing toolbox.…”