Two‐Dimensional Spectra of Radar Returns From Sea: 2. Analysis of the Group Line From Experimental Data

Abstract

Wavenumber-frequency spectra are obtained by performing a two-dimensional Fourier transform of range-time normalized radar cross-section (NRCS) or Doppler velocity maps. In such diagrams, some energy is present at low space-time frequencies, resulting from the nonlinear behavior of the measured quantity related to the sea surface geometry. This feature is called the group line, since for a narrow-band wave packet the energy is concentrated along a straight line with group velocity as slope. Which physical nonlinear process generates the group line remains an open question. Breaking waves have been proposed as the most probable contributor. However, numerical simulations from weakly nonlinear surfaces without breaking events and experiments performed at low winds also provide such feature. In a companion paper, a theoretical and numerical analysis has permitted to predict the energy distribution of the group line depending on the kind of nonlinearity. It provides some means to characterize a group line in a rigorous way. In this paper, it is used to analyze the group lines derived from the experimental MARLENE data. The group lines computed from the backscattering amplitude behave as the one of the square of sea surface slopes. The analysis of the Doppler velocities provides similar results, which significantly differ from what is expected if breaking waves are the main contributors and do break at velocities reported in the literature. Our results suggest that the group line mainly reflects the asymptotic behavior of the scattering amplitude at grazing incidence, of which the leading nonlinear term is proportional to the square of surface slope.

Two‐Dimensional Spectra of Radar Returns from Sea: 1. Theoretical and Numerical Study of the Group Line

“…Since the computations performed here are restricted to weak hydrodynamic nonlinearities, we have not been able to exhibit group lines with such high slopes. This will be investigated in a companion paper focusing on in situ data analysis Platzer, Fabbro, & Saillard, . (submitted).…”