Observations of the broadening and coherence of MF/lower HF surface-radar ocean echoes

Abstract

Known broadening mechanisms for the first-order Bragg peaks in high-frequency (HF: 3-30 MHz) and very-high-frequency (VHF: 30-300 MHz) Doppler spectra of radar echoes from the sea surface are summarized. Observations of medium-frequency (MF: 0.3-3 MHz)/lower HF sea echoes were made with a surface-wave phased-antenna array transmitting Gaussian pulses of width /spl ap/40 and 70 /spl mu/s at frequencies 1.98, 3.84, and 5.80 MHz (sensitive to sea waves of length 76, 39, and 26 m, respectively). Some of the broadening of first order Bragg peaks in Doppler spectra was consistent with the effects of nonuniform surface-current fields across radar footprints. At these MF frequencies, however, the dominant cause of first-order broadening was the spread in Doppler shifts caused by the phase speeds of first-order gravity sea waves being modified by the comparatively shallow radar footprints. Second-order sea echoes were usually not observed with good signal-to-noise ratio because the sea wave heights were not a sufficient fraction of the radio wavelengths. However, data were obtained in which bathymetrically amplified, higher order sea echoes also contributed to the broadening. Examples of this confused data are presented to illustrate that the interpretation of data from a coastal surface-wave radar at MF is difficult, hence limiting the quality of wave heights estimated by inverting second-order sea echoes. Furthermore, the phase shifts of gravity sea waves caused by propagation through nonuniform surface-current and bathymetry fields helps to explain the considerable incoherence of sea echoes in coastal regimes. To this end, hundreds of Doppler spectra were analyzed from records taken during 2-3 day periods and across a range of frequencies and sea states. The results show that the coherence of Bragg peaks decreases with frequency and with water depth and confirm that they integrate less coherently than a stationary tone. There is also a suggestion that coherence depends on wave age. Consequently, on practical time scales (/spl sim/1000 s) incoherent spectral averaging is more effective at revealing spectral information otherwise buried in noise (e.g., second-order echoes containing sea-state information).