Anisotropy of the sea surface height wavenumber spectrum from altimeter observations

Abstract

In this paper, the zonal and meridional sea surface height (SSH) wavenumber spectra are systematically calculated using along-track and gridded altimeter products, and the slopes of the SSH wavenumber spectra over the mesoscale band, which is defined by the characteristic length scale of mesoscale signals, are estimated. The results show that the homogeneous spectral slopes calculated from the along-track and gridded altimeter datasets have a similar spatial pattern, but the spectral slopes from gridded altimeter data are generally steeper than that from the along-track data with an averaged difference of 1.5. Significant differences are found between the zonal and meridional spectra, which suggest that SSH wavenumber spectra are indeed anisotropic. Furthermore, the anisotropy exhibits strong regional contrast: in the equatorial region, the zonal spectrum is steeper than its corresponding meridional spectrum, while in the eastward-flowing high EKE regions the meridional spectrum is steeper than its zonal counterpart. The anisotropy of SSH wavenumber spectral slopes implies that EKE distributes anisotropically in different directions, and this distribution is closely associated with the generation and nonlinear evolution of mesoscale movements.

Can the Surface Quasi‐Geostrophic (SQG) Theory Explain Upper Ocean Dynamics in the South Atlantic?

“…Both SQG and QG theories predict different exponents for the spectra of KE and SSH, which is an indicative of the region's dynamics. Although spectral slopes over the mesoscale range from along-track altimetry are consistent with the turbulence theories (Vergara et al, 2019;Wang et al, 2019) in highly energetic regions, the reasons for the flatter spectra found in low-energy regions still poses an ongoing debate (Richman et al, 2012;Sasaki et al, 2014;Uchida et al, 2017). In this context, improved theoretical framework should be pursued.…”

“…Wang et al. (2019) pointed out that these slopes are correlated with eddy kinetic energy (EKE), being steeper when the energy is high. Both studies found regional differences between the spectra, and Vergara et al.…”

“…Using along-track data, the authors identified flatter slopes (k −2 ) in the intertropical band, slopes close to k −11/3 at midlatitudes, and around k −5 in the subpolar regions, showing different dynamical regimes associated with the sea surface height signal. Wang et al (2019) pointed out that these slopes are correlated with eddy kinetic energy (EKE), being steeper when the energy is high. Both studies found regional differences between the spectra, and Vergara et al (2019) observed a marked seasonal variation in the spectral slopes.…”

Improving the Estimation of the Wavenumber Spectra From Altimeter Observations

“…Estimation of the SLA spectral slope α is usually a two-step procedure [8], [10], [11], [13], [14]. First, an estimation of the PSD is done.…”

“…To perform a reliable slope estimation, the PSD estimate has to have a variance sufficiently low. Therefore, the widely used spectral method is based on a cumulated (Welch) periodogram which consists of averaging a large number of periodograms of along-track SLA segments [8], [10]- [14]. This PSD averaging determines the spatial and temporal resolution of the resulting slope map, depending on the size of the area in latitude and longitude and on the time period from which the SLA signals are issued.…”