Rosemary Morrow (LEGOS, France); Isabelle Pujol (CLS, France); Gerald Dibarboure (CNES, France); Clement Ubelmann (Ocean Next, France)

We use three years of Sea Surface Height (SSH) observations over the World Ocean to analyze its spectral characteristics: spectral slope and noise levels. Our analyses are based on previous works that focused on the spectral slope characterization through a linear regression over a geographically dependent wavelength range, corresponding to the mesoscale wavelength range. In our current analysis, we adopt a methodology that allows for the characterization of the spectral slope by optimal fit of a linear function, considering the statistical uncertainty of the observed spectra. This is particularly important in the wavelength range of the instrumental noise (15-30 km wavelength): the noise level influence the spectral shape over the geophysical spatial scales. The observed values of spectral slope are within the range of previous works however we observe a higher spatial variability in regions where the eddy kinetic energy varies sharply, with spectral slope values that drop significantly in low energy regions. Noise level distribution is also within the range of previous results, although we observe higher values in the mid latitudes (10 to 15%) and intertropical band. In addition, our current approach allows us to analyze (when present) the break in the mesoscale spectral slope that occurs at wavelengths shorter than 100 km, partly due to the internal tides and their energy cascade to smaller scales, but also other dynamical signal and instrumental errors. Preliminary results on this topic are also presented.