Atlantic sea surface height and velocity spectra inferred from satellite altimetry and a hierarchy of numerical simulations. Issue 6 (18th June 2016)

• Record Type: Journal Article
• Authors: Biri, Stavroula; Serra, Nuno; Scharffenberg, Martin G.; Stammer, Detlef
• Is Part Of: Journal of geophysical research. Volume 121:Issue 6(2016:Jun.)
• Page Start: 4157
• Page End: 4177
• Publication Date: 2016-06-18
• Subjects: ocean variability -- spectral analysis -- altimetry -- ocean modeling; Oceanography -- Periodicals; 551.4605
• DOI: 10.1002/2015JC011503
• Languages: English
• ISSNs: 2169-9275

Abstract

Abstract: Frequency and wavenumber spectra of sea surface height (SSH) and surface geostrophic velocity are presented, as they result for the Atlantic Ocean from a 23 year long altimeter data set and from a hierarchy of ocean model simulations with spatial resolutions of 16, 8, and 4 km. SSH frequency spectra follow a spectral decay of roughly f −1 on long periods; toward higher frequencies a spectral decay close to f −2 is found. For geostrophic velocity spectra, a somewhat similar picture emerges, albeit with flatter spectral relations. In terms of geostrophic velocity wavenumber spectra, we find a general relation close to k −3 in the high‐resolution model results. Outside low‐energy regions all model spectra come close to observed spectra at low frequencies and wavenumbers in terms of shape and amplitude. However, the highest model resolution appears essential for reproducing the observed spectra at high frequencies and wavenumbers. This holds especially for velocity spectra in mid and high latitudes, suggesting that eddy resolving ocean models need to be run at a resolution of 1/24° or better if one were to fully resolve the observed mesoscale eddy field. Causes for remaining discrepancies between observed and simulated results can be manifold. At least partially, they can be rationalized by taking into account an aliasing effect of unresolved temporal variability in the altimetric observations occurring on periods smaller than the 20 days Nyquist period of the altimetric data, thereby leading to an overestimate of variability in the altimetric estimates, roughly on periods below 100 days. Key Points: Altimeter data errors Spectral analysis Dynamics and modeling