How well can we measure the ocean's mean dynamic topography from space?

R. J. Bingham*, K. Haines, D. J. Lea

*Corresponding author for this work

Research output: Contribution to journalArticle (Academic Journal)peer-review

16 Citations (Scopus)
419 Downloads (Pure)


Recent gravity missions have produced a dramatic improvement in our ability to measure the ocean's mean dynamic topography (MDT) from space. To fully exploit this oceanic observation, however, we must quantify its error. To establish a baseline, we first assess the error budget for an MDT calculated using a 3rd generation GOCE geoid and the CLS01 mean sea surface (MSS). With these products, we can resolve MDT spatial scales down to 250 km with an accuracy of 1.7 cm, with the MSS and geoid making similar contributions to the total error. For spatial scales within the range 133-250 km the error is 3.0 cm, with the geoid making the greatest contribution. For the smallest resolvable spatial scales (80-133 km) the total error is 16.4 cm, with geoid error accounting for almost all of this. Relative to this baseline, the most recent versions of the geoid and MSS fields reduce the long and short-wavelength errors by 0.9 and 3.2 cm, respectively, but they have little impact in the medium-wavelength band. The newer MSS is responsible for most of the long-wavelength improvement, while for the short-wavelength component it is the geoid. We find that while the formal geoid errors have reasonable global mean values they fail capture the regional variations in error magnitude, which depend on the steepness of the sea floor topography.

Original languageEnglish
Pages (from-to)3336-3356
Number of pages21
JournalJournal of Geophysical Research: Oceans
Issue number6
Early online date4 Jun 2014
Publication statusPublished - Jun 2014


  • mean dynamic topography
  • geostrophic ocean currents
  • error analysis
  • mean sea surface


Dive into the research topics of 'How well can we measure the ocean's mean dynamic topography from space?'. Together they form a unique fingerprint.

Cite this