Uncertainties in the mean ocean dynamic topography before the launch of the Gravity Field and Steady-State Ocean Circulation Explorer (GOCE)

In anticipation of the future observations of the gravity mission Gravity Field and Steady-State Ocean Circulation Explorer (GOCE), the present-day accuracy of mean dynamic topography (MDT) is estimated from both observations and models. A comparison of five observational estimates illustrates that RMS differences in MDT vary from 4.2 to 10.5 cm after low-pass filtering the fields with a Hamming window with wavenumber N = 120 (corresponding to an effective resolution of 167 km). RMS differences in observational MDT reduce to 2.4-8.3 cm for N = 15 (1334 km). Differences in data sources (geoid model, in situ data) are mostly visible in the small-scale oceanic features, while differences in processing (filtering, inverse modeling techniques) are reflected at larger scales. A comparison of seven different numerical ocean models demonstrates that model estimates differ mostly in western boundary currents and in the Antarctic Circumpolar Current. RMS differences between modeled and observed MDT are at best 8.8 cm for N = 120, and reduce to 6.4 cm for N = 15. For models with data assimilation, a minimal RMS difference of 6.6 cm (N = 120) to 3.4 cm (N = 15) is obtained with respect to the observational MDTs. The reduction of differences between MDTs with increasing filtering scales is smaller than expected. While it is expected that GOCE will improve MDT estimates at small spatial scales, improvement of mean sea surface estimates from satellite altimetry may be needed to improve MDT estimates at larger scales.