130 likes | 310 Views
How well do we know the mean ocean dynamic topography?. Femke C. Vossepoel (IMAU/SRON), Peter Jan van Leeuwen (IMAU), and Radboud Koop (SRON). 3rd International GOCE Workshop, Frascati, 6-8 November 2006. Comparison of observational and modeled ocean mean dynamic topographies. Observed:
E N D
How well do we know the mean ocean dynamic topography? Femke C. Vossepoel (IMAU/SRON), Peter Jan van Leeuwen (IMAU), and Radboud Koop (SRON) 3rd International GOCE Workshop, Frascati, 6-8 November 2006
Comparison of observational and modeled ocean mean dynamic topographies • Observed: • Le Grand • Rio05 • Maximenko-Niiler • Naeije • Chambers-Zlotnicki • Modeled: • ORCA • MPI-OM • OCCAM • POP • Hycom • ECCO • NCOM RIO05 (see Rio&Hernandez, 2004)
Observational MDT error sources • MSS • Time-averaging periods • Interpolation errors • Observational errors
Low-pass filtering with Hamming filter • N=15 (1334 km) • N=30 (667 km) • N=60 (334 km) • N=120 (167 km)
Possible sources of error in modeled MDTs • Averaging period • Errors in atmospheric forcing • Spatial resolution • Mixing parameterization
Conclusions • RMS differences between MDTs in the order of 5-10 cm at N=15, still 5 cm at N=120 • Reduction differences is smaller than expected • Results obscured by differences in: • Mean Sea Surface processing • Averaging period observations and models • Atmospheric forcing (models) • Model resolution and formulation • …. Details: http://www.phys.uu.nl/~vossepl/DraftVossepoel.pdf, under revision for JGR
Ongoing and future work • Regional model for Agulhas: study impact of current strength on ring shedding • Regional model for ACC • Assimilate SSH using different geoid models • Investigate importance bottom topography (pressure, friction)