Harley E. Hurlburt(1), Jay F. Shriver(1), and Ole Martin Smedstad(2)
(1) Naval Research Laboratory, Stennis Space Center, MS 39529
(2) Planning Systems, Inc., Stennis Space Center, MS 39529
A data-assimilative ocean model was used to investigate the ability of satellite altimetry to map ocean eddies as a function of eddy size and to determine the lower bound on eddy size that could be reliably depicted. A quantitative assessment was performed by comparing eddy center locations from the ocean model with an unusually clear ocean color image in the northwestern Arabian Sea. Twenty well-defined eddy centers were identified in the ocean color imagery that could be used in the assessment. The eddy diameters ranged from 25 to 300 km and half of them were small eddies, 25-75 km in diameter. The 1/32 degree global NRL Layered Ocean Model (NLOM) (4 km resolution at 20N) was used to assimilate altimeter track data from the ERS-2, GFO, and JASON 1 altimeters with the ocean model forecast as the first guess. The model without assimilation of altimeter data was used as a control run to help assess data/assimilation impact. The model with assimilation showed 70% reliability in mapping the small eddies vs 20% reliability without assimilation, a surprising and unanticipated result. These results have significant implications for nested finer resolution coastal region models, including the value of high-resolution boundary conditions from an outer model.