A Puzzling Disagreement between Observations and Numerical Models in the Central Gulf of Mexico

Abstract

wo large, independent sets of direct observations in the central Gulf of Mexico show a mean near-surface flow of ~10 cm s?1 to the west, concentrated in the northern and southern Gulf. Numerical models that the authors have examined do not produce this mean westward flow. The observed speeds appear to be almost an order of magnitude larger than the estimated errors; this paper studies the observations to estimate carefully the possible errors involved and compares the observations with model results. The flow to the west in the southern Gulf is presumably wind driven on the shallow parts of the shelf, and, in slightly deeper water at the outer edges of the shelf, is possibly the result of southward Sverdrup interior flow driven by the negative curl of the wind stress. In another possibly related issue, long-term deep current-meter observations in the northern Gulf at ~1000 m and below find flow to the west, whereas some models find flow to the east. The flow proposed here assumes a mean flow to the west above roughly 300 m, with a required return flow in deep water. The difference between the deep observations and the models will produce a slope of pressure surfaces of the opposite sign below 1000 m, reversing the direction of upper-layer geostrophic flow in the models.