The ability of a hydrodynamic model to reproduce the results of a dye release experiment conducted in a wide shelf sea environment was investigated with the help of the Massachusetts Institute of Technology general circulation model (MITgcm). In the field experiment a fluorescent tracer, Rhodamine WT, was injected into the seasonal pycnocline, and its evolution was tracked for two days using a towed undulating vehicle equipped with a fluorometer and a CTD. With a 50 m horizontal resolution grid, and with three different forcings initialized in the model (viz: tides, stationary current, and wind stress on the free surface), it was possible to replicate the dye patch evolution quite accurately. The mechanisms responsible for the enhancement of horizontal dispersion were investigated on the basis of the model results. It was found that enhancement of the dye dispersion was controlled by vertically sheared currents that, in combination with vertical diapycnal mixing, led to a substantial increase in the “effective” horizontal mixing. The values of “effective” horizontal mixing found from the model runs were in good agreement with those obtained from in-situ data, and the probable degree to which the observational techniques undersampled the dye patch was revealed.