An Oceanic Mixed Layer Model Suitable for Climatological Studies Results over Several Years of Simulation

Abstract

A simple version of the Niiler and Kraus (1977) mixed layer model has been tested by simulating the upper ocean evolution at Ocean Station Papa for several years. As suggested by the authors themselves, the weakest point of this model is the parameterization of the turbulent dissipation. Consequently, the performed simulations show that the model’s wind-mixing efficiency constant (m) has to be tuned to the observations and exhibits seasonal and interannual changes. In an attempt to solve this problem a new parametric formulation of the dissipation is proposed. The resulting model (CMO) is calibrated independently of the observations to be simulated. When compared to the Niiler-Kraus model on a four-year simulation of the mixed layer changes at Station Papa, CMO presents the best overall agreement with the observations. The maximum error in the CMO-predicted monthly-mean sea surface temperature is 0.5 K. Nevertheless CMO, like the Niiler-Kraus model, systematically overestimates the sea surface temperature during spring and summer and underestimates it during fall and winter. Some explanations for this are suggested.