A three-layer model of ocean sea surface skin temperature, warm layer and mixed layer (SWaM) for improved representation of the upper ocean warm layer in modeling and data assimilation

We present a numerically efficient three-layer model of the ocean, incorporating a surface skin layer, a warm layer, and a mixed layer. The model reproduces key processes that force local change in the upper ocean mixed layer in response to atmospheric fluxes and stresses at the surface, vertical mixing of the warm and mixed layers, and mass entrainment from the thermocline. Numerical efficiency of this model, as well as its capability to reproduce observed cycles, makes it an effective tool for use in global models such as NAVGEM (NRL global model) in applications which require forecasts that are sensitive to diurnal temperature variability. The model is an extension of Zeng and Beljaars' (2005) prognostic scheme of sea surface skin temperature for modeling and data assimilation. We show how the new SWaM model improves representation of the upper ocean thermal structure under a range of environmental conditions, subsequently increasing skill in SST prediction in 2-10 day forecasts. The SWaM, original Zeng and Beljaars model, and multilayer PWP model results are compared against data collected during the CINDY-DYNAMO field experiment with an underwater SeaGlider.