Open Access
On the seasonal mixed layer simulated by a basin‐scale ocean model and the Mellor‐Yamada turbulence scheme
Author(s) -
Ezer Tal
Publication year - 2000
Publication title -
journal of geophysical research: oceans
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.67
H-Index - 298
eISSN - 2156-2202
pISSN - 0148-0227
DOI - 10.1029/2000jc900088
Subject(s) - mixed layer , thermocline , turbulence , climatology , forcing (mathematics) , environmental science , mixing (physics) , meteorology , dissipation , atmospheric sciences , surface layer , geology , physics , layer (electronics) , materials science , quantum mechanics , composite material , thermodynamics
Seasonal changes and vertical mixing processes in the upper layers of the North Atlantic Ocean are simulated with a basin‐scale sigma coordinate ocean model that uses the Mellor‐Yamada turbulence closure scheme. The cause of insufficient surface mixing and a too shallow summertime thermocline, common problems of ocean models of this type, is investigated in detail by performing a series of sensitivity experiments with different surface forcing conditions and different turbulence parameterizations. A recent improvement in the parameterization of the dissipation term in the Mellor‐Yamada turbulence scheme, which has shown a significant improvement in one‐dimensional calculations, had a positive but relatively small influence on the three‐dimensional calculations. The results quantify the improvement in the model upper ocean thermal structure as surface forcing becomes more realistic from one experiment to another, for example, when monthly mean winds are replaced by 6 hour variable winds. The inclusion of shortwave radiation penetration is especially important to prevent overly shallow model mixed layers during the summer and seems to affect not only the surface layer but also the thermal structure of the upper 200 m of the ocean. The difficulty of evaluating turbulent mixing processes in three‐dimensional models due to errors in surface fluxes, spatial changes, and three‐dimensional effects, as shown here, points to the important role still left for one‐dimensional turbulence models in improving parameterizations used in three‐dimensional realistic models.