Solomon, A., and I. Wainer, 2006: Pacific tropical-extratropical thermocline water mass exchanges in the NCAR CCSM3. Ocean Modelling, j.ocemod.2006.04.003.


ABSTRACT

In this study we document how model biases in extratropical surface wind and precipitation, due to oceanÐatmosphere coupling, are communicated to the equatorial Pacific thermocline through Pacific Subtropical Cell (STC) pathways. We compare the simulation of climate mean Pacific Subtropical Cells (STCs) in the NCAR Community Climate System Model version 3 (CCSM3) to observations and to an uncoupled ocean simulation (the ocean component of the CCSM3 forced by observed wind stress and surface fluxes). We use two versions of the CCSM3 with atmospheric resolution of 2.8° (T42) and 1.4° (T85) to investigate whether the climate mean STCs are sensitive to the resolution of the atmospheric model.

Since STCs provide water that maintains the equatorial thermocline, we first document biases in equatorial temperature and salinity fields. We then investigate to what extent these biases are due to the simulation of extratropicalÐtropical water mass exchanges in the coupled models. We demonstrate that the coupled models' cold and fresh bias in the equatorial thermocline is due to the subduction of significantly fresher and colder water in the South Pacific. This freshening is due to too much precipitation in the South Pacific Convergence Zone. Lagrangian trajectories of water that flows to the equatorial thermocline are calculated to demonstrate that the anomalously large potential vorticity barriers in the coupled simulations in both the North and South Pacific prevent water in the lower thermocline from reaching the equator. The equatorial thermocline is shown to be primarily maintained by water that subducts in the subtropical South Pacific in both the coupled and uncoupled simulations. It is shown that the zonally integrated transport convergence at the equator in the subsurface branch of the climate mean STCs is well simulated in the uncoupled ocean model. However, coupling reduces the net equatorward pycnocline transport by not, vert, ~4 Sv at 9°S and not, vert, ~1 Sv at 9°N. An increase in the atmospheric resolution from T42 to T85 results in more realistic equatorial trades and off-equatorial convergence zones.