An Observational and Numerical Investigation of the Climatological Heat and Salt Balances at OWS Papa

Abstract

The climatological heat and salt budgets at OWS Papa are examined over the years 1960 to 1981. The average surface fluxes of heat and freshwater are estimated to be 25±15 W m?2 and 4.4±4.3 mg m?2 s?1, respectively. Year to year changes in these fluxes are found to be uncorrelated with changes in ocean heat and salt content above 200-m depth. Different hypotheses of how the surface fluxes are balanced are tested using a one-dimensional model of vertical mixing with prescribed horizontal and vertical heat and salt transport. The surface freshwater flux must be at least 3 mg m?2 s?1, on average, in order to balance model-computed vertical diffusion. Additional surface input is balanced by the vertical advection of salt from below. Other processes are required to balance the heat budget. Assuming that the surface flux estimates are in error and correcting them to give zero net fluxes of both heat and freshwater leads to a systematic erosion of the main pycnocline and deepening of winter mixed layers, such that the surface waters become too cold and too salty. These trends are reduced but not eliminated by including a steady vertical advection due to Ekman pumping and accounting for vertical diffusion in the salt and heat budgets. Balancing the remaining heat by horizontal advection throughout the water column results in a local minimum of too cold water at about 165-m depth. However, acceptable long-term simulations are achieved if the required cold water is transported into the seasonal thermocline and isothermal layer only during the fall and winter months. Observations supporting this scenario are reviewed. Model sensitivity experiments with this balance show which combinations of surface heat and freshwater fluxes produce the observed average amplitude of the annual cycle of sea surface temperature while achieving a realistic salt balance. Coupling the model by using its upper-layer temperature in the surface flux calculations greatly improves the simulators compared to observations and only changes the average heat flux by 1 W m?2. It is concluded that using constant surface flux corrections does not properly account for oceanic transport deficiencies that are not both concentrated in the mixed layer and steady throughout the year.