Oceanic fluxes of mass, heat, and freshwater : a global estimate and perspective

Abstract

Data from fifteen globally distributed, modern, high resolution, hydrographic oceanic transects are combined in an inverse calculation using large scale box models. The models provide estimates of the global meridional heat and freshwater budgets and are used to examine the sensitivity of the global circulation, both inter and intra-basin exchange rates, to a variety of external constraints provided by estimates of Ekman, boundary current and throughflow transports. A solution is found which is consistent with both the model physics and the global data set, despite a twenty five year time span and a lack of seasonal consistency among the data. The overall pattern of the global circulation suggested by the models is similar to that proposed in previously published local studies and regional reviews. However, significant qualitative and quantitative differences exist. These differences are due both to the model definition and to the global nature of the data set. The picture of the global circulation which emerges from the models IS a complex, turbulent flow. When integrated across ocean basins not one, but two major cells emerge. The first connects an Atlantic overturning cell (estimated at 18± 4x 109 kg s- 1) to the Southern Ocean where the Antarctic Circumpolar Current carries lower deep waters to the Indian and Pacific basins where they are converted to upper deep and intermediate waters before returning to the Atlantic. The second cell connects the Pacific and Indian Basins to the north and south of Australia. In t his cell deep waters pass into the Pacific and return within the Indian Basin as intermediate waters after passing through the Indonesian Passages. The two cells are found to be independent of one another, i.e. within the models, the Indonesian Passages do not represent a significant element in a net global circulation. While there is ample evidence of westward flow around the southern tip of South Africa which would support a "warm" water path scenario, the variability of flow in this region, rich with eddies makes hydrography a poor estimator of the relative strengths of the controversial "warm" and "cold" water paths. All existing estimates of Indonesian Passage throughflow, including the smallest (O x 106 m3 s-1) and the largest (20 x 106 m3 s-1), are consistent with the model constraints. When the Pacific- Indian throughflow is not constrained, the model produces an estimate of 11 ± 14x 109 kg s-1. The model heat flux estimates are both significantly different from zero and quite robust to changes in initial assumptions, with the exception of the choice of wind field. Although in this work it was not possible to compute freshwater fluxes which were significantly different from zero, future inclusion of salinity anomaly constraints along with terms describing vertical diffusion may yet make it possible to compute significant freshwater :flux estimates from hydrography.