Heat and Freshwater Budgets From Two Intensive Upper Ocean Surveys During JASMINE

During the May-June 1999 JASMINE expedition, two five-day long intensive surveys of upper ocean temperature, salinity and velocity were conducted in the southern Bay of Bengal. The first took place during calm winds, relatively clear skies, and little rainfall. The second was conducted during strong southwest monsoon winds, cloudy skies and heavy rainfall. The deepest surface mixed layers observed were during the second survey, and did not exceed 45 m.

All terms in the upper ocean heat and freshwater budgets were estimated from observations except for vertical heat and salt fluxes due to turbulence. These are estimated as a residual, assuming that surface heat and freshwater fluxes are known accurately. Vertical velocity was estimated below 80 m assuming that density surface displacements are not due to turbulent mixing. Above 80 m, constant divergence was assumed, which is appropriate for large vertical scales associated with baroclinic waves, but which neglects local Ekman pumping.

Integrating from the surface downward to 80 m, and assuming that the vertical turbulent fluxes at 80 m are negligible, we estimate the surface fluxes of heat and freshwater. For the first survey, we estimate 74 Wm^-2 heating, and 0.4 psu m d^-1 salinization (-12 mm d^-1 net freshwater flux). For the second survey, we estimate 109 Wm^-2 cooling, and 1.6 psu m d^-1 freshening (48 mm d^-1 net freshwater flux). The heat fluxes are within about 15 Wm^-2 of the observed fluxes (preliminary shipboard estimates from C. Fairall and F. Bradley). The budget estimate of net freshwater flux during the second survey is within 20% of the shipboard surface flux estimates. However, the net evaporation estimated from the upper ocean salt budget during the first survey is substantially higher than the shipboard surface flux estimate.

The discrepancies may be attributable to uncertainties in our budget estimates or to the assumptions employed to reach our results. The assumption of negligible vertical turbulent heat and salt fluxes at 80 m is likely too strong. The neglect of Ekman pumping during the first survey is likely not an error source because the winds were generally calm over a large region surrounding the survey area.