Further refinements to models for the bulk‐skin sea surface temperature difference

Abstract

The bulk‐skin sea surface temperature difference, ΔT, has an important impact on both satellite remote sensing and air‐sea interaction studies. Extensive in situ data from diverse environmental conditions are used to explore the dependence of ΔT on environmental conditions and to evaluate and refine models for ΔT during nighttime. Data from 10 different cruises covering large ranges of latitudes and seasons in both the Atlantic and Pacific Oceans were accumulated and processed consistently. Comparison of the measured ΔT values with the environmental parameters generally confirmed an observed tendency for ΔT to approach a near‐constant value as the wind speed increases but also demonstrated a dependence of the residual variability on the net heat flux. To adequately predict ΔT, it is necessary to use a model that incorporates a dependence on all these parameters. Existing models for ΔT are all shown to be unable to fully reproduce the observed variability under all conditions. To improve the ability to predict ΔT, a new surface renewal‐type model was developed incorporating four additional physical processes. New expressions are proposed for renewal timescales, accounting for microscale and large‐scale wave breaking, the effects of capillary waves, and a saturated shear regime. To incorporate the simultaneous effects of multiple physical processes within the sampling region and period, the modeled ΔT is computed as a weighted combination of the contributions of the relevant regimes. An initial specification of the weights, derived from the observed ΔT values and environmental parameters, demonstrates that a suitable combination of the timescales can reproduce the majority of the observed variability in ΔT. A second, objective determination of the weights, using conjugate gradient minimization, produces slightly poorer results but demonstrates that the model can be used in a prognostic manner to significantly improve upon the existing predictions of ΔT.