Evaluation of Surface Heat Flux Uncertainties and Their Impacts on the Study of Ocean Mixed Layer Temperature Variability

A major theme of recent research is the investigation of the nature of climate variability and the current capability to measure, model, and predict it. This is a formidable task that involves understanding complex interactions and exchanges of energy between the major elements of the Earth system. With their ability to store and release vast quantities of heat, the oceans are an integral element of climate variability. Accurately modeling coupled atmosphere-ocean variability relies upon a proper characterization of the exchanges of heat and momentum across the air-sea interface. The exchange of heat takes place through net shortwave and terrestrial radiative fluxes and turbulent exchanges of heat and moisture. Estimating these interactions with sufficient accuracy is a difficult challenge. These processes contain inherent errors due to insufficient knowledge of physics, observational uncertainty, and parameterization deficiencies. Uncertainties arising from the estimation of the surface turbulent and radiative processes generate limitations to the understanding of the primary mechanisms governing oceanic variability. This work elucidates the impact of uncertainties in the estimation of turbulent and radiative heat fluxes on the analysis of the mixed layer temperature balance, an effect that has not been properly quantified although recognized in most previous analyses. In particular, this work focuses on variability at seasonal and intraseasonal time scales. The analyses of this work include: i) an updated characterization of uncertainties in current state-of-the-art estimates of the turbulent and radiative heat fluxes, ii) an examination of the closure of the mixed layer temperature balance on seasonal and intraseasonal time scales, iii) an evaluation of the sensitivity of the mixed layer temperature balance to differences between surface heat flux estimates, iv) the development of a flexible approach by which to determine required accuracies of the net surface heat flux, and v) an exploration of the role of mixed layer depth variability on the mixed layer temperature balance. Taken together, the results of these analyses provide a framework to understand the impact of surface heat flux uncertainties within the context of upper ocean mixed layer variability. The analyses performed in this study have exploited a set of eight turbulent and six radiative heat flux estimates. An intercomparison of these products has revealed that the typical spread between products has been reduced relative to previous generations of estimates. Differences between radiative and turbulent heat flux estimates are typically within 15-20% of one another on regional and seasonal scales although larger uncertainties remain in traditionally problematic regions (e.g., cloud-topped boundary layers, western boundary currents). On both intraseasonal and seasonal time scales, the ocean mixed layer is controlled most strongly by the net shortwave and turbulent latent heat fluxes over the world oceans with the exception of the deep tropics wherein oceanic processes are also important. The current ensemble mean estimates of the net surface heat fluxes and oceanic process are capable of resolving the upper ocean mixed layer temperature seasonal cycle quite well in many locations; areas of strong net heat flux warming are somewhat problematic. On intraseasonal time scales, small signal to noise ratios and large residual imbalances leave little room to make definitive conclusions on the role of individual elements of mixed layer forcing. However, general features of the relative importance of surface heat flux variability versus oceanic variability are supported from previous studies. The mixed layer temperature balance is found to be most sensitive to uncertainties in the latent and net shortwave heat fluxes. The timing of the shoaling of the mixed layer depth is also important to the sensitivity of the mixed layer temperature balance. Taking into account mixed layer depth variability is found to be important to understanding the role of the net surface heat fluxes in generating mixed layer temperature warming and cooling. Current estimates of the net heat flux uncertainty are outside of the traditional 10 W m &minus 2 goal on seasonal time scales and spatial scales on the order of 1000 km. The approach designed within this investigation suggests that a 10 W m &minus 2 limit is somewhat too restricting if the aim is to resolve the seasonal mixed layer temperature evolution. In short, the use of the ocean mixed layer temperature balance has provided a unique framework for translating uncertainties in the surface heat flux estimates into a practical context. It is hoped that a better appreciation of these uncertainties will lead to an improved ability to model and understand the mechanisms by which the oceans contribute to variability of Earth's climate.