Estimating eddy heat flux from float data in the North Atlantic: The impact of temporal sampling interval

Acoustically tracked float data from 16 experiments carried out in the North Atlantic are used to evaluate the feasibility of estimating eddy heat fluxes from floats. Daily float observations were bin averaged in 2 degrees by 2 degrees by 200-db-deep geographic bins, and eddy heat fluxes were estimated for each bin. Results suggest that eddy heat fluxes can be highly variable, with substantial outliers that mean that fluxes do not converge quickly. If 100 statistically independent observations are available in each bin (corresponding to 500-1000 float days of data), then results predict that 80% of bins will have eddy heat fluxes that are statistically different from zero. Pop-up floats, such as Autonomous Lagrangian Circulation Explorer (ALACE) and Argo floats, do not provide daily sampling and therefore underestimate eddy heat flux. The fraction of eddy heat flux resolved using pop-up float sampling patterns decreases linearly with increasing intervals between float mapping and can be modeled analytically. This implies that flux estimates from pop-up floats may be correctable to represent true eddy heat flux.