Uncertainty of North Atlantic Current observations from altimetry, floats, moorings, and XBT

Abstract

Large-scale geostrophic flow across an oceanographic section can be computed from in situ density profiles and satellite-borne sea level anomaly measurements at the section endpoints. Here, uncertainties are estimated and compared for observations that span the North Atlantic Current. Out of multi-year mooring records from two sites, Porcupine Abyssal Plain and Central Irminger Sea, 16 months from the 2002–2005 period are available that have sufficient data coverage to observe this current. Transport and uncertainty estimates from these moorings are compared to results from Argo floats and expendable bathythermographs. The latter have longer data sets and provide decadal time series with semiannual resolution in time. Typical uncertainties, which take sensor errors and sub-sampling the eddy field into account, are below 2 Sv of volume transport for the mooring- and float-based estimates and just below 3 Sv for the expendable bathythermographs. Peak-to-peak variability occurs on decadal time scales and is about 11 Sv, well above the observational uncertainty. We can therefore have confidence that the observed signal reflects true changes in the ocean. The time series is further extended to 25 years duration, using solely the altimetry data by proxy. The aim of this study is to add value to such time series by understanding and quantifying the uncertainties and consistency between methods. The methodology is applicable at other locations as well.