How Salty Is the Global Ocean: Weighting It All or Tasting It a Sip at a Time?

Global ocean mean salinity S is a key indicator of the Earth's hydrological cycle and the exchanges of freshwater between the terrestrial water and ice reservoirs and the ocean. We explore two different ways of determining how salty the ocean is: (1) use in situ salinity measurements to taste the ocean a sip at a time and obtain a sample average; (2) use space gravimetry to weigh the whole ocean including sea-ice, and then separate sea-ice effects to infer changes in liquid freshwater content and thus S. Focusing on the 2005-2019 period, we assess monthly series of S derived from five different in situ gridded products, based mostly but not exclusively on Argo data, versus a series obtained from GRACE and GRACE Follow-On data and available sea ice mass estimates.

There is little consistency in S series from the two methods for all time scales examined (seasonal, interannual, long-term trend). In situ series show larger variability, particularly at the longest scales, and are somewhat incoherent with the GRACE-derived series. In addition, there are wide spread differences among all the in situ S series, which denote their considerable sensitivity to choice of data, quality control procedures, and mapping methods. Results also suggest that in situ S values are prone to systematic biases, with most series showing increases after around 2014 that are equivalent to a drop in barystatic sea level of tens of centimeters! Estimates derived from GRACE are much smaller in magnitude and consistent with contributions of freshwater to the global mean sea level budgets, and they are thus more reliable than in situ-based S estimates. The existence of GRACE-derived estimates can serve as a consistency check on in situ measurements, revealing potential unknown biases and providing a way to cross-calibrate the latter data.