How well do we know ocean salinity and its changes?

Abstract

An understanding of the quality and accuracy of state-of-the-art time-varying salinity information is being established by comparing existing products against each other as well as against existing ocean reanalyzes and satellite retrieval. The focus the paper is on large-scale salinity structures and their temporal variations from the sea surface down to 700 m depth. Uncertainties are quantified through the spread between all existing estimates on various time scales including long-term trends. Results suggest that uncertainties of time mean climatological salinity estimates are as large as 0.3 within the first 50 m of high vertical salinity gradients and even larger in some boundary regions. Over the open (deep) ocean, a near-surface spread of 0.1 can be observed; below 300 m depth it stays below 0.03 over most part of the world ocean. Measuring salinity in near-costal and highly dynamic regions remains problematic and calls for a better integration of existing in situ and satellite capabilities there. There remain challenges in estimating the amplitude of the salinity seasonal cycle from the existing in situ database. This holds especially within the tropics, where the disagreement between the seasonal salinity amplitude estimated from individual situ data sets is typically around 30%, locally even up to 50%. Regarding interannual and longer-term salinity changes, inferred trend patterns agree well with the standard deviation of the decadal climatologies, suggesting that it remains difficult to obtain statistically significant estimates of long-term trends from the existing short data base over the world ocean. Overall, our results call for further improvements of ocean reanalysis products and their expanded use in studies of ocean variability and changes in the future.