Abstract
Monitoring sea surface salinity and density variations is crucial to investigate the global water cycle and the ocean dynamics, and to analyse how they are impacted by climate change. Historically, ocean salinity and density have suffered a poor observational coverage, which hindered an accurate assessment of their surface patterns, as well as of associated space and time variability and trends. Different approaches have thus been proposed to extend the information obtained from sparse in situ measurements and provide gap-free fields at regular spatial and temporal resolution, based on the combination of in situ and satellite data. In the framework of the Copernicus Marine Environment Monitoring Service, a daily (weekly sampled) global reprocessed dataset at ¼°x¼° resolution has been produced by modifying a multivariate optimal interpolation technique originally developed within MyOcean project. The algorithm has been applied to in situ salinity/density measurements covering the period from 1993 to 2016, using satellite sea surface temperature differences to constrain the surface patterns. This improved algorithm and the new dataset are described and validated here with holdout approach and independent data.
Highlights
The sea surface salinity (SSS) is recognized as one of the Essential Climate Variables (ECVs) by the Global Climate Observing System (GCOS)
The aim of this paper is to present this new dataset and the modifications of the multivariate optimal interpolation (OI) algorithm carried out to account for the sparseness of in situ data before 2002
To verify that the higher variance found at the large mesoscale range is not given by noise or artifacts, we compared co-located SSS TSG and SSS products spectra from repeated high-resolution TSG measurements collected along a merchant shipping route, which provides enough data to minimize the error in spectral computation
Summary
The sea surface salinity (SSS) is recognized as one of the Essential Climate Variables (ECVs) by the Global Climate Observing System (GCOS). It must be kept in mind, anyway, that the comparison between point-wise in situ measurements and FIGURE 2 | Matchup positions between Global L4 dataset of SSS and TSG measurements. To verify that the higher variance found at the large mesoscale range is not given by noise or artifacts, we compared co-located SSS TSG and SSS products spectra from repeated high-resolution TSG measurements collected along a merchant shipping route, which provides enough data to minimize the error in spectral computation To this aim, we started from same data used by Kolodziejczyk et al (2015), i.e., observations from Toucan and Colibri ships (their Table 1), which span over almost 10 years and cover a transect in the North Atlantic, looking at data collected. −30 < lon < −5; −60 < lat < −30 60 < lon < 90; −60 < lat < −5 −180 < lon < −100; −60 < lat < 0
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
