How well do we know ocean salinity and its changes?
Introduction
In the ocean salinity is a very important ocean parameter, which, through its impact on density, determines many processes such as the horizontal circulation, vertical mixing, double diffusion, or even the near-surface stability. Through its impact on boundary layer processes, salinity also plays a role in modulating the air-sea interaction, including surface fluxes of heat and CO2. Despite this significance, salinity, historically, was under-sampled.
During the pre-electronic era, obtaining accurate salinity observations required great chemical analysis skills to precisely determine the salinity content of water samples on board of a research vessel or later in a laboratory at home resulting in very sparse salinity observations, historically. The situation improved dramatically through the invention of the CTD (conductivity-temperature-depth) technology, which allows inferring salinity in situ through conductivity measurements. Since then, obtaining routine salinity profiles became practical, but respective observations were still limited in space and time due to the sparseness of hydrographic sections performed by research vessels. It is only through the establishment of the autonomous Argo profiling float array (Argo Steering Team, 1998) that obtaining a routine sampling of high-quality salinity profiles down to 2000 m depth became possible on a global scale (Weller et al., 2019, Stammer et al., 2019).
Today, the Argo array provides more than 3500 profiles every 10 days on a nearly global nominal 3° geographic grid. The evolution of respective in situ salinity profiles available since the advent of the Argo array is displayed in Fig. 1a, showing the total number of salinity observations obtained globally per 5 m depth cells building up over the period 2004 to 2017. The figure clearly illustrates the unprecedented density of salinity profiles available today relative to the time before 2010 and indicated the sparseness of salinity data available prior to this period, especially before 2004. Also shown in Fig. 1b is the spatial distribution of historical salinity observations. As can be seen, the salinity observations per grid cell remain small not only over parts of the Southern Ocean but also in the northeast Pacific and some parts of the subtropical Atlantic.
Based on essentially the same historical salinity observations, several gridded salinity products of the historical data sets (usually called climatology) and reanalysis salinity products were generated which today are commonly used for salinity and freshwater change studies (e.g., Curry et al., 2003, Palmer et al., 2019). Best known among those climatologies is that of Levitus (Levitus, 1982), also known as the World Ocean Atlas (WOA), which regularly gets updated by new or newly recovered historic data sets (Zweng et al., 2018). Another often-used product is the EN4 data set (Good et al., 2013). However, as with temperature climatologies, substantial differences exist between individual salinity data sets, arising especially from the way the historical database is being quality controlled, processed and gridded. As a result, substantial differences remain between existing salinity data sets and thus also in our knowledge of the background against which we have to measure freshwater content changes in the ocean resulting from climate variability or anthropogenic climate change.
Recently, it became feasible inferring surface salinity from satellite born measurements of microwave radiance emitting from the top mm of the ocean. Respective missions now provide a global coverage of surface salinity roughly every 3 days with spatial resolution of about 50 km (Vinogradova et al., 2019). Remote sensing systems thus provide the highest available spatial and temporal resolutions of salinity variations, making them unique in capturing the temporal changes of salinity in dynamically complex regions (e.g. Tong et al., 2015, Boutin et al., 2012, Gourrion et al., 2011, Reul et al., 2012). However, satellite salinity retrievals are limited to the sea surface. Moreover, respective data have strong geographically varying biases, which can reach 0.5 in warm waters and about 1 or more in cold waters as will become obvious below.
A third important source of information about time-varying ocean salinity and its impact on various parameters such as geostrophic flow fields or sea level comes from ocean models, which, driven by atmospheric forcing fields, simulate salinity with and without incorporating salinity observations as constraints. Ultimately, respective salinity synthesis fields obtained through the assimilation of salinity observations into numerical models in combination with altimetry and other data should provide the most accurate climatologies. However, through an intercomparison of upper-ocean salinity fields resulting from several existing WCRP/CLIVAR ocean reanalyses, Shi et al. (2017) documented that substantial uncertainties remain in our information about time mean salinity fields, let alone their temporal changes. The authors conclude that most of the tested reanalysis products agree best in their salinity fields in the tropical Pacific concerning the mean and standard deviation (STD). Estimates agree the least in the Southern Ocean, where differences are typically of the order 0.1 and can reach 0.2 in the top 700 m depth-averaged salinity. Results confirm that the quality of reanalysis products depends on the data availability and shows clear temporal variations. Nevertheless, the authors hypothesize that resulting spatial patterns of temperature and salinity variability may be indicative of real dynamical changes of the large-scale ocean circulation.
To obtain more reliable salinity estimates for the understanding of ocean variability, improved salinity products are required. The goal of this paper is therefore to establish an understanding of the consistency of existing salinity products and to quantify remaining uncertainties by comparing a wide variety of products against each other. In detail, we will document the state of art of ocean salinity observing and investigate to what extent salinity variability and trends can be inferred reliably against the degree of uncertainties remaining in the existing data sets. Our focus will be especially on an evaluation of large-scale salinity structures and their temporal variations in the global ocean as well as knowledge gaps. The paper’s scope covers an analysis of salinity fields down to 700 m and the surface salinity as observed in situ and through remote sensing. Subsequently, we will use the combined information to investigate interannual and longer salinity changes in the global ocean. Trends in the surface and upper layer salinity are analyzed using the salinity observations within the last decades and using long reanalysis products, which go back to 1960 or beyond.
Questions that will be addressed entail:
How similar or dissimilar are existing salinity climatologies in describing the time mean ocean salinity field?
How well do existing salinity data sets, including satellite products, describe salinity changes?
Can a comparison of all available salinity products, including those from ocean reanalyses, document the state of our knowledge about salinity changes in the ocean?
The structure of the remaining paper is as follows: Section 2 describes input data sets. Section 3 analyzes the consistency and the spread of existing climatologies in describing time mean salinity structures. The analysis will also quantify our understanding of the overall degree of RMS salinity variability. Moreover, some emphasis will be placed on the knowledge added by using satellite salinity data. Section 4 focuses on an analysis of the seasonal salinity cycle while Section 5 analyses interannual and longer-term salinity variations. Also provided will be a comparison of the long-term changes with individual long-term time series stations. Concluding remarks will be given in Section 6.
Section snippets
Data sets
Used in this study are gridded in situ and sea surface salinity products, local time-series from moored instrumentation, as well as several ocean reanalysis products
1) Global gridded in situ products: Among the in situ data products used, some are based only on data collected by Argo profilers, others also include near surface data originating from surface drifters, mammals and thermosalinograph instruments installed on research vessels or also from voluntary operating merchant ships; others
Full-depth salinity climatologies
As reference for the later discussion, shown in Fig. 3 (left column) is the ensemble and time mean in situ salinity fields representing the Argo period 2004 – 2017. Fields are presented at 5 m depth and for three subsurface depth ranges representing the layers 10 to 50 m, 10 to 300 m and 300 to 700 m, respectively. As can be expected, all fields reproduce the major well-known structures of the salinity climatology, notably the subtropical salinity maxima, the relative salinity minima in the
Consistency of surface salinity information
An obvious advantage of satellite SSS measurements above existing in situ technologies is the quasi-synoptic global coverage with unprecedented spatial resolution (order 50 km) allowing improved observations of horizontal SSS gradients (e.g., D'Addezio et al., 2015). This benefits especially the observational capabilities of time varying SSS phenomena such as Tropical Instability Waves (e.g., Yin et al., 2014) or meanders/eddies in frontal systems (Gulf Stream, river outlets, Agulhas current
Annual cycle
An important component of ocean variability is its fluctuation occurring on the annual period. Largely driven by external forcing, this variability typically represents a significant peak in a spectral description of ocean variability, especially for temperature and sea level. Previously, the amount of annual variability in salinity was less well known, partly due to the lack of respective information in the in situ observational databases. Here, we revisit the question of how large the
Interannual salinity variability and trends
In the context of climate change of specific interest is the knowledge of long-term salinity changes and trends as they would be expected from a change in a global hydrological cycle. Several authors argued already that an enhancement of the hydrological cycle may have left an imprint in salinity anomalies at the sea surface (Durack et al., 2012, Skliris et al., 2016). This would imply an overall negative trend in the tropics and a positive trend in the subtropics. Respective changes might
Discussion and concluding remarks
In this paper, we established an understanding of the consistency among existing salinity products and quantified remaining uncertainties by comparing a variety of salinity products against each other. Focusing on the large-scale salinity structures of the top 700 m, we find that all existing climatological data sets reproduce the major well-known structures of the salinity climatology, notably the subtropical salinity maxima, the relative salinity minima in the upper layer of the tropics and
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
Funded in part through a DFG grand to Universität Hamburg as part of the Forschergruppe FOR1740, the BMBF funded project 03F0795C and through the ESA-CCI project on surface salinity. Contribution to the DFG funded excellence cluster CLICCS of Universität Hamburg.
References (61)
- et al.
The “great salinity anomaly” in the Northern North Atlantic 1968–1982
Prog. Oceanogr.
(1988) - et al.
Surface salinity in the Atlantic Ocean (30°S–50°N)
Prog. Oceanogr.
(2007) - et al.
The 2004–2008 mean and annual cycle of temperature, salinity, and steric height in the global ocean from the Argo Program
Prog. Oceanogr.
(2009) - et al.
A modeling study of the processes of surface salinity seasonal cycle in the Bay of Bengal
J. Geophys. Res. Oceans
(2014) On the Design and Implementation of Argo - an Initial Plan for the Global Array of Profiling Floats
(1998)- et al.
Evaluation of the ECMWF ocean reanalysis system ORAS4
Q. J. R. Meteorolog. Soc.
(2013) - et al.
The Ocean Reanalyses Intercomparison Project (ORA-IP)
J. Oper. Oceanogr.
(2015) - et al.
Characteristics of the seasonal cycle of surface layer salinity in the global ocean
Ocean Sci.
(2012) - et al.
First assessment of SMOS data over open ocean: Part II—Sea surface salinity
IEEE Trans. Geosci. Remote Sens.
(2012) - Boutin, J., J.L. Vergely, J. Köhler, F. Rouffi, and N. Reul, 2019. ESA Sea Surface Salinity Climate Change Initiative...
SODA3: a new ocean climate reanalysis
J. Clim.
An assessment of oceanic variability for 1960–2010 from the GFDL ensemble coupled data assimilation
Clim. Dyn.
Seasonal variability of salinity and salt transport in the northern Indian Ocean
J. Phys. Oceanogr.
Decadal trends of the upper ocean salinity in the tropical Indo-Pacific since mid-1990s
Sci. Rep.
Ocean salinities reveal strong global water cycle intensification during 1950 to 2000
Science
An enhanced PIRATA dataset for Tropical Atlantic Ocean-Atmosphere Research
J. Clim.
SMOS first data analysis for sea surface salinity determination
Int. J. Remote Sens.
Combined active/ passive retrievals of ocean vector wind and sea surface salinity with SMAP
IEEE Trans. Geosci. Remote Sens.
El Niño variability in simple ocean data assimilation (SODA), 1871–2008
J. Geophys. Res. Oceans
EN4: quality-controlled ocean temperature and salinity profiles and monthly objective analyses with uncertainty estimates
J. Geophys. Res. Oceans
A monthly mean dataset of global oceanic temperature and salinity derived from Argo float observations
JAMSTEC Report Res. Develop.
Interannual to Decadal Variability of Upper-Ocean Salinity in the Southern Indian Ocean and the Role of the Indonesian Throughflow
J. Clim.
Evaluating the GECCO3 1948–2018 Ocean Synthesis-a configuration for initializing the MPI-ESM climate model
Q. J. R. Meteorolog. Soc.
Mechanisms of mixed-layer salinity seasonal variability in the Indian Ocean
J. Geophys. Res. Oceans
Quality assessment of space-born sea surface salinity observations over the subpolar North Atlantic
J. Geophys. Res. Oceans
In-situ based reanalysis of the global ocean temperature and salinity with ISAS: Variability of the heat content and steric height
J. Clim.
Comparison of the Atlantic Meridional Overturning Circulation between 1960 and 2007 in six ocean re-analysis products
Clim. Dyn.
Cited by (15)
An observing system experiment framework for the tropical Indian Ocean salinity: A case study using a constellation of three satellites
2023, Deep-Sea Research Part II: Topical Studies in OceanographyAquarius sea surface salinity retrieval in coastal regions based on deep neural networks
2023, Remote Sensing of EnvironmentCitation Excerpt :Sea surface salinity (SSS) is an important indicator reflecting changes in the global water cycle and a critical parameter for detecting climate change related to the water cycle (Schmitt, 2008; Emery and Meincke, 1986; Lagerloef et al., 2010; Gould and Cunningham, 2021; Olmedo et al., 2022). SSS is also significant in determining the seawater density that affects ocean circulation (Koblinsky et al., 2003; Martin, 2014), and in regulating air-sea coupling through controls on boundary layer processes (Stammer et al., 2021; Durack, 2015). At present, the main methods to measure SSS include the global array of Argo floats, thermosalinograph (TSG), satellite observations, etc.
Atlantic Niño induced sea surface salinity variability as observed from the satellite
2024, Environmental Research LettersValidation of the multi-satellite merged sea surface salinity in the South China Sea
2023, Journal of Oceanology and LimnologyRemote sensing of sea surface salinity: challenges and research directions
2023, GIScience and Remote Sensing