This paper presents the first Soil Moisture and Ocean Salinity (SMOS) Sea Surface Salinity (SSS) dedicated products over the Baltic Sea. The SSS retrieval from L-band brightness temperature (TB) measurements over this basin is really challenging due to important technical issues, such as the land–sea and ice–sea contamination, the high contamination by radio-frequency interference (RFI) sources, the low sensitivity of L-band TB at SSS changes in cold waters, and the poor characterization of dielectric constant models for the low SSS range in the basin. For these reasons, exploratory research in the algorithms used from the level 0 up to level 4 has been required to develop these dedicated products. This work has been performed in the framework of the European Space Agency regional initiative Baltic+ Salinity Dynamics.
Two Baltic+ SSS products have been generated for the period 2011–2019 and are freely distributed: the Level 3 (L3) product (daily generated 9 d maps in a
The impact assessment of Baltic+ SSS products has shown that they can help in the understanding of salinity dynamics in the basin. They complement the temporally and spatially very sparse in situ measurements, covering data gaps in the region, and they can also be useful for the validation of numerical models, particularly in areas where in situ data are very sparse.
The Baltic Sea is a strongly stratified semi-enclosed shallow sea that has several sub-basins, which are mostly separated from each other by underwater sills. The water balance is positive with large freshwater supply from rivers and precipitation and with occasional high-saline water input from the North Sea through the narrow and shallow Danish straits. The propagation of the saline water inflows in the deeper layers is hampered by bathymetry; the basins are connected to each other through narrow channels and shallow sills and by hydrodynamic restrictions including brackish water outflow, fronts and mixing. The mean depth of the Baltic Sea is only 54 m, which yields to highly variable ocean dynamics mainly controlled by local atmospheric forcing
The surface layer salinities in the southern and central basins are between 6.5–8.5 psu, being highest in the southern part and decreasing towards the north. Due to the voluminous river discharge the salinity decreases towards the ends of the sub-basins in the northern and eastern extremity. Salinities in the basins also differ from each other clearly. In the Bothnian Sea, the surface salinity is typically between 5–6 psu, in the Bothnian Bay between 2–4 psu and in the Gulf of Riga 4.5–6 psu. The Gulf of Finland is an exception, because it is a direct continuation of the central basin and resembles a very large estuary, having a continuous salinity gradient in the surface salinity decreasing from 6 psu in the western part close to 0 psu in the eastern part. Surface salinity is thus an indicator of the dynamics and changes in the conditions of the basins and of the exchange between them. More detailed descriptions of the salinity variation and dynamics in the Baltic Sea can be found for example in
Complex oceanographic conditions within the Baltic Sea are a challenge for oceanographic models and, for example, the salinity dynamics cannot be comprehensively simulated by the present model systems (e.g.,
Remote sensing has been used for decades in the Baltic Sea to follow the ice conditions, surface temperature and algal blooms. However, salinity conditions have remained outside of an overall synoptic view so far. There is a need to put in situ data in context because of the strong seasonal cycles and strong meso-scale dynamics with fronts and eddies, which have horizontal dimensions of the order of kilometers to tens of kilometers. Remotely sensed salinity information would be a valuable addition to the available tools for understanding the changes.
For all the above, Earth observation sea surface salinity (SSS) measurements have a great potential to help in the understanding of the dynamics in the basin
the contamination of ocean brightness temperature (TB) measurements close to land, particularly crucial since few points are further than 110 km from the nearest coast the contamination of ocean TB close to ice edges, since the Bothnian Bay and the eastern part of the Gulf of Finland are ice-covered every year, and also the Baltic Proper in severe winters; the high contamination by radio-frequency interference (RFI) sources the low sensitivity of L-band TB to SSS changes in the cold waters dielectric constant models that relate the TB and the SSS that were derived from salinity measurements in the range of the global ocean (32–38 psu) and are not fully tested in the low-SSS and low-SST regimes of the Baltic Sea.
For all the above conditioning factors, essential modifications have been required in the algorithms used from the very low level of processing up to the SSS retrieval to develop dedicated SSS products over the Baltic Sea:
In the brightness-temperature generation, the ALL-LICEF calibration approach and the correction of the correlators' efficiency errors proposed by In the SSS retrieval, two major changes have been introduced with respect to the original debiased non-Bayesian retrieval the empirical correction of the dielectric constant model for the low SSS regimes of the Baltic Sea; the characterization and correction of SSS systematic errors, depending not only on the acquisition conditions, but also on the SST.
In this work, we present the dedicated algorithms used to develop the Baltic+ L3 and L4 SSS products and their quality assessment. The article is structured as follows: Sect.
This section is devoted to explaining the datasets and the main algorithms used in the generation of the Baltic+ L3 and L4 SSS products (see Fig.
computation of brightness temperatures at antenna reference frame (ARF) from level 0 data by using the ALL-LICEF calibration and applying the computation of the measured TB at the bottom of the atmosphere (BOA); computation of the difference between SMOS TB and modeled TB and inversion to retrieve SSS; correction of systematic biases on SSS by means of SMOS-based climatological data; generation of the Baltic L3 salinity maps; correction of temporal biases found in L3 SSS maps; multifractal fusion of L3 SSS maps with an SST field to generate the L4 SSS maps.
Block diagram of the Baltic+ SSS processor.
We generate the TB dataset starting from the SMOS ESA Level 0 data (
The auxiliary data used for the SSS retrieval come from the European Centre for Medium range Weather Forecast (ECMWF)
We use a regional climatology as an annual reference SSS field, which is added to the debiased SMOS SSS anomalies (see Sect.
Since the SST is one important driver of the SSS errors, we analyzed the errors of all the available SST datasets over the Baltic sea: (i) ECMWF
The ESA CCI SST combines data from both the Advanced Very High Resolution Radiometer (AVHRR) and Along-Track Scanning Radiometer (ATSR) SST_CCI Climate Data Records, providing daily global SST on a 0.05
The OSTIA dataset uses satellite data provided by international agencies via the Group for High Resolution SST (GHRSST). These products include data from microwave and infrared satellite instruments. The OSTIA dataset also has daily global coverage on a 0.05
These SST products are used in the SSS retrieval (Sect.
A sea-ice mask is required to discard those SSS retrievals in ice-covered regions. This sea-ice mask is created from the sea-ice fraction (SIF) information provided by OSTIA (product ID “OSTIA-UKMO-L4-GLOB-v2.0”,
We use the Baltic Sea physics reanalysis (CMEMS_product_ID: BALTICSEA_REANALYSIS_PHY_003_011,
We use the daily SSS of the BSIOM hindcast simulation using the model configuration described in
Some of the corrections we propose to improve the quality of TBs over the Baltic Sea are not included in the current operational ESA L1B products. For this reason, we have used the MIRAS Testing Software (MTS)
We use the ALL-LICEF mode as the calibration approach
The application of this correction leads to an overall reduction of the TB contamination close to the coasts
In the Baltic, the ALL-LICEF calibration approach and the
The impact of the
A 9 d
Before the salinity retrieval process, the corrected brightness temperatures are transformed from antenna to ocean surface
The SSS retrieval is based on finding the appropriate value of raw SSS that makes the GMF (geophysical model function) of TB closer to the actually measured TB. The GMF is derived from a dielectric constant model for sea water. All the dielectric constant models found in the literature are built by empirical fitting of laboratory measurements. The dielectric constant model of
In a first analysis of the retrieved raw SSS, a low number of retrievals was obtained in some regions of the Baltic, especially in regions where the SSS values are very low.
Figure
It was found that raw SSS values were only retrieved if
The M&W dielectric constant model is reviewed for the SST and SSS conditions of the Baltic Sea. Figure
For very diluted solutions, the conductivity depends almost linearly on the salinity
However, as shown in Fig.
The debiased non-Bayesian (DNB) SSS retrieval
A single SSS value is retrieved for each TB measurement at a given incidence angle, unlike the conventional Bayesian retrieval, where a single SSS is retrieved from the entire set of multi-angular TB. Details on the SSS retrieval (referred to as raw SSS, since they need to be corrected from systematic biases and filtered) can be found in Sect. 2.2 of
We want to characterize the systematic errors in SMOS SSS. This characterization is based on the hypothesis that systematic errors are the same for all those SSS acquired under the same conditions. In total, 7 years of SMOS SSS retrievals (2013–2019, the cleanest period in terms of RFI contamination) are used for the characterization of those systematic biases on raw SSS that do not depend on time.
The raw SSS are grouped together according to their geolocation in the same fixed grid of TB measurements (coordinated by the latitude and longitude), overpass direction (ascending or descending, denoted by
When we applied the original DNB SSS retrieval to the Baltic Sea, we observed that seasonal variations were much higher than in the global ocean
Maps of monthly mean differences between SMOS SSS and the salinity field of CMEMS Baltic reanalysis [psu] for the year 2013.
Then, we analyzed the dependence of these differences on SST. SMOS SSS fields retrieved in 2013 were collocated with the salinity and temperature outputs from the CMEMS Baltic reanalysis. Figure
Difference between the SMOS SSS and reanalysis salinity variability as a function of the SST.
Therefore, for each given 6-tuple (instead of the 5-tuple of the original DNB),
Bins of SST for the computation of SMOS-based climatological data and the corresponding ranges of SST to be applied.
Still, the classification of the raw SSS for the 6-tuple leads to SMOS-based climatological distributions with a significantly reduced number of events. For this reason, the strategy for computing the SMOS-based climatological data, i.e., the central estimator of all the raw SSS acquired under a given 6-tuple, is changed with respect to the original DNB. We base the correction of systematic biases and filtering criteria only on the first- and second-order moments. In the Baltic Sea, the presence of outliers in the raw SSS highly impacts on the estimation of the statistical parameters that characterize the SMOS-based climatological distributions. To avoid this, the statistics are computed only with raw SSS belonging to the interval between the 5-quantile (IQ5) and the 95-quantile (IQ95). Hence, the mean (
Examples of maps of the mean and the standard deviation of the SMOS-based climatological distributions are shown in Fig.
SMOS-based climatological distributions for descending overpasses,
For the generation of the debiased non-Bayesian SMOS SSS values, each raw SSS acquired at a time
Then, a time-independent SSS reference is added to the SMOS SSS anomalies to obtain the final debiased SSS values. The annual reference SSS field used is the Baltic regional climatology provided by SeaDataNet (see Sect.
We study now whether the multi-annual mean of the salinity (required for the bias mitigation) changes with SST. For this, the impact of adding the regional climatology computed per bin of SST versus using a unique regional climatology as the annual reference field is analyzed. We use the salinity and temperature provided by CMEMS Baltic reanalysis in the period 2013–2019 to compute the averaged salinity for each bin of temperature. The mean error when using the single regional climatology as the annual reference field, instead of using the mean salinity value per bin of temperature (taking into account the frequency of each temperature value), is shown in Fig.
Mean error when applying a single annual reference climatology instead of a different climatology computed for each bin of temperature.
Errors in SSS retrievals over the Baltic Sea are expected to be much larger than in the global ocean, due to the low sensitivity of SSS to TB at cold waters. Moreover, residual errors caused by land–sea and ice–sea contamination, as well as perturbations by RFI sources, are also affecting the salinity retrievals. For this reason, the filtering criteria defined for the BEC global product
The filtering criteria are the following:
Any raw SSS out of the range [ For a given 6-tuple, The histogram has less than 30 measurements. The standard deviation is greater than 35 psu. Raw SSS are discarded if they deviate too much from the SMOS-based climatological data. That is, any raw SSS value outside the interval defined by In order to improve the quality of L3 SSS maps, all SSS values with an associated SSS uncertainty (estimated as detailed in Sect. 2.2.2 of SSS retrievals in the Skagerrak and the Kattegat straits (grid points with longitudes lower than 14
If the SMOS-based climatological distribution corresponding to a given 6-tuple has been discarded following the previous criteria, then all the associated raw SSS are discarded.
The Baltic+ L3 SSS data product is provided in a regular longitude–latitude grid of
An example of a L3 SSS map and its associated error are shown in Fig.
SMOS measurements are affected not only by spatial biases, but also by biases that depend on time
In the BEC global product
Temporal bias correction computed for the SSS product during 2013 by using the CMEMS Baltic reanalysis (red) and SDN in situ measurements (blue). Note that the peaks in the correction computed from SDN in situ measurements are due to the very scarce and fragmentary spatial distribution of collocated in situ data.
L4 SSS product has been generated by applying multifractal fusion techniques
The same SST data that are used as auxiliary data in the SSS retrieval are used here as template in the fusion scheme. L4 SSS maps are produced with
the same spatiotemporal resolutions as the template, i.e., daily maps at a spatial grid of
Ratio of time when the SSS from CMEMS Baltic reanalysis is used in those grid points where the SMOS SSS L3 product is not available (from
To assess the inherent uncertainty of the L4 SSS product, the correlated triple collocation (CTC) method is used
When applying CTC, the data are assumed to represent similar spatiotemporal scales with two of the datasets possibly having correlated errors. Under these conditions, CTC can be used to obtain maps of error variances of triplets of remote sensing SSS maps.
We consider three sets of collocated SSS maps in the period 2016–2018: (i) Baltic+ L4 SSS product, (ii) CMEMS Baltic reanalysis product
Error standard deviations computed by CTC for Baltic+ L4 SSS during the
We compare the performance of the new Baltic+ SSS to those of other existing EO SSS products. The satellite SSS products used for this inter-comparison are the following:
Figure
Spatiotemporal coverage of year 2016 (percentage of valid SSS retrievals with respect to the total number of maps in the year) for each satellite product:
Ship tracks from the FerryBox voluntary network measure both temperature and salinity in mounted thermosalinographs (TSGs) on voluntary vessels routinely making transects in the Baltic Sea. These data were available at CMEMS under the product identifier INSITU_BAL_TS_REP_OBSERVATIONS_013_038 (it was retired in March 2020 and replaced by INSITU_GLO_TS_REP_OBSERVATIONS_013_001_b;
The data collected from these vessels pass quality control checks before being distributed to the science community. All the ship routes available for the validation of Baltic+ SSS products are collected in Table
FerryBox ship routes and periods of operation.
The SeaDataNet (SDN) Temperature and Salinity historical data collection for the Baltic Sea V2 (
The in situ data in the period 2015–2019 were downloaded from ICES (International Council for the Exploration of the Sea) Oceanography CTD and bottle data (nowadays ICES Data Portal
Furthermore, to keep consistency with the other datasets, the uppermost available SSS measurements are used for this validation, lying in the range of 1–5 m depth.
The collocation strategy we follow for the comparison to in situ is the following:
Spatial collocation
Temporal collocation
For all the datasets, all the in situ data available in the 9 days (for L3 product) of SMOS data used to generate the product and in the same day (for L4 product) of the map are considered in the comparison.
The quality assessment of the SSS satellite retrievals results from the comparison against the reference datasets presented in Sect.
The following metrics are computed both for Baltic+ L3 and L4 SSS products:
The three satellite SSS products with the best temporal and spatial coverage (see Sect.
The objective of this assessment is to analyze the SSS dynamics captured by Baltic+ SSS products and the CMEMS Baltic reanalysis
Time series of Baltic+ L3 and L4 SSS products are analyzed and compared to the salinity provided by the CMEMS Baltic reanalysis and the in situ measurements. For that, we define boxes over given regions of interest, where both the reanalysis salinity and the Baltic+ L3 and L4 SSS products are averaged and compared to the in situ stations that are located in the region defined by each box. The boxes used for each region are shown in Fig.
Map of the in situ stations and the boxes used in the analysis of the time series per region. Red: Arkona Basin, grey: northern Baltic Proper, blue: Bothnian Sea, cyan: eastern Gotland Basin, pink: western Gotland Basin, green: Gulf of Finland, yellow: Gulf of Riga.
All the in situ measurements from the different ferry routes are analyzed per year. The statistics are computed considering all the collocations available for the Baltic+ L3 SSS product and FerryBox data (see Table
Global statistics for Baltic+ L3, L4 and filtered L4 (not considering extrapolated measurements from reanalysis) SSS products against FerryBox in situ data. Note the high variability in the number of match-ups is due to the different cruises operated each year.
To analyze the spatial distribution of the differences (
To analyze the spatial distribution without the effect of the non-homogeneous spatial sampling, the histograms of the spatial distributions of the mean and the standard deviation of
Comparison of L3 SSS and ferry data: spatial distribution of the mean of
Comparison of L3 SSS and ferry data: spatial distribution of the standard deviation of
Comparison of L3 SSS and ferry data: number of match-ups for each grid point in the map per year (from
Global statistics are also computed considering all the collocations available for the Baltic+ L4 SSS product and FerryBox data (see Table
We also compute global statistics of the collocations of Baltic+ L4 SSS and FerryBox data per year, considering only those Baltic+ L4 SSS that come from the L3 SSS (i.e., extrapolated data from reanalysis are filtered out) (see Table
The spatial differences between the L4 SSS and the SSS provided by ferry data are computed in
Comparison of L4 SSS and ferry data: spatial distribution of mean
Comparison of L4 SSS and ferry data: spatial distribution of the standard deviation of
Comparison of L4 SSS and ferry data: number of match-ups for each grid point in the map per year (from
Global statistics are computed considering all the collocations available for the Baltic+ L3 SSS product and SeaDataNet data per year (see Table
Global statistics of Baltic+ L3, L4 and filtered L4 (not considering extrapolated measurements from reanalysis) SSS products against SeaDataNet in situ data.
The spatial distribution of the SeaDataNet in situ measurements allows us to analyze
the performances of the Baltic+ L3 SSS product in the whole Baltic Basin and the influence of the proximity to land and ice edges in the quality of the Baltic+ SSS products. We compute the mean of
Comparison of L3 SSS and SDN: Spatial distribution of
Global statistics are computed considering all the collocations available for the Baltic+ L4 SSS product and SeaDataNet data per year (see Table
Global statistics are also computed considering all the collocations available for the Baltic+ L4 SSS product when the extrapolation of the reanalysis data is not considered and SeaDataNet data per year (see Table
The spatial distribution of the differences between the Baltic+ L4 SSS product (considering all the measurements) and the in situ data provided by SeaDataNet is also analyzed. For that, we compute the mean of
Comparison of L4 SSS and SDN, spatial distribution of
Maps of the estimated error standard deviations for each SSS dataset are shown in Fig.
Error standard deviations [psu] for the satellite SSS products computed by CTC for all the collocated maps in 2016:
The analysis of the Baltic+ L3 SSS product and the comparison with the other satellite products reveals that the Baltic+ L3 SSS product is currently the satellite-derived SSS product with the lowest salinity error among the currently available products, highlighting especially the improved spatial coverage and oceanographic resolution.
Figure
Spatiotemporal collocations of Baltic+ L3 SSS, Baltic+ L4 SSS, CMEMS Baltic reanalysis salinity fields with in situ salinity. Regions:
Baltic+ SSS products can be very useful to validate the models in areas where in situ data are sparse. Also, the location of the salinity gradients and their variability is valuable knowledge in evaluating the model performance. For example,
Access to the data is provided by the Barcelona Expert Center, through its FTP service. The DOI of the L3 product is
In this work, we present the first regional satellite-derived SSS maps over the Baltic Sea. To date, these are unique dedicated remote sensed SSS products available over the region, mainly due to the technical difficulties of retrieving SSS from satellite measurements over this basin. Several technical improvements have been required, the major ones being (i) the study of the dielectric constant models for the low-salinity regimes of the Baltic Sea, and (ii) the characterization of SMOS SSS systematic errors depending also on the SST. These improvements developed in the context of the Baltic+ Salinity Dynamics project have a clear impact on other regional initiatives (such as EO4SIBS (4000127237/19/I-EF) and SO-Fresh (4000134536) projects) and in the SSS retrieval from satellite L-band measurements in general.
Baltic+ SSS products have been proven to have a good spatiotemporal coverage with an accuracy of 0.7–0.8 psu for the L3 product (9 d,
Several scientific studies with Baltic+ SSS data are currently in progress, such as (i) the analysis of the consistency between the structures detected in the Baltic+ SSS products with the ones detected in the SST and in the DOT (Dynamic Ocean Topography) and (ii) the use of Baltic+ SSS time series as part of the HELCOM indicators to study the correlation between the SSS variability and the extreme events of different species in the Baltic Sea.
Interactions with the scientific communities working in the Baltic, and in particular with Baltic Earth Working Group on Salinity Dynamics, has allowed us to identify that Baltic+ SSS products can help fill some knowledge gaps
VGG generated the BEC product and is the main contributor to the writing of this paper. EO and CGH were the main contributors to the editing of this paper. VGG, EO, AT and JM were responsible for the conceptualization and development of the algorithms used in the generation of the product. CGH was responsible for the distribution of the products. The validation of the products was carried out by CGH, AGE, NH, MU, CG and VGG. PA and LT provided quality-controlled in situ data and participated in the discussions about the quality of the product and potential applications. MA and RC were responsible for the management of the project. DF and RS were the ESA project officers. All authors have reviewed the paper.
The contact author has declared that neither they nor their co-authors have any competing interests.
Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The authors would like to thank Andreas Lehman (Chair of the Baltic Earth Working Group on Salinity Dynamics, from GEOMAR Helmholtz-Zentrum für Ozeanforschung) for his valuable help on the definition of the scientific requirements and the scientific impact assessment of the products and to Klaus Getzlaff (also from GEOMAR Helmholtz-Zentrum für Ozeanforschung) for kindly providing the BSIOM hindcast simulation data.
They also would like to thank Jannica Haldin, Joni Kaitaranta, Kemal Pinarbasi and Owen Rowe (from The Baltic Marine Environment Protection Commission- HELCOM) for the fruitful discussions about the potential scientific applications of the Baltic+ SSS products and for integrating the seasonally averaged Baltic+ L4 SSS maps in HELCOM web map service. Last but not least, the authors want to acknowledge the anonymous reviewers and the editor for their valuable and helpful comments.
This work has been carried out as part of the Baltic+ Salinity Dynamics project (4000126102/18/I-BG), funded by the European Space Agency. It has been also supported in part by the Spanish R&D project INTERACT (PID2020-114623RB-C31), which is funded by MCIN/AEI/10.13039/501100011033. We also received funding from the Spanish government through the “Severo Ochoa Centre of Excellence” accreditation (CEX2019-000928-S). This work is a contribution to the CSIC Thematic Interdisciplinary Platform Teledetect.
This paper was edited by Giuseppe M. R. Manzella and reviewed by two anonymous referees.