To give you the best possible experience, this site uses cookies. Continuing to use this site means you agree
to our use of cookies. If you'd like to learn more about the cookies we use please find out more


 project header 2


Project name: Observing System Experiment of Satellite SSS & El Niño 2015/16

Acronym: SMOS-Nino15

Duration: 24 months

Project end: Sep 2018

Principal Investigator and Co-PIs: Benoit Tranchant (CLS), Gilles Larnicol (CLS), Eric Greiner (CLS), Elisabeth Remy (Mercator Océan), Matt Martin (Met Office), Rob King (Met Office) and Kirsten Wilmer-Becker (Met Office)

Website: https://www.godae-oceanview.org/projects/smos-nino15/


Project objectives & scientific and technical background

The main objectives of this project are:

OBJ-1To assess and analyse the onset and evolution of the 2015/16 El Niño event
OBJ-2To prepare the Observing System Experiment;
OBJ-3To make appropriate simulations
OBJ-4To refine the GODAE requirements for SSS (sea surface salinity) after the analysis of the errors associated with these experiments
OBJ-5To produce an Observing System Impact Statement on behalf of the GOV OSEval-TT
OBJ-6To create a data base containing the input and output data from the OSE and make this publicly available through the project web-site
OBJ-7To prepare and co-ordinate the submission of, peer-reviewed journal articles detailing the impact of satellite SSS observations in each operational ocean forecasting systems
OBJ-8To promote the SMOS-Niño15 project and the use of ESA SMOS data



One scientific question for this project is the understanding of the mechanisms at work during one of the strongest El-Niño events in recent years. Although the 1997/1998 event was similarly strong the ARGO array and the satellite SSS were not available at the time, providing an opportunity now to progress in the closure of the water budget. This is a major issue because the water cycle is impacted by anthropogenic change. There is also evidence from previous studies that salinity can play a significant role in the evolution of El Nino events, and that improved initialisation of salinity in ocean models could improve the forecast skill.

The assimilation of satellite SSS observations is still a difficult challenge because of the various and complex biases that affect it. Data assimilation systems do not have yet all the ingredients to benefit from this data. The main issue is the bias between the upper SSS seen by satellite, and the salinity within the mixed layer seen by ARGO. This study allows new solutions to be tested in the real time systems. The analysis of the improvement and/or degradation in an Observing System Experiment framework will provide valuable feedback to the data centres as well as to the satellite designers. The use of SMAP and/or AQUARIUS will help inform the setting of future requirements for satellite SSS and to verify that different satellite configurations produce different results.

The Sea Surface Salinity (SSS) is a key component for determining the global water balance. More generally, salinity variations play a crucial role in the upper tropical ocean dynamic. The upper ocean buoyancy accumulation from excess rainfall in the western Pacific is a significant factor in the El-Niño dynamics [RD 4][RD 5].

The SSS has been measured from space for the past 6 years by the SMOS mission (ESA), which is still returning observations. The Aquarius mission (NASA) has delivered almost 4 years of SSS (August, 2011 to June, 2015) and now, the SMAP mission (NASA) has been providing SSS since April 2015 [RD 6].

Before the launch of the SMOS mission, early studies had shown the potential benefit of assimilating SSS from space [RD 7][RD 8][RD 9]. Since the launch of SMOS, different attempts have shown that it was difficult to assimilate satellite SSS data into an operational ocean forecasting system [RD 10][RD 11]. The difference between the forecast and the SSS can be 5 times larger than the difference between the forecast and near surface ARGO salinity [RD 12]. The signal to noise ratio is not high today, and data and methods must be improved. Nevertheless, several studies [RD 1][RD 2][RD 3] show that SSS measured from space can bring new information. More precisely, [RD 13] and [RD 12] demonstrate the ability of SMOS data to identify mesoscale structures in the Gulf Stream that correspond to features measured by sea level satellite data in the summer months. More generally, when the surface waters are warm, a sharper contrast may be seen on the satellite SSS than on the surface temperature. A typical example is the retroflection of the North Brazil Current: Amazon river discharges plus precipitations from the Inter Tropical Convergence Zone produce high-contrast structures in SSS fields [RD 2]. Similar contrasts exist in the tropical Pacific near the fresh water pool and near Panama. The resolution of the satellite SSS is expected to be valuable near river mouths and in the equatorial wave guide where the resolution of the ARGO array is reduced.

Beyond regional improvements, it has been shown [RD 14] that the assimilation of AQUARIUS SSS gives a significant improvement for all ENSO forecast lead times after 5 months. Hence, satellite SSS can be complementary to satellite SST and SSH, and it can fill the gaps in the ARGO array. Benefits are expected for real time analysis up to seasonal forecast.

The onset of the El-Niño 2015/16 event is evident in the SSS from both the Mercator Océan and Met Office systems (see Figure 1). The most interesting recent period is the year 2015. In early 2014 the El-Niño conditions were neutral and a weak El-Niño developed through winter 2014/15, which may have preconditioned the on-set of the 2015/16 El-Niño. The El-Niño then became significant in late spring 2015, peaking in December 2015 and returning to neutral in late spring 2016. We therefore propose to work on the 2014-early 2016 period.

SSS along equator 2013-2015
















Figure 1: Hovmöller diagrams of salinity along the Equator in the Pacific from the Mercator- Ocean Psy3v3r3 real time system (left) and from the Met Office FOAM system (right). The freshening increases in the West from 2013 to 2015.

Figure 2 [RD 15] shows the trend in SSS over recent decades. The salty North Atlantic gets saltier whereas the fresh pool of the Western Pacific gets fresher due to increased precipitation near the dateline. This freshening is the fingerprint of anthropogenic warming [RD 15]. The point is the El- Niño 2015/16 event produces a freshening which is superimposed on this slow signal. It is not too surprising since the decadal trend could be the result of increased occurrences of central El-Niño (Modoki).

The underlying point is that SMOS, as a gap filler, could also help to monitor more precisely than ARGO alone the fingerprint of anthropogenic change in the tropical oceans. The period considered is certainly too short to conclusively demonstrate this, but one must consider the impact on the forecast both in terms of the meso-scale, and in terms of the very large scale.

The drawback with satellite SSS is that it suffers from various and complex sources of biases and errors [RD 6]. In order to assimilate SSS data from SMOS, [RD 12] show the importance of using unbiased satellite SSS data while using rigorous quality control in an upstream process. To date, there are various satellite SSS data sources, with Level 2 and 3 products, ascending or/and descending orbits, from daily to monthly products, with/without observation based bias corrections. A broad set of data sources can be used in the OSEs planned as part of this project. The experiments will be carried out for 2015 with the traditional data sources, and with and without assimilation of various satellite SMOS data sources. The experiments will be complemented by the use of Aquarius and/or SMAP data (from early 2015) depending on the limitations of the datasets.

Observed SSS trends and means

Figure 2: Fig. 1 from [RD 18]: Observed SSS trends and means. (a) The 33-yr (1970–2002) linear surface salinity trend (century21) computed from monthly anomalies. White grid boxes indicate regions with insufficient or no data and stippling denotes areas where trends are statistically significant from 0 at the 5% level using a two-sided Student’s t test. (b) Climatological mean observed surface salinities estimated over 1950–2008 and 1970–2002 for the Pacific and Atlantic, respectively. Geographical masks were applied in order to prevent mixing of data across the land boundaries.


In order to explore the various potential impact of the SSS in the forecasting systems, experiments shall be performed using different forecasting systems (several centres will be involved), evaluating performance on short term forecast (7 days) or mid-term forecasts (monthly).

These experiments will follow GOVST community best practices as described in [RD 16], [RD 17] and [RD 18] and will be fully designed during the course of the project (task 3). In particular, traditional assessment methodologies as used in the GOVST community (RD 19, RD 20) will be set up to assess the results of these experiments.

At the end of this set of experiments, feedback will be provided to the satellite SSS data centres. Typical information is the impact of the new data types on the level of bias and error diagnosed by the data assimilative systems. This feedback will be consolidated into Observation Impact Statement Reports, such as those of [RD 21], and will be written up as journal papers. Also, the experiments will be used to assess and update the requirements for satellite SSS data as set out by GOVST.





Home Contact Login CLS Met Office Mercator Ocean