In situ data

Historical SSS knowledge essentially comes from the analysis of single water samples and ther-mosalinograph (TSG) measurements on board research vessels and ships of opportunity. Since 2000, moored buoy arrays have been deployed for long term monitor of climatic events and

pro-Figure 2.3: Aquarius Satel lite

vide a reference data for remote sensing validation. This global network of offshore moorings is composed of NOAA/TAO(Tropical Atmosphere Ocean) array, JAMSTEC/TRITON (Triangle Trans-Ocean Buoy Network) array in the Tropical Pacific, PIRATA (Prediction and Research Moored Array in the Atlantic) array in the Atlantic, and RAMA (Research Moored Array for African-Asian-Australian Monsoon Analysis and Prediction) in the Indic Ocean. Additionally, the deployment of Argo automated devices for long term monitoring of the physical and bio-chemical properties of the ocean are increasing the coverage of in situ data. The Argo program has allowed the first global maps of in situ salinity maps (Figure2.5).

Argo is an international collaboration with more of 30 countries contributing to maintain a global array of more than 3500 free drifting profiling floats that measure temperature and salinity from the upper 2000 meters of the sea-ice free global ocean. The floats drift neutrally buoyant at a 1000 meters ’parking depth’. At 10-day intervals, the float sinks to 2000 m and then rises to the surface over 6 hours while measuring the vertical profile of T and S. The GPS determines the position and the float transmits the data via satellite. Then, the float sinks to their parking depth until the cycle is repeated. There are several float models: PROVOR and ARVOR are built by NKE-INSTRUMENTATION in France in collaboration with IFREMER, the APEX float are produced by Teledyne Webb Research, the SOLO float designed and built by Scripps Institution of Oceanography (USA).

The Argo salinity measurements are not taken, in most cases, not shallower than 3-5 m from the surface. Under most conditions (moderate to high winds) the surface ocean mixed layer extends much deeper, and the buoy provides an accurate measurement of the 1-2 cm of the surface layer that emits the microwave signal seen by the satellite. Therefore Argo floats provide a global quasi-SSS dataset considering the difference in the 1 cm and 5 m depths of the satellite and Argo measurements, respectively, is most of the time smaller than the expected

accuracy of Aquarius and SMOS SSS measurements. However, under tropical rain conditions, stratification leads to gradients between the surface and the buoy measurement depth higher of 0.1 pss (Drucker and Riser,2014).

The Argo array of profiling floats has a repeat cycle of one profile every 10 days and the current array extent provides a coverage of one profiler for an area of 300 km. Vinogradova and Ponte(2013) estimated the daily SSS mean differences between modeled and in situ SSS in boxes of 1 degree to be about 0.1, but greater than 0.2 in higher dynamical regions as near strong currents, outflows of major rivers and in coastal regions. When comparing in situ and satellite data it is also necessary to take into account that the horizontal SSS variability within the footprint of SSS remote sensing missions is typically larger than the vertical variations between the Argo and satellite estimate (Drucker and Riser, 2014). SSS variability with spatial scales smaller than the nominal 150 km footprint (in the case of SMOS), will contribute to differences between the satellite and in situ measurements. Therefore Argo is not only valuable to assess the quality of SMOS retrievals but also to highlight the differences between the two observation types.

Figure 2.4: Argo floats distribution for the September 6th 2015, 3881 active floats. Image source: Argo Program webpage

To validate the satellite products presented in this thesis we use global in situ data for the year 2012, both coming from TSG and Argo floats. The in situ salinity data used were collected and made freely available by the Coriolis project and programs that contribute to it (Coriolis project webpage). The Argo data used have been compiled, quality controlled and distributed to users through the French Global Data Assembly Center (GDAC) of Coriolis Project. Then, the profiles are selected following these criteria: (i) Only Delayed Mode data are being used.

(ii) Only primary CTD measurements (N-PROF=1) from each profile are used. (iii) Only data with the highest Quality Control value (QC=1) are used. (iv) The Quality Control of geographical position and date position is accepted if it has been set to 1 (good), 2 (probably good), 5 (Value changed) or 8 (Interpolated value). (v) Uppermost valid salinity measurement, without accepting any salinity value closer than 0.5 m to the surface, because some authors have observed erroneous measurements likely due to inaccuracies in the pressure measurement (Boutin et al.,2012b). (vi) Salinity data from PROVOR, SOLO instruments are not considered at depths shallower than 5 m, as these profilers types did not pump water at a depth shallower than 5 m (Wong et al.,2014).

A set of in situ data coming from the SPURS-MIDAS cruise (Las Palmas de Gran Canaria

16 March 2013 – Ponta Delgada, A¸cores 12 April) on board the Spanish R/V Sarmiento de Gamboa is also used. TheSPURS experiment(Salinity Processes in the Upper ocean Regional Study) aimed at understanding the processes that drive the upper ocean dynamics and the role that salinity plays on them in the area of maximum salinity in the center of the North Atlantic subtropical gyre. The experiment was coordinated by WHOI (R. Schmitt) and sponsored by NASA (E. Lindstrom), and includes intensive field work with a large variety of state-of-the-art instrumentation, the use of satellite remotely sensed salinity information (Aquarius and SMOS), as well as dedicated numerical modeling.

The Sarmiento de Gamboa was at the SPURS site together with the US R/V Endeavor in a coordinated program to continue the SPURS field work initiated by the French R/V Thalassa (STRASSE cruise, August-September 2012) and US R/V Knorr (September-October 2012). The main role of the Sarmiento de Gamboa cruise was to first run a general survey for mesoscale mapping of the area by means of thermosalinograph (TSG), Acoustic Doppler Current Profil-ers (ADCP) and Seasoar ondulating CTD (conductivity temperature depth sensor) while the Endeavor arrived at the site and serviced the moorings and gliders deployed in the September cruise. At the same time a total of 48 surface salinity drifters were deployed.

Figure 2.5: Schematics of SPURS field campaign measurementsSpurs experiment