1 Combining Argo and remote-sensing data in the North Atlantic Stephanie Guinehut, Gilles Larnicol and Pierre-Yves Le Traon CLS - Space Oceanography Division, 8-10 rue Hermès, 31526 Ramonville St Agne, France E-mail: Stephanie.Guinehut@cls.fr Despite the impressive increase of the number of temperature and salinity profiles from the Argo array, in-situ data still undersample the temporal and spatial variability of the ocean thermohaline structure. In contrast, remote-sensing measurements provide synoptic observations of sea level and sea surface temperature (SST) over the world ocean, but with no direct estimation of the ocean’s vertical structure. In order to reconstruct instantaneous temperature (T) fields at high temporal and spatial resolution, a merging method is developed to combine the accurate but sparse in-situ T profiles with the high-resolution but less accurate (as synthetic T profiles) altimeter and SST measurements. Three sources of data from the year 2002 are used. In-situ T profiles are from the Coriolis center 1 and correspond to real-time data including Argo profiling floats and also XBT and CTD measurements. Altimeter sea level anomalies are from the SSALTO/DUACS center 2 and are weekly combined maps of Jason-1 (Topex/POSEIDON), ERS-2 and GFO with a 1/3° Mercator horizontal resolution. SST data are from the NAVOCEANO center 3 and correspond to maps of weekly means of MCSST AVHRR with an 18-km horizontal resolution. The first step of the method consists in deriving synthetic T profiles from the surface down to 700-meter depth from altimeter and SST data through a multiple linear regression method. Pre- processing of altimeter SLAs includes the extraction of the steric part of the SLA using regression coefficients deduced from an altimeter/in-situ comparison study (Guinehut et al., 2002). Validation of the vertical projection of SLA and SST is performed on a subset of 3,500 T profiles from the year 2002. Results indicate that extracting the steric part of the SLA greatly reduces the differences between the in-situ reference profiles and the reconstructed synthetic profiles (Figure 1). The impact of SST is also clearly visible from the surface down to 300-meter depth where SST compensates the weak correlations between SLA and temperature, which means that SST is highly 1 http://www.coriolis.eu.org/coriolis 2 http://www.aviso.oceanobs.com/duacs 3 http://podaac.jpl.nasa.gov/navoceano_mcsst complementary to SLA when deriving T profiles from remote-sensing measurements. Figure 1: Rms of in-situ T anomalies (red) – anomalies calculated from the Levitus monthly mean climatology, rms of the differences between in-situ T fields and the synthetic profiles deduced from total SLA (green) and steric SLA (blue), and from a simple (SLA - dotted line) or a multiple (SLA+SST – unbroken line) regression method (in °C). The second step of the method consists in combining the synthetic profiles with in-situ T profiles using an optimal interpolation method (Bretherton et al., 1976). Analyses are performed at a weekly period on a 1/4° horizontal grid on each Levitus vertical level from the surface down to 700- meter depth. To gain maximum benefit from the qualities of both data sets, namely the accurate information given by in-situ T profiles and the mesoscale variability given by the T synthetic profiles, a precise statistical description of the errors of these observations must be introduced in the optimal interpolation method. For the in-situ profiles, since these observations are considered almost perfect, a very low white noise is applied. For the synthetic profiles, simulating remote- sensing (altimeter and SST) observations, since these observations are not direct measurements but are derived from the regression method, correlated errors have to be applied to correct long- wavelength errors or biases present in the synthetic fields and introduced by the regression method.