Improved Sea Surface Height from Satellite Altimetry in coastal zones: A case study in the Southern Patagonia.

LORELEY S. LAGO; MARTIN SARACENO; LAURA A. RUIZ ETCHEVERRY; MARCELLO PASSARO; FERNANDO OREIRO; ENRIQUE E. D'ONOFRIO

Florencia

Workshop; 10th Coastal Altimetry Workshop; 2017

ESA/CNES

In this work satellite altimetry data and its corrections terms are evaluated in a complex coastal environment. Altimetry data are compared with data obtained from a bottom pressure recorder (BPR) deployed during 22 months. The instrument is moored at 2.5km from the coast in San Matias Gulf, Argentina, at only 1.5km from the nominal intersection of satellite tracks 52 (descending, land-to-ocean transition) and 189 (ascending, ocean-to-land transition) of Jason 2. The 20-Hz S-GDR altimetry product is considerd to evaluate how close to the coast altimetry data are reliable. The BPR also provided the first long-term record of accurate sea level in the region, allowing a precise evaluation of tide models. We first compare how two retracking algorithms affect satellite altimetry data: we considered the classic Brown model (MLE4) and a more recent developed method: ALES (Adaptive Leading Edge Subwaveform Retracker). ALES has the ability to recover more data than MLE4 close to the coast, especially for the track that has a transition from land to ocean. Correlation between the 20-Hz S-GDR altimetry product and the in-situ dataset is 0.99 (95% Confidence Level) when all corrections except DAC and ocean tide are applied to the altimeter data, until a distance of 1.6km to the coast for track 189, and 4km for track 52. The large difference is explained by the fact that the satellite flies from land to ocean along track 52 while it flies from ocean to land along track 189. ALES and MLE4 show similar correlation with in-situ data when applied to satellite altimetry data for distances larger than 17km from the coast. Results also show that both solid earth tide and ionosphere corrections increase the correlation between altimetry and in-situ data near the coast: a correlation value of 0.9 is found at a distance from the coast of 4.1km (track 189) and 1.6km (track 52) when they are applied, instead of 4.6km and 8km respectively. Tide correction also bias the sea level anomaly constructed with satellite data, no matter which tide model is considered. Among the three global and one regional tidal models considered, the model with the lower root sum square of the difference (RSS) is the regional one (TPXO8, 4.8cm). Yet the lowest difference with in-situ tidal constituents is obtained by harmonic analysis of the complete (1992-2014) altimetry data set (RSS 4.1cm) highlihtying the potential of altimetry data to compute tides. Considering data from both ascending and descending tracks we finally show that using ALES and TPXO8 tidal model it is possible to construct a sea level anomaly with a root mean square difference of 13cm as close as 4km from the coast.