High resolution coastal modeling in Kerguelen Island for CAL/VAL operations.

Yann-Treden Tranchant (Littoral ENvironnement et Sociétés, UMR 7266 LIENSs CNRS, France)


Mathilde CANCET (Noveltis, France); Etienne SAHUC (Noveltis, France); Simon MILLET (Noveltis, France); Florent LYARD (OMP/CNRS/LEGOS, France); Laurent TESTUT (OMP/CNRS/LEGOS, France); Valerie BALLU (Littoral ENvironnement et Sociétés, UMR 7266 LIENSs CNRS, France); Gérald DIBARBOURE (CNES, France); Nicolas PICOT (CNES, France); Olivier LAURAIN (OCA-GEOAZUR, France); Pascal BONNEFOND (Observatoire de Paris – SYRTE, France)

Event: 2020 Ocean Surface Topography Science Team Meeting (virtual)

Session: Coastal Altimetry

Presentation type: Type Forum

The sea surface height (SSH) measured by the satellite altimeters contains the contribution of the geoid, the dynamic topography and the ocean’s response to tides and atmospheric forcing. Calibrating and validating past and future altimetry missions, such as SWOT, implies to significantly increase the resolution and accuracy of local geoids.

In the frame of the FOAM project (From Ocean to inland waters Altimetry Monitoring), CalNaGeo, a geodetic GPS on a towed blanket, was developed to map the sea surface. Two CalNaGeo campaigns were carried out in 2016 and 2018 in Kerguelen Island, near historical tide gauges and Jason-2&3, Sentinel-3A and SARAL/AltiKa altimetry tracks. In order to recover the geoid contribution in the total SSH obtained from a precise GNSS survey, the contributions of the other physical processes have to be evaluated. Due to the non-colocation of permanent in-situ observations with the CalNaGeo track, a high-resolution model is needed to correct the GPS measurements and isolate the marine geoid.

To meet this objective, we have implemented a high-resolution barotropic modeling configuration in the Kerguelen Island region, which we run with two different models: SCHISM and TUGO-m. Each model has specificities in terms of ocean processes approximations and comparing the simulation results enable to identify the driving processes in the area. This study is based on an unstructured grid with a resolution ranging from 10km in the deeper ocean to 100m near the coastline, providing a coastal/open ocean continuum. Both models are forced by tides from the global tide model FES2014.

After harmonic analysis, the simulations are compared with coastal altimetry products (along track observations from Topex/Jason1/Jason2 missions) and tide gauge observations. An indicator of the model stability to remove the non-stationary part of the sea level in CalNaGeo measurements is to investigate the differences at the crossover points before and after applying the model correction. The RMSE of these errors is computed for 2 areas, which correspond to TP/Jason and SARAL/Altika calibration areas. A comparison is made with FES2014 in order to assess the benefit of a regional model on a global tide atlas.

After removing the simulated SSH to CalNaGeo measurements, the first results are promising, highlighting a strong correlation between time-invariant residuals and geoid patterns.
Yann-Treden Tranchant
Littoral ENvironnement et Sociétés, UMR 7266 LIENSs CNRS