Geomed2: High-Resolution Geoid Models of the Mediterranean

Sean Bruinsma (CNES, France)

Event: 2019 Ocean Surface Topography Science Team Meeting

Session: The Geoid, Mean Sea Surfaces and Mean Dynamic Topography

Presentation type: Type Poster

The GEOMED2 project started in 2014 with the aim of estimating the geoid of the Mediterranean. In this work, an overall view on the GEOMED2 project is presented and the different steps in managing the project are revised and commented. The way gravity data has been checked for possible outliers and biases is discussed and a comparison among Global Geopotential Models (GGM) used for reducing the low frequency component of the gravity signal is given. The different methods for terrain effect computation (RTC) and the outcomes of the applied geoid estimation methods, i.e. Collocation, Stokes-Wong&Gore and the KTH methods are presented and commented.
Marine gravity data is not available for large parts of the Mediterranean and consequently the gravimetric geoid solution is significantly less accurate there. Gravity inferred from altimetry data can be used to fill the gaps. However, ocean dynamic signal may contaminate the derived gravity or geoid, which is why a pure gravimetric solution is preferred in an ideal world.
Geoid solutions using altimeter-inferred data only, such as DTU15 and DTU17 and UCSD V24 and V27 gravity, were computed too. These solutions were computed using the Stokes-Wong&Gore method. By interpolating the altimeter-inferred data on the ship measurements’ positions, i.e. replacing ship data, and then computing the geoid solution allowed drawing surprising solutions on the quality of both datasets.
All models are compared with a marine geoid constructed with models for the MSS and the MDT (based on drifter data), which is our benchmark. The difference in geoid quality due to geoid estimation method is rather small, fortunately, but the impact of the GGM used in the remove-restore procedure is very big. The RTC method equally has a big impact, and best results are obtained with the Kuehn-Hirt model.
Comparing model-inferred to drifter-inferred current velocities constitutes an independent quality evaluation. This type of evaluation leads to a very detailed quality assessment of the models, notably as a function of spatial scale.

 

Poster show times:

Room Start Date End Date
The Gallery Tue, Oct 22 2019,16:15 Tue, Oct 22 2019,18:00
The Gallery Thu, Oct 24 2019,14:00 Thu, Oct 24 2019,15:45
Sean Bruinsma
CNES
France
sean.bruinsma@cnes.fr