Investigation of the contribution of topographic effects on regional geoid modeling within the Geomed2 project

With nowadays high resolution Global Geopotential Models (GGMs), traditional geoid approximation with the Remove-Compute-Restore (RCR) procedure needs revisiting both in terms of the numerical and the methodological steps followed. This becomes evident if one considers the high degree of expansion of the latest combined GGMs, corresponding to frequencies as short as 9.2 km for a maximum degree of expansion equal to 2160, and the limited spatial resolution of bathymetry models, being 30 arcsec (~1 km) for the SRTMPlus model. In the above schema, if a high-resolution GGM is used to reduce the available surface gravity data, then the bathymetry model used for the terrain reduction should be able to represent adequately the spatial frequencies between 10 and 1 km, so that the corresponding gravity signal presents in the reduced gravity anomalies can be removed. This step is quite significant since depending on the correctness, resolution and accuracy of the bathymetry model, the resulting residual gravity anomalies can be either useful for geoid determination, or can introduce biases, noise exaggeration and aliasing in the predicted gravimetric geoid model. Within the GEOMED 2 project, which aims at a high resolution and accuracy geoid determination for the entire Mediterranean basin, in situ gravity anomalies are used within the RCR procedure employing DIR-R5 and EIGEN6c4 as reference geopotential models and various bathymetry/topography models (SRTMPlus, DTU2013, MISTRAL and EMODNET) in order to investigate the influence of the used terrain representation on the gravity and geoid signal. The used Digital Terrain and Bathymetry Models (DTBMs) are of various resolutions, smoothness, and accuracy, while some of them are directly derived from satellite data and some of them from in situ echo sounding observations. In this work the Residual Terrain Model (RTM) reduction is used to model the contribution of topography/bathymetry to gravity and the geoid, while the contribution of each model and its appropriateness for use in geoid modeling are investigated both in the frequency and the space domain.