Validation and fusion of different databases in preparation of high-resolution geoid determination

An up to date determination of a high-resolution geoid requires the use of best available databases concerning digital terrain model (DTM), bathymetry, global geopotential model and gravity field. The occasion to revisit methods to validate and merge different data sets has been created by a new project for the determination of a new European Geoid.
Since the computation of the latest European geoid and quasi-geoid model (EGG97), significant new or improved data sets have become available, such as new global geopotential models from CHAMP and GRACE missions, new national and global DTMs and new or upgraded gravity data sets.
In the context of the new European Gravity and Geoid Project (EGGP), within the IAG Commission 2, some data validation tests have been performed in the Italian zone.
In the area 19°× 17° wide, covering Italy, three kinds of tests have been performed: comparison among different DTMs in order to choose the best one to be used; comparisons in terms of geoid computation in some coastal areas, to evaluate bathymetry effects, and the validation of the EIGEN-CG01C and EIGEN-CG03C new global models up to degree and order 360.
These preliminary tests lead to the choice of SRTM DTM (integrated in no-data holes), with an added bathymetry derived by the Italian 1:25 000 official cartography near the coasts and the NOAA bathymetry in high seas. The validation of the new global models and the comparison with EGM96 model show that, in terms of geoid computation, the EGM96 yields better results. Moreover, the validation of new available land gravity data and the cross-validation of two sets of gravity data on sea have been completed.     

The reduction of aliasing in gravity anomalies and geoid heights using digital terrain data

Observations of gravity can be aliased by virtue of the logistics involved in collecting these data in the field. For instance, gravity measurements are often made in more accessible lowland areas where there are roads and tracks, thus omitting areas of higher relief in between. The gravimetric determination of the geoid requires mean terrain-corrected free-air anomalies; however, anomalies based only on the observations in lowland regions are not necessarily representative of the true mean value over the topography. A five-stage approach is taken that uses a digital elevation model, which provides a more accurate representation of the topography than the gravity observation elevations, to reduce the unrepresentative sampling in the gravity observations. When using this approach with the Australian digital elevation model, the terrain-corrected free-air anomalies generated from the Australian gravity data base change by between 77.075 and −84.335 mgal (−0.193 mgal mean and 2.687 mgal standard deviation). Subsequent gravimetric geoid computations are used to illustrate the effect of aliasing in the Australian gravity data upon the geoid. The difference between 'aliased' and 'non-aliased' gravimetric geoid solutions varies by between 0.732 and −1.816 m (−0.058 m mean and 0.122 m standard deviation). Based on these conceptual arguments and numerical results, it is recommended that supplementary digital elevation information be included during the estimation of mean gravity anomalies prior to the computation of a gravimetric geoid model.