Analysis of Errors in Gravity Derived from the FALCON Airborne Gravity Gradiometer

FALCON airborne gravity gradient data can be converted to vertical gravity via direct integration techniques (Gunn, 1975) or equivalent source methods (Dampney, 1969), and hence compared with conventional gravimetry data. Either conversion process amplifies long wavelength features from the gravity gradient data and therefore makes regional scale features vulnerable to distortion. Recent FALCON surveys over areas with pre-existing ground gravity have allowed proper assessment of when conversion problems may arise. Comparison of FALCON and ground data from three different areas indicated that derived vertical gravity is very good at wavelengths less than the minimum survey dimension, spectral analysis showing errors are typically 0.1 mGal / km for wavelengths below this limit. A similar analysis by Bruton et al. (2001), of profile gravity data from three airborne gravimetry systems, showed error of spectral character different to that achievable via FALCON, with steadily decreasing contribution to the total error from longer wavelengths. This independent analysis of the airborne gravity systems indicated that the limiting wavelength, beyond which airborne gravimetry profiles becomes more accurate than a detailed FALCON gradiometry survey data, is in the range 10-20 km. Regional-scale gravity databases, sparse marine track data, or satellite-derived gravity can be used to conform the FALCON gravity to the regional data beyond a certain wavelength. This overcomes the limitation of FALCON gravity at long wavelengths. Given typical survey dimensions of 10-100 km, and provided that the reliable range of wavelengths in the FALCON and regional data overlap sufficiently, it is possible to derive gravity data having errors of approximately 0.1 mGal / km and sampling all features with wavelengths greater than 300 m. Introduction The FALCON airborne gravity gradiometer (AGG) measures the two horizontal differential curvature gradients (Lee, 2001), quantities also measured by the original Eötvös torsion balances (Heiland, 1968; Telford, 1990). Within a north, east, down coordinate system and with subscripts denoting derivatives, these quantities are second derivatives of the gravitation potential referred to as GNE and GUV = ((GNN GEE)/2). Differential curvature gradients measure the degree of distortion of the gravitational equipotential surface from a spherical shape. Traditional gravimeters, on the other hand, measure the vertical acceleration due to gravity (which we denote by gD), arguably a less abstract quantity for interpretation. Gravimeters and gravity gradiometers both measure derivatives of the gravitational potential. Well known algorithms allow conversion between these various derivatives to produce theoretically equivalent quantities. Such conversions, based on integration of the gravity gradients and equivalent source techniques, are applied routinely to FALCON gravity gradient data to derive the more familiar vertical