The determination of the gravity field, together with knowledge of the surface topography, provides one of the primary means of inferring the density structure and dynamics of planetary interiors [
Phillips and Lambeck, 1980; Hager, 1985]. After removing the topographic gravity signal, the distribution of internal density anomalies caused by thermal or compositional differences can be studied. Gravity field models can also be used to study the compensation of surface topography, which can provide information on the mechanical properties and state of stress of the lithosphere.
The determination of these models has been accomplished using a wide variety of different measurement types and solution techniques. The near‐Earth measurement types can be divided into satellite tracking measurements, surface gravity measurements, and satellite altimeter measurements. In both near‐Earth and interplanetary contexts, tracking data are used to measure gravitational perturbations affecting satellites, with the accuracy and spatial/temporal distribution of the data being the most important factors in the resulting field accuracy and resolution. Surface gravity data provide a more direct measurement of the gravity field, but acquiring data uniformly over the Earth has always been difficult. Satellite altimetry provides precise measurements of the marine gravity field, provided that satellite orbit errors and non‐geoidal sea surface height variations can be adequately modeled. Comprehensive gravity field solutions must incorporate these disparate data types in order to estimate mathematical parameters describing the gravity field, such as spherical harmonic coefficients. The proper combination of these data, along with the desire to acquire an accurate representation of the model errors, require that complex procedures be developed for computing these solutions.