For geophysical purposes, gravity is measured in many ways, from static-point observations, using a gravimeter, to mean-value determinations from gravimeter and gravity gradiometer data collected by airplanes, ships, and satellites. We tested estimates of vertical and horizontal components of the gravity vector by combining Global Positioning System (GPS) data with a Honeywell H764G inertial navigation system (INS) on a land vehicle traversing highways in southwestern Montana. The estimation methods were based on techniques applied successfully to airborne INS/GPS data. In addition, we used wavelet denoising and wavenumber correlation procedures to enhance the estimates. Analyses of multiple traverses along the roads verified levels of repeatability as good as 0.64mGal (all numerical accuracy values refer to standard deviations) in the vertical gravity-disturbance component. Control data, interpolated onto each road segment from an available database of gravity values, had an accuracy better than 2–4mGal. Compared with this interpolated control, our vertical gravity estimates compare as well as 2–3mGal. Resolution of the estimated vertical component is about 2.2km, based on 180-s data smoothing and a vehicle averaging about 90km/hr. Large scale-factor errors exist in the horizontal gravity estimates. Removing those on the basis of extensive deflection-of-the-vertical (DOV) control yields repeatability in the horizontal components of 2–15mGal and agreement with the control at 5–9mGal. Our tests confirm that a land-vehicle INS/GPS system is capable of along-track gravity mapping with precision and resolution that have used in local geophysical modeling (e.g., continental rifts) as well as in mineral exploration.

The current official North American Vertical Datum of 1988 (NAVD 88) and the International Great Lakes Datum of 1985 (IGLD 85) will be replaced by a new geoid‐based vertical datum in 2022. The Gravity for the Redefinition of the American Vertical Datum (GRAV‐D) project collects high‐quality airborne gravity data to improve the quality of the gravitational model that underpins the geoid model. This paper validates the contribution of GRAV‐D data in the Great Lakes region. Using the lake surface height measured by satellite altimetry as an independent data set, Global Gravity Models (GGMs) with/without the GRAV‐D data are compared. The comparisons show that the improvement reaches decimeters over Lake Michigan where the historic gravity data have significant errors. Over all lakes, except Lake Erie, the GRAV‐D data improve the accuracy of the gravitational model to 1–3 cm.