The data is a CSV table containing the following columns: station: station ID order: station order point: point ID method: gravity determination method ('absolute' for primary points, 'relative' for secondary points) station_name: point fullname latitude: approximate latitude longitude: approximate longitude elevation: approximate elevation gravity_eff: gravity at effective position ste_eff: standard error of gravity at effective position h_eff: height of effective position tie: gravity differences from the primary (absolute) point ste_tie: standard error of gravity differences a: first-degree coefficient of vertical gravity gradient approximation b: second-degree coefficient of vertical gravity gradient approximation ua: standard error of the first-degree coefficient ub: standard error of the second-degree coefficient covab: correlation coefficient gravity: final gravity value, which is related to point benchmark total_uncert: total uncertainty of the final gravity value expanded_uncert: total expanded uncertainty of the final gravity value

This version supersedes all previous ones. Previous versions contained errors in the computation of "gravity" column,