Investigations of a Suspected Jump in Swedish Repeated Absolute Gravity Time Series

Since 2006 the gravity change at 17 points in Sweden has been observed with Lantmateriet's absolute gravimeter FG5-233. The main purpose of the observations is to study the postglacial rebound in Fennoscandia. In 2010, a suspected jump of a few mu Gal can be seen in the gravity time series that significantly affects the estimated gravity rate of changes. It is shown that if the jump is not considered, then the absolute value of the gravity rate of change is systematically underestimated compared to the land uplift model NKG2014LU_test. In this paper two different ways to estimate and apply corrections of the jump are demonstrated. The first is to estimate the jump from the observations themselves within a least squares adjustment, while the second is to assume the instrument bias obtained in international comparisons of absolute gravimeters. The best agreement between land uplift model and estimated rates of change of gravity is achieved by correcting the data with the official biases reported from the international comparisons.