Superconducting gravimeter and seismometer shedding light on FG5’s offsets, trends and noise: what observations at Onsala Space Observatory can tell us

Journal article, 2020

Ten-year worth of absolute gravity (AG) campaigns at Onsala Space Observatory (OSO), Sweden, are simultaneously reduced
using synchronous data from a superconducting gravimeter (SG). In this multi-campaign adjustment, the a priori models
commonly applied for each setup in AG-alone experiments are sidestepped in favour of SG records and a model to estimate
its drift. We obtain a residual (hourly samples) at the 5 nm/s2 RMS level, reducing the SG data with a range of ancillary data
for the site’s exposure to ocean and atmospheric loading, and hydrology effects. The target quantity in AG projects in the
Baltic Shield area is the secular change of gravity dominated by glacial isostatic adjustment with land uplift as its major part.
Investigating into the details of the associated processes using AG requires a long-term stable reference, which is the aim of
international comparison campaigns of FG5 instruments. Two of these have been campaigning at OSO since 2009 when the
SG had been installed. In the simultaneous inversion of all sixteen campaigns, we identify weaknesses of AG observations,
like varying systematic offsets over time, excess microseismic sensitivity, trends in the AG data and side effects on the SG’s
scale factor when campaigns are evaluated one by one. The simultaneous adjustment afforded us an SG scale factor very
near the result from a campaign with a prototype quantum gravimeter.Whence, we propose that single-campaign results may
be biased and conjectures into their variation, let alone its causes misleading. The OSO site appears to present manageable
problems as far as environmental influences are concerned. Our findings advocate the use of AG instruments and procedures
that are more long-term stable (reference realization), more short-term stable too (setup drifts), less service craving and more
resilient to microseismic noise.