<p>The GRACE mascon solutions, called the advanced products of GRACE data, are widely used in cryosphere science or hydrology research. It has been demonstrated that the mascon solutions have the same or better performances compared with the spherical harmonic (SH) solutions at the basin scale. However, although the mascon solutions are expected to have the ability to recover the transient gravity signals due to large earthquakes, few studies have investigated the performances of the mascon solutions in studying seismic deformations systematically. In this study, we attempt to examine the performances of the mascon solutions for transient gravity signals induced by three M9 class earthquakes: the 2011 Tohoku-Oki, 2004 Sumatra and 2010 Chile earthquakes, and compare them with the SH solutions and theoretical gravity changes modelled by dislocation theory. We analyse the co-seismic gravity changes and conclude that the mascon solutions contain almost identical information as the SH solutions and can retrieve the co-seismic gravity change signals in the resolutions equivalent to the Gaussian filter radii of 210~270 km. However, the mascon solutions have other strengthening gravity change signals, with magnitudes that are the same order as that of the SH solutions and contain pre-seismic gravity change signals in the Tohoku earthquake. These strengthening and pre-seismic signals are both considered artificially introduced noise due to the mathematical treatment before releasing rather than real geophysical signals.</p>
Seasonal gravity changes estimated from GRACE data
