Concentrated Slip and Low Rupture Velocity for the May 20, 2012, M W 5.8, Po Plain (Northern Italy) Earthquake Revealed From the Analysis of Source Time Functions

We analyse the rupture properties of the May 20, 2012, MW 5.8, Po Plain (Northern Italy)
earthquake by using two different modeling procedures based on the source time functions: a forward
modeling and a global inversion Bayesian method. While the forward modeling allows to retrieve general
information on the source characteristics, the global inversion allows to explore a substantially larger
number of possible solutions, with more parameters, providing a quantitative estimate of the misfit. We
invert for the spatial slip distribution and for the rupture velocity on a planar fault model. The unknown
slip is given at the nodes of the subfaults (control points) and then given at the elementary subfaults
through a bilinear interpolation. The number of control points is progressively increased to move from
a high- to low-wavelength description of final slip on the fault plane. The optimal model parameter set
is chosen according to the Akaike Information Criterion. The uncertainty on the slip distribution and
rupture velocity has been estimated by a statistical analysis of the model ensemble and, in particular,
through the weighted mean model and the standard deviation.
We find that the most earthquake slip occurred in the regions located northeast and southwest of
the hypocenter, consistent with the forward modeling. Moreover, we find a low rupture propagation
velocity (0.4 compressional Mach number) similarly to what has been observed for the close 29 May,
MW 5.6, and radiation efficiency suggesting that half of the strain energy was used to create new
fracture.