TY  - JOUR
ID  - epos1590
UR  - http://dx.doi.org/10.1093/gji/ggv083
IS  - 3
A1  - Debski, Wojciech
N2  - The spatial location of sources of seismic waves is one of the first tasks when transient waves from natural (uncontrolled) sources are analysed in many branches of physics, including seismology, oceanology, to name a few. Source activity and its spatial variability in time, the geometry of recording network, the complexity and heterogeneity of wave velocity distribution are all factors influencing the performance of location algorithms and accuracy of the achieved results. Although estimating of the earthquake foci location is relatively simple, a quantitative estimation of the location accuracy is really a challenging task even if the probabilistic inverse method is used because it requires knowledge of statistics of observational, modelling and a priori uncertainties. In this paper, we addressed this task when statistics of observational and/or modelling errors are unknown. This common situation requires introduction of a priori constraints on the likelihood (misfit) function which significantly influence the estimated errors. Based on the results of an analysis of 120 seismic events from the Rudna copper mine operating in southwestern Poland, we propose an approach based on an analysis of Shanon's entropy calculated for the a posteriori distribution. We show that this meta-characteristic of the a posteriori distribution carries some information on uncertainties of the solution found.
VL  - 201
TI  - Using meta-information of a posteriori Bayesian solutions of the hypocentre location task for improving accuracy of location error estimation
AV  - none
EP  - 1408
Y1  - 2015///
PB  - Oxford University Press
JF  - Geophysical Journal International
KW  - Inverse theory
KW  -  computational seismology
KW  -  theoretical seismology
SN  - 0956-540X
SP  - 1399
ER  -