It  has  been  showed  by  several  researchers (Harris, 1998; King et al., 1994; Stein et al., 1993) that  small  static  stress  changes  due to  permanent fault displacement can alter  the  likelihood  of, or  trigger,  earthquakes  on  nearby  faults. 

For   example   King et al. (1994)  discovered  that  the  spatial  distribution  of aftershocks   for   the  Landers  earthquake  (1992) could be  explained by  the  Coulomb  criterion:

Places,where the Coulomb stress on optimally oriented faults increased by more than 0.5 bar showed an increased aftershock frequency, and aftershocks were sparse where the Coulomb stress dropped by a similar  amount. This was a very good example for the fact  that  in some cases, the analysis of Coulomb  stress  changes  seems to be a good  indicator for the  appearance of  aftershocks  or  future  events.

Difference in
Coulomb
failure
stress
change
between
model 1
and
model 20
(data set
from
Sudhaus &
Jonsson
2008)

The Coulomb stress change calculations in the  previous  studies  (for example Stein et al. 1993) are based on fault parameters without considering any errors. In this paper we want to investigate the influence of fault parameter errors on the Coulomb stress change  calculations   by    incorporating  the  fault  model  errors. We perform this investigation   in  a  case  study    of  the  June   2000  Kleifarvatn  earthquake   that served as the underlying subject in the study of Sudhaus and Jonsson 2008.

Kleifarvatn fault
mechanism:
almost N-S-
striking right
lateral strike
slip

We use 2500 sets of model parameters of the Kleifarvatn  earthquake provided by the paper of Sudhaus and Jonsson (2008). The variability of  these  models represent the posterior probability density distribution for the Kleifarvatn fault model. With these model parameters in conjunction with the software disloc3d and Coulomb3.1 we intend to calculate not only the Coulomb stress change of  the Kleifarvatn  event for the best model  but also incorporate the model varia- bility to  reveal  the  influence of  errors in the models on the Coulomb stress patterns. The  results  shall  lead  us to a better understanding of the need to incorporate errors in the Coulomb calculations.

Standard deviation
of Coulomb failure
stress change
distribution
(data set:
Sudhaus &
Jonsson 2008,
first 100 models)
UTM
coordinates