Jökull - 01.01.2010, Side 82

Horálek and Fischer
1E-3
1E-2
1E-1
1E0
1E1
MPa
-1.0 -0.5 0 0.5 1.0 kma) b) c)
Figure 11. Angular dependence of the rate of IA evaluated using a step of 10◦; all aftershocks in a); aftershocks
linked by speed higher than 100 m/s in c). The slip axis is indicated by a dash line. Space-time distribution of
the complete stress filed (dynamic and static) resolved on the fault plane surrounding the rupture induced by an
instantaneous stress drop of 10 MPa due to strike slip on a circular area with radius of 100 m in a homogeneous
half-space. The method of Bouchon (1997) was used to calculate the stress field. Only the shear stress vector
(traction) is depicted because the normal stress is zero. The stress field breaks with time into the permanent
(static) part as a result of the near-field deformation and the transient (dynamic) part carried by the propagation
carried by the transition of seismic waves. – Þéttleiki stefnunnar á milli hvers skjálfta og eftirskjálfta hans
sem fall af snúningshorni í sameiginlegum brotfleti þeirra. Fyrir hvert skjálftapar er jafnframt mæld fjarlægð
og tímamunur. Þessar stærðir skilgreina hraða. Myndin sýnir stefnudreifinguna fyrir alla eftirskjálfta í a) en
bara fyrir þá eftirskjálfta sem hafa hraða hærri en 100 m/s. Litmyndin í miðju sýnir einfalt líkan af skerspennu
umhverfis hringlaga misgengi með fast 10 MPa spennufall.
lar plot of the ML>0.5 IAs displays nearly oval form.
If only fast IAs are taken into account then the IA-
angular-distribution patterns change to the lobe-like
character: they show elongation with two lobes along
the slip direction, whereas the occurrence of IA in the
slip-perpendicular direction markedly diminishes.
To explain how a prior earthquake can bring sub-
sequent earthquakes to failure, we use the Coulomb
failure stress criterion ∆CFS=∆τslip − µ(∆σn −
∆P ), where ∆CFS is the Coulomb stress increment;
∆τslip is the change in shear stress due to the first
earthquake resolved in the slip direction of the second
earthquake; ∆σn is the change in normal stress due to
the first earthquake, resolved in the direction orthog-
onal to the second earthquake fault plane; ∆P is the
change in pore pressure; µ is the coefficient of fric-
tion. To examine the triggering effect we calculated
the time-space distribution of ∆CFS (both dynamic
and static) on the fault plane surrounding the rupture.
The earthquake was represented by an instantaneous
and uniform stress drop of 10 MPa on a circular area
with a radius of 100 m (earthquake with ML∼2.8) in
a homogeneous half-space. In such a case the nor-
mal stress change ∆σn is zero and ∆CFS reflects
only the slip-parallel component ∆τslip. To calcu-
late ∆CFS there was used the method of Bouchon
(1997) slightly modified and implemented after Bur-
jánek (personal communication). As depicted in Fig-
ure 11, the ∆τ field breaks up into two parts with
time: the permanent one (static ∆τ ) as a result of the
near-field deformation and the transient one (dynamic
∆τ ) carried by the transition of seismic waves. The
separation of the static and dynamic parts increases
with time. As can be seen in Figure 11, the static
∆CFS distribution pattern manifests a distinct elon-
gation in the slip direction, whereas the dynamic ∆τ
oscillates (∆τ takes on positive and negative values
in each point of the fault plane), which implies that
all points on the fault plane are shaken at respec-
tive times.Thus we infer that the static stress changes
82 JÖKULL No. 60