No abstract

No abstract

Yang et al. Reply: In the Comment, Corral points out that ''the only conclusion that can be drawn from this (i.e., our results presented in [1]) is that the seismicity time series is not uncorrelated, and there exists some dependence between magnitudes and recurrence times.'' We agree with Corral's view on how to explain our results of earthquake data analysis and the notion that a system of self-organized criticality (SOC) does not necessarily have an exponential distribution of recurrent times. In fact, the latter has been known to the geophysical community for quite a long time [2,3]. In [1] we should have stated that earthquake does not exhibit sandpile-type SOC.Corral proceeds to identify possible reasons why reshuffling the earthquake catalog will change the recurrence statistics. One of his findings is that ''when an earthquake starts, its magnitude is undetermined.'' This finding is not consistent with our analysis [1] nor with what has been well known for aftershocks: the higher the magnitude of an aftershock, the less frequently the aftershock will occur [4]. In other words, for aftershocks, the magnitude is not totally undetermined, but, rather, the magnitude is to some extent dependent on the magnitude of the main shock [4].Two separate data analyses are made in [1]. In the first one the seismicity time series is rearranged in the following way: the new time series consists of earthquakes occurring at the same instants as in the actual catalog, but the magnitudes of all quakes are drawn randomly from the Gutenberg-Richter distribution. The first-return-time probability distribution calculated from the rearranged catalog is clearly different from what is obtained from the actual catalog, strongly suggesting that the magnitude of an earthquake is not random but constrained by other quakes.The objective of [1] is to express our concerns that earthquakes being of SOC have two important implications: (i) earthquake can occur randomly anywhere at any PRL 95, 159802 (2005) P H Y S I C A L

[1] We determined the internal structure and mineral composition of the Yingxiu-Beichuan fault zone at the Zhaojiagou exposure and measured frictional and transport properties of the fault rocks collected to gain a better understanding of dynamic weakening mechanisms during seismic fault motion. This fault is a major fault in the Longmenshan fault system that caused the 2008 Wenchuan earthquake. The exposure studied is located midway along the surface rupture, close to where the largest coseismic displacement occurred. High-velocity friction experiments reveal exponential slip weakening from a peak friction toward a steady state value. Slip weakening is more pronounced for water-dampened gouge than dry gouge, suggesting thermal pressurization. The fault gouge has a very low permeability (< 10 À21 m 2 at 165 MPa effective pressure) and is surrounded by fault breccia with a permeability of 10 À19 to 10 À17 m 2 , grading into less permeable, fractured country rocks. The fault zone thus exhibits a "conduit/barrier" structure, allowing fluid flow only in the breccia zone. We numerically modeled coseismic slip weakening including thermal pressurization and mineral dehydration/decarbonation, basing our calculation on measured frictional and transport properties and on the slip history inferred for the Wenchuan earthquake. The results indicate that (1) thermochemical pressurization played an important role in causing dynamic slip weakening, (2) the slip-weakening distance is similar to the seismologically determined values, and (3) pore pressures might have exceeded the normal stress, thus maintaining temperatures below 600°C. Interestingly, enough heat was generated to fully remove and thermally pressurize the interlayer water from smectite, contributing an excess pore pressure of~6 MPa. In addition, we found that the incorporation of state-dependent fluid properties predicts much more efficient fluid pressurization than using constant properties. The dramatic weakening predicted probably offers a compelling explanation for the large coseismic displacement and slip acceleration observed near Beichuan city.

No abstract