I have been studying fault rocks and fault mechanics for 40 years, trying to understand
mechanisms of earthquakes. A basic strategy has been to study fault rocks for understanding
deformation and transport processes in fault zones at depths, to reproduce the same processes
in laboratory experiments and determine mechanical and transport properties of faults, and to
conduct earthquake modeling based on measured properties and compare with natural
earthquakes. I will try to give an overview of the progress of fault studies in the last 25 years,
emphasizing the importance of such integrated studies. The following four topics will be
covered from my own perspectives of fault and earthquake studies. High-velocity
frictional properties of faults in relation to earthquake rupture dynamics will be the
main focus, but the lecture will cover lithosphere rheology, initiation processes of
earthquake-induced landslides, and a basin evolution and pore-pressure development as
relevant topics.
[1] Friction to flow law
A simple friction to flow law merges strength profiles of lithosphere and velocity-dependency
models of faults that have been used widely in the last three decades to characterize the
thickness and internal structures of the lithosphere and to model earthquake cycles and
earthquake rupture propagations, respectively. The law allows analyzing earthquake
generations including frictional, transitional and flow properties at shallow to deep faults
across a lithosphere. Analyses shows how strength profile evolve during earthquake cycles.
The law can be extended to describe brittle to high-temperature flow properties across a
lithosphere, and realistic analysis of earthquake generation and interseismic deformation,
including postseismic deformation, will be possible.
[2] High-velocity weakening of fault and a source of diverse seismic activities
Extensive studies in the last two decades demonstrated that faults undergo dramatic
weakening at seismic slip rates, through mechanisms such as flash heating/bulk heating of
gouge, frictional melting, and thermochemical pressurization. It is likely that earthquake
nucleation is controlled by rate and state frictional properties at slow slip rates, whereas
high-velocity weakening can affect the growth processes of large earthquakes. Combinations
of those slow and high-velocity properties can produce very diverse seismic and aseismic
fault motions. I will also discuss technical problems in building friction apparatuses to extend
the high-velocity friction studies to wet environments and to higher normal stresses and
temperatures.
[3] Catastrophic landslides triggered by earthquakes
Tsaoling landslide in Taiwan is the largest and well-documented landslide among many
landslides triggered by the 1999 Chi-Chi earthquake. The landslide occurred along a very flat
bedding surface, and a Newmark analysis using high-velocity frictional properties
quantitatively reproduced the initiation and runaway processes of the landslide, elucidating
the importance of slip weakening of landslide surface. The method has many applications to
earthquake-triggered landslides.
[4] Basin evolution and pore pressure development
Pore-pressure distribution in the earth is poorly known at present. Analysis of basin
evolution, using measured permeability and storage capacity of all formations of a basin,
revealed the sedimentation and fluid flow processes in the last 30 Ma that lead to the
development of abnormal pore pressure below about 4 km in depths. Modeling based on
measured transport properties will be useful to solve many problems such as fluid
flow in the earth, effect of water on earthquake generation, waste isolation, and
CCS. |