Date: Thu, 26 Oct 1995 15:03:54 +0900
To: eq-geo-net@gsjtmws8.gsj.go.jp
From: arrows@asuvm.inre.asu.edu (Ramon Arrowsmith)
Subject: Future of Task Group II-3, Post Erice Meeting
Discussion of possible future directions of International Lithosphere Program
(ILP) efforts in paleoseismology
INTRODUCTION
In a recent discussion at the Erice conference (Active Faulting Studies for
Seismic Hazard Assessment), Bob Yeats requested comments and discussion of
future efforts to be taken within the ILP program on paleo-seismology. In
particular, the need may arise for a follow-on proposal to ILP as the existing
project is being successfully concluded. The effort to organize two
conferences (Marshall and Erice) and the associated opportunities to share
ideas and new results and to establish collaborative links have been extremely
valuable to the community. The production of the JGR Special Volume will bring
together much new information and provide a comprehensive view of the state of
understanding of the paleoseismology community. International training courses
for paleoseismologists worldwide may prove to be the most significant legacy of
the current efforts.
Through this contribution we mean to stimulate the discussion of new
directions of ILP efforts in paleoseismology by pointing out some of the
outstanding questions and identifying possible ways to address these problems.
We also provide a partial summary of previous comments. Thus we hope to give
people the opportunity to disagree with our inferences and to suggest
additional or alternative important focus areas for the international
paleoseismology community.
First, we comment on the importance of gathering more data so that we may
better understand the processes controlling the timing and distribution of
earthquakes. We are entering a time of more comprehensive data gathering.
Consider Robert Muir-Wood's picture of how little of the paleoseismology rock
we have quarried. We should work to carefully catalogue those data and include
quality/uncertainty measures (another important lesson from Erice). Secondly,
we should continue to apply and refine the existing methods typically used in
paleoseismology. The data quality and density of some places (such as the
integrated hazard model of SCEC, or Kerry Sieh's efforts to increase dating
resolution and quantity in Sumatra) are commendable, but also represent special
cases, and in many other places, standard paleoseismological techniques are
just being or have yet to be applied. Finally, we should work to develop new
methods to broaden our paleoseismological horizons. We need to focus efforts on
new methods and approaches, as even in those areas where we applied
state-of-the-art methods (such as the San Andreas Fault system) our current
knowledge is still surprisingly incomplete.
GATHERING MORE DATA
Focus on a few places and a few problems?
Kerry Sieh pointed out his frustration with the paleoseismologic record in
most areas when trying to use this record to establish fundamental
relationships and fault behavior. Therefore we may want to advocate
concentrating on focused studies where the most data and the best methods are
combined to solve some of these first order problems. As we are beginning to
utilize a fully interdisciplinary approach along the San Andreas fault system
for example (encouraged by the occurrence of earthquakes where we didn't expect
them) we realize problems in our early estimates of earthquake hazard in
California and NW Mexico. Other regions, however, may be better candidate sites
for such fundamental studies. This comment is not meant to distract from the
need to conduct paleoseismologic work in all earthquake-prone areas near our
communities.
Comprehensive paleoseismology
In order to learn the most about the activity of a particular structure, and
to contribute most to the paleoseismological database, a comprehensive effort
must be undertaken that includes (where possible) mapping and documentation of
the offset of slip in previous events along the structure, archeo-, historical,
and instrumental seismicity, geodesy, geomorphology, mesoscale structure,
regional seismotectonics (migration of earthquakes, contribution to plate
deformation), and the traditional paleoseismology of some trench sites.
APPLICATION AND REFINEMENT OF CURRENT METHODS
Repeatability and established criteria in paleoseismological methods
Jim McCalpin identified the importance of establishing "well-documented,
accepted criteria for relating deposits and faulting events." Of course, we do
not want to develop a cookbook approach that will lead paleoseismologists to
blindly interpret trench logs, but we should continue to evaluate the methods
of paleoseismology and ensure that much of the raw data from projects is made
available for the consideration and further analysis by fellow
paleoseismologists. Not only will that provide an opportunity for discussion
of specific results, it will also provide an opportunity to learn from each
other the "tricks of the trade."
What if we did a paleoseismology experiment in which we took a suitable site
or series of sites and had several different groups go through complete
paleoseismological investigations, without mutual consultation (i.e., blind to
each others efforts)? Such an effort might permit us to comment regarding
repeatability and the effectiveness of different techniques/styles.
Observation of original forms and initial modifications of historic surface
ruptures
Clearly, interpretation of the paleoseismic record requires observation of
historic surface ruptures in order to understand their original forms (cross
section and plan view) and the modifications of those forms by geomorphic
processes. These observations provide the opportunity to understand both the
change in the landform, as well as the stratigraphy developed adjacent to fault
scarps. This was identified in the 9/4/95 comment of Jim McCalpin to this
group.
Rate constraints from tectonic geomorphology
Along with developing criteria for the identification of slip events from the
sedimentological record, we should begin to constrain some geomorphic criteria
(i.e., the "Machette constraint" of Hanks, et al, 1984, which indicated that a
scarp greater that 100 ka in the Rio Grande rift portion of the Basin and Range
province, USA will not be preserved in the landscape, or landform assemblages
along strike-slip faults indicating the order of magnitude of the slip rate (1
mm/yr vs. 30 mm/yr)).
Integrated investigations of blind thrust faults
As several recent earthquakes in California have demonstrated (Coalinga,
Whittier Narrows, Loma Prieta, and Northridge), the threat of buried thrust
faults presents a significant hazard. Many have suggested further work on
these structures; we advocate studies integrating geology, geomorphology,
seismology, and geodesy. Not only does this provide the opportunity to
constrain the geometry and rates of deformation associated with these
structures, but also it permits us to learn more about the development of
permanent deformation in areas of active tectonics, and the relation between
earthquakes, geomorphology, and geologic structure. This approach was
advocated and some of the pitfalls identified by Gianluca Valensise in his
Erice presentation.
Mechanical models of faulting
Much effort is oriented toward the elucidation of the timing and magnitude of
slip in paleoearthquakes. The interpretation of the observations of slip with
regard to the process of faulting is not well developed however. Mechanical
models of faulting that permit interaction, variable material properties, and
fault strength may be used to aid in these interpretations. In other words,
how can we relate the slip rate, the slip per event, and surface offset? How
should we compare the surface offset, the seismological slip distribution, and
the geodetic slip distributions (if we can determine them)? These models will
also be essential in the interpretation of point observations of paleo-fault
offsets with regards to likely rupture scenarios and segment lengths.
Quantitative models of qeomorphic and tectonic processes
Along with the mechanical modeling identified above, quantitative simulation
of geomorphic processes and their interaction with tectonic processes will aid
both in the development of our understanding, as well as in the quantitative
interpretation of the landscape and the activity and location of active
structures.
DEVELOPING NEW METHODS
Some of the methods presented above have yet to be comprehensively applied,
and so may be considered new. What are the promising new directions in
paleoseismology? New directions must provide increased data density, accuracy,
and precision. The methods that Kerry Sieh has developed for using corals and
that Shmuel Marco and Amotz Agnon have found from the episodic deformation of
soft sediments to provide high resolution data over long time spans may
represent the types of methods and ingenuity needed here.
OUTSTANDING QUESTIONS
What are the outstanding questions for paleoseismology? Some have been
identified above. Here are three from Tony Crone: 1) How can we decide the
likely extent of future ruptures considering the complex interaction between
pre-existing faults and currently active faults? 2) If we note discrepancies
between geodetic, seismologic, and geologic data, which do we accept in the
consideration of regional seismic hazard? 3) What is the role and contribution
of less prominent second order faults to seismic hazard?
PROPOSED DIRECTIONS
Regarding a possible proposal to the International Lithosphere Project, it
should strike a balance between the three directions presented in this comment.
We should increase the spatial and temporal depth, breadth, and precision of
our paleoseismological database. That may be achieved by application and
refinement of current methods, and the development of new ones.
In a proposal to the ILP, we should continue the following:
1) Yearly workshops of the Marshall and Erice sort.
2) Field trips of the Mongolia sort.
3) International training courses.
We should develop the following:
1) Establish a database of paleoseismological data and methods. This could be
made accessible to the community via the World Wide Web.
2) Documenting and testing or at least discussing the criteria for identifying
events in the paleoseismic record.
3) Implement integrated and comprehensive paleoseismology.
4) Integrate quantitative models of deformation and surface processes into the
interpretation of deformation rates and geometries.
5) Encourage innovative techniques.
Respectfully submitted,