Using GIS to characterize fault populations and seismicity of an active plate boundary

Term Project
CE 394K: GIS in Water Resources (and more!)

Tip Meckel, Ph.D. Candidate
Department of Geological Sciences / Institute for Geophysics                       
The University of Texas at Austin
Austin, TX  78712

Any comments can be sent to tip@mail.utexas.edu

If you are interested in a more developed treatment of the geologic interpretation of fault characteristics in structural settings, I have written a paper summarizing the significance of the fault lengths adjacent to the Macquarie Ridge that incorporates the GIS database that was developed for this term project.

Project Description and Outline of Goals

INTRODUCTION:    

   The Macquarie Ridge is a topographic feature on the sea floor south of New Zealand. It constitutes the plate boundary between the Pacific plate to the east and the Indo-Australian plate to the west.  Over the last thirty million years, this boundary has experienced deformation resulting from the differential movement of the two plates.  The recent seismicity of the region also indicates that the plate boundary is active and deformation continues to occur. 

    Past and present deformation has formed a vast network of faults and complex topography on the sea floor.  Images of the sea floor collected from ships during recent cruises have been used to interpret the location and extent of submarine faults related to the plate boundary.  Maps have been created showing fault distribution.  Digitized versions of these maps have been created and provide an opportunity to conduct spatial analysis of fault characteristics using GIS.

This goals of this GIS term project are to integrate mapped fault populations adjacent to the Macquarie Ridge with seismicity data to allow a geologic interpretation of the tectonic state of the plate boundary to be made.   Take a COOL flyover of the Macquarie Ridge courtesy of the Australian Geological Survey Organization.


 Shaded bathymetry over the Macquarie Plate Boundary

 

PROCEDURE:

    Sidescan sonar, swath bathymetry, seismic reflection, gravity, and magnetic data collected between latitudes 49º and 57º south during the 1994 cruise of the R/V  Rig Seismic (124) (170,000 km2) and 1996 cruise of the R/V Maurice Ewing (9513) (13,000 km2) in the region surrounding Macquarie Island have been used to locate the currently active plate boundary along the MRC and to study the structures and sedimentation that occurred during the evolution of this boundary (Massell, 1997; Schuur, 1997). A map of >16,000 interpreted faults that exist on the sea floor in the study area was compiled by Massell (1997). These faults were interpreted from high resolution sidescan sonar and bathymetry data.  These faults have been digitized by hand and converted to an ASCII file of the endpoints of the lines in latitude and longitude coordinates.  These data were processed to arrive at lengths for each of the faults.  The ASCII file was imported into GIS ArcView and the data were projected using the projection wizard.  A table was created that codified the length and position of each fault with a fault number.      
    Seismicity data were downloaded from the USGS for the are of interest.  These data come in table format with each epicenter identified by latitude and longitude coordinates.  Additional data in the table included magnitude, depth, time, and other standard desciptors of the event.  A table showing the attributes as they arrive from USGS is here.  These data were then projected onto the map of the faults using the projection wizard and knowing the projection of the fault data set.  The results can be seen below.

 

This is an example of the output from the GIS database incorporating both the faults and the seismic data.  Although it is not seen here, latitude and longitude coordinates have been imported and exist in current layouts of this figure.

This is a map of over 16,000 faults on the sea floor.  The dots are earthquake 
epicenters for events that occurred within the last 30 years.  The epicenters are 
colored according to their depth.

 

  

   ANALYSIS:  

One interesting result of the length data for the faults follows an exponential distribution, as seen to the right.  The equation for the line is not as important as the shape.  It is likely that this shape indicates fundamental processes of faulting at spreading ridges. 

 

A really useful aspect of using GIS was that I could query the table of fault attribute data to quickly generate plots of the distribution of faults of different length.  To the right, the yellow faults are the smallest faults which are most described by the exponential distribution.  The black faults are a subset of the entire population that are more accurately described by a power-law scaling.  The interesting observation is that there are few black faults right in the narrow zone of the active plate boundary.  For a detailed geologic discussion of this fascinating topic, see an accompanying paper I wrote.

  

The earthquake data was also analyzed using ArcView.  To the right, are two plots of the density of different attributes of the earthquakes.  The upper plot is a density plot of the depth of earthquakes.  The lower plot is a density plot of the year and date that the earthquake occurred.  It is clear from both of these plots that not only can powerful analysis be done with data within ArcView, but that earthquake characteristics are not random in space or time.  While to plots do have a bias where there is no data, it is significant that there is no data there.  Unlike other sampling efforts, we generally assume that ALL earthquakes over a certain size are observed.  That means that the distribution in the adjacent plots are significant for understanding the current and past (VERY recent past...30 years) stress state at the plate boundary. 

 

CONCLUSIONS:

A GIS database of sea floor faults has proved beneficial for this research in numerous ways.  Initially, the length of the faults was of interest.  There is some debate as to whether faults follow specific scaling distributions with regard to attributes such as length. The lengths of faults follow an exponential distribution for the study area.  This is significant because this has not been documented and may be suggestive of fundamental processes at plate boundaries.  After it was identified that the fault lengths followed an exponential distribution, it was especially useful to be able to query the fault table and observe the spatial distribution of the faults.  In an accompanying paper to this database, this has proved useful for analyzing the processes of fault formation at spreading ridges. 

    In addition, seismic events (earthquakes) could be integrated into the data base.  This allowed some conclusions to be drawn about the distribution of earthquakes along the plate boundary.  It is evident from density plots shown above that earthquakes are occurring in clusters both in time and space.   These concentrations of activity are suggestive of the current stress state at the plate boundary.

 
 

 

REFERENCES:

Massell, C. G., 1997.  The neotectonics of the Macquarie Ridge Complex, Pacific-Australia plate boundary, B.S. thesis, The University of Texas at Austin, 101 pp.

Massell, C. G., Coffin, M. F., Mann, P., Mosher, S., Frolich, C., Schurr, C. L., Karner, G. D. R., and Lebrun, J.-F., in review.  Neotectonics of the Macquarie Ridge Complex, Australia-Pacific plate boundary, Journal of Geophysical Research.

Schuur, C. L., 1997.  Sedimentary regimes at the Macquarie Ridge Complex south of New Zealand: interaction of Southern Ocean circulation and tectonism.  M.A.
            thesis, The University of Texas at Austin, 104 pp.

Back to the GIS Class Page