ABSTRACT

The Denali Fault in Central Alaska is one of the longest in the world and it has been critical to the formation of North America since the time of dinosaurs. The fault is an important source of hazard - it caused a magnitude 7.9 earthquake in 2002 - but only a handful of geophysical studies have examined its structure. We know very little about how the fault is shaped and behaves within the Earth's crust, and we do not know how the rocks change along the length of the fault. We use new 3D imaging techniques to see the structure of the fault 2-50 kilometers below the surface, which help us to determine the geologic history of Alaska. To add further detail to our imaging, we are placing hundreds of coffee-can sized seismometers in two separate locations to measure the shaking from earthquakes both locally and from across the planet. This allows much more accurate estimates of earthquake hazards for the region's residents and millions of annual visitors. On a larger scale, it shows us how faults work at the edges of continents, and whether this fault forms the true edge of the North American geologic plate. Our use of inexpensive, portable instruments and our techniques to decode specific earthquake wave types are likely to change the way people study fault zones around the world in the future.


Technical Summary: We are imaging the Denali fault at regional scale using joint tomographic inversion of body and surface waves, and to locally image the damage zone of the 2002 magnitude 7.9 Denali earthquake. This work is motivated by several unanswered questions specifically about the Denali fault as well as general fault properties. Does the fault have a signature in the Moho or upper mantle? Are there persistent contrasts in across-fault velocity in the crust? What is the relative tectonic importance of the different fault strands? Are there unique damage zone patterns that result from supershear ruptures? Where are the preferred nucleation sites for large earthquakes, and what are the patterns of shaking that can be expected?

To date, only a handful of geophysical studies have examined the Denali fault in isolated sections as part of regional-scale linear transects across central Alaska. Multiple tomographic studies have successfully imaged the structure of the subduction zone and underlying Yakutat plate beneath the region, but they have not achieved sufficient horizontal resolution to image the crustal and upper mantle structure of the Denali fault. There is little constraint on the fault zone structure at depth and the along-strike variation of deformation patterns. To resolve the regional-scale structure of the Denali fault, we are tomographically imaging the fault from the surface to 70km depth using newly-developed joint inversion techniques combining ambient noise Rayleigh wave phase velocities and body wave double-difference measurements which achieve kilometer-scale resolution in both Vp and Vs. In the process, we are also building improved catalogs of arrival times for P, S, and fault zone head waves.

At the local scale, no seismological study has been conducted to examine the fault damage zone and principal slip surface structure. At other major strike-slip faults, linear or two-dimensional arrays of seismometers with spacing less than 100m have successfully detected fault zone head waves that propagate along (semi-)vertical interfaces with contrasting velocities and fault zone trapped waves confined inside fault-related low-velocity zones consisting of damaged rocks. These two seismic phases, unique to fault zones, can be exploited to provide constraint on fault structure relevant to the mechanics of earthquake rupture. Leveraging the recent development of low-cost highly-portable three-component seismometers, we are deploying hundreds of instruments in dense fault-straddling arrays for one month each at two locations along the Denali fault. The density and richness of the first dataset, which we've already recorded, are unique not just to the Denali fault zone, but to all major strike-slip faults worldwide. Our preliminary analysis, including trapped wave detection and high-frequency ambient noise cross-correlation, contains critical information about the damage zone structure in a section of the fault that recently slipped in a supershear rupture. These data demonstrate the feasibility of further analysis including ambient noise tomography, trapped wave normal modes, and fault zone head wave detection.

Please report errors in award information by writing to: awardsearch@nsf.gov.