Home > Academics > Institutes & Centers > CREST > Research > Geophysical Modeling > Activities

Activities

a) Major Research Activities

Shear Wave Splitting Measurements In The New Madrid Seismic Zone With Local Earthquake Data

This project investigates crustal anisotropy in the New Madrid seismic zone (NMSZ) by analyzing shear wave splitting from local earthquake data. Located in the central United States, the NMSZ spans portions of western Tennessee, northeastern Arkansas, and southeastern Missouri. Preliminary results derived through analysis of 540 suitable events from 1151 earthquakes (M2) spanning the years 2003-2011 consist of 64 high quality shear wave splitting measurements. Two fast shear-wave polarization directions have been identified within the Reelfoot and portions of the Western Rift Margin fault region: N55-65 and N110-115, of which the first is the most dominant. The observed time delays between fast and slow split shear waves vary between 22 to 240 ms, or 1 to 25 ms/km when normalized to hypocentral distance. Most time delays larger than 14 ms/km are associated with shallower events at 5 to10 km hypocentral depth and are located within the southern region of the Reelfoot fault.  The same region is also associated with low shear wave velocity anomalies (Powell, 2010) and microswarm activity (Bisrat et al., 2012); split shear wave time delays give an estimate of about 6% average crustal anisotropy and support the proposed existence of high pore-fluid pressure in the shallow crust.

Ambient noise cross-correlation in the eastern Tennessee seismic zone.

We applied this technique to the continuous data recorded at the eastern Tennessee short-period seismic array. We first collected the continuous data during a period of 3 months, usually considered as a minimum recording time to obtain sufficiently high signal-to-noise ratio for imaging purpose. We computed the cross-correlation on the vertical component for all simultaneously recording station pairs. The obtained cross-correlograms were then analyzed by means of a standard multiple narrow bandpass filter technique typically used to extract group wave speed from surface waves in seismic signals. We thus obtained an average Rayleigh wave group dispersion curve for the region at periods ranging from 0.5 to 30 seconds, i.e. at higher frequencies than what earthquake studies allow because of the attenuation properties of the Earth’s continental crust.

P-wave teleseismic travel time residual mapping in the eastern Tennessee seismic zone.

In this work we applied the adaptive stacking technique to the waveform data recorded at CERI short-period array from 1999 to 2011. We thus selected 224 events with epicentral distances ranging from 30 to 100 degrees (P-waves are not observed between 105 and 142 degrees because of their refraction in the Earth’s outer core) and with magnitudes greater than 5.5 (Fig. 4). Locations from the National Earthquake Information Center (NEIC) and reference velocity model AK135 were used to compute theoretical arrival times. A total of 2855 P-wave residuals were obtained for the entire observation period. Fig. 5 shows the efficiency of the adaptive stacking procedure to align seismograms and extract the travel time residuals. On Fig. 6, individual residual distributions are shown for each station, along with its mean value and associated standard deviation. These results are compiled on the map on Fig. 7. Results obtained for individual events will be used to determine the 3D velocity structure of the deepest part of the crust and upper mantle.

Kraljevo earthquake sequence analysis

On November 3., 2010, a magnitude 5.4 (Mw) earthquake occurred near the city of Kraljevo, Serbia, in the Balkan Peninsula, and was followed by more than 650 aftershocks. In order to better characterize its epicentral location, its depth and the orientation of the fault on which that earthquake took place, we collected waveform data for the main shock and its aftershock sequence from several seismological institutions of Serbia  and neighboring countries. We first of all obtained a 1D velocity model for the area, which can be used in the future to locate other earthquakes that will occur in the vicinity of Kraljevo. This velocity model could also be used as an initial model for a 3D local earthquake tomography study in the region. The accurate locations that we obtained show that the mainshock was located at 43.743N and 20.708E, at a depth of 15.4 km. The aftershocks clearly depict a subvertical EW-trending plane which allow to identify the direction of the fault involved in Kraljevo earthquake. Its strike-slip nature is furthermore accredited by the focal mechanisms determined for the aftershocks of magnitude greater than 3.0, although some of them seem to be associated with different mechanisms and may have occurred on secondary, differently oriented faults. This is also suggested by a NS alignment of epicenters in the southwestern part of the region affected by the aftershocks.

Joint Bayesian inversion of hypocentral and velocity model determination with the reversible jump algorithm

Markov chain Monte Carlo (McMC) inversion algorithms offer an interesting alternative to linearized inverse methods as they allow to sample the parameter space according to the posterior probability distribution by combining a priori information with the information contained in the data without any need for regularization. Such algorithms also have the interesting property of showing no dependence on the initial model configuration and are furthermore well known to provide more realistic estimates of parameter variance and covariance. A feature of our algorithm is that, besides P-wave velocity values in each layer of the velocity model, the number of layers itself, as well as their respective thicknesses, are also considered as unknowns and inverted for in the procedure. This is achieved by means of transdimensional partition modeling and yields a parsimonious solution whose complexity (the number of layers in the model) is driven by the data and their associated uncertainties. The velocity inversion part of the algorithm is still in progress but the earthquake location part has already been successfully tested.  One experiment we carried out consisted in generating synthetic P-wave arrival times at the 97 seismic stations of the Northern California Seismic Network (NCSN) for 18 hypothetical earthquakes having occurred in the middle of the array.  The total resulting dataset consists of 1747 arrival times to which Gaussian noise was added to simulate real data noise. The obtained results show that in each case, the true location is located within 1 standard deviation around the mean hypocentral solution. A comparison with the results obtained with a commonly used earthquake location software, namely Hypocenter, based on a linearized inversion method shows that in the latter case, the depth of some earthquakes can be erroneous if the chosen starting location was not appropriate (Fig. 16). This illustrates the fact that such techniques may strongly depend on initial models and get easily trapped in local minima of the objective function. This simple experiment clearly demonstrates the superiority of the method we developed with respect to conventional techniques.

b) Training and Development Activities for Students

All students, post-doc and faculty participating on Project 5 had multifaceted training and development opportunities. During the summer of 2011 three graduate students, Sanja Knezevic-Antonijevic, Jubril Davies and Tolulope Agbaje, completed two-month summer internships at the University of Nevada, Reno. Their summer field projects included paleo-seismic studies using sonar in Lake Tahoe and study of 2011 Hawthrone earthquake sequence. Students presented results of summer internship research at the December 2011 American Geophysical Union Meeting in San Francisco, CA (funded by CREST Project 5).

All graduate students had a chance to attend at least one national research conference. Sanja Knezevic-Antonijevic presented results of her research at the North Carolina ArcUsers Group Fall Conference in  Carolina Beach, NC (September 14-16, 2011) and at the Fall AGU meeting in December of 2011 in San Francisco, California.  In the spring of 2012 five NCCU graduate students, Sanja Knezevic-Antonijevic, Jubril Davies, Tolulope Agbaje, Carl Stearns and Ayodeyi Kuponiyi presented their CREST sponsored research at the Seismological Society of America Meeting in San Diego, CA. In addition, Tolulope Agbaje won first place for her poster at the 4th Annual CST North Carolina Central University Student Research Symposium (March, 2012).

In addition to courses offered at NCCU, several students completed Principles of Seismology and Data Analysis in Earth Science graduate courses at UNC-CH and participated in the weekly student-faculty Journal Club. One graduate student (Philip Martin) was also able to complete summer 2012 volcano seismology field course through New Mexico Tech and the Instituto Geofisico in Ecuador. During the course he participated in deployment of  six seismometers and an infrasound array on the flank of the active volcano Tungurahua. During spring semester of 2012 Philip also participated in the Central North Carolina Seismic Survey lead by Dr. Lara Wagener (UNC-CH) during witch 12 seismic monitoring stations were installed in Lee County, mostly in the Triassic basin. The goal of the study is to document background seismic activity for possible comparison with future levels of seismicity, if extraction of natural gas using the hydrologic fracturing method in approved in North Carolina.  He will share both of his experiences with other graduate students and faculty during brown bag seminar in the Fall of 2012. 

With the help of other external funding (“Technology Assistance with Implementation and Operation of Transportable Array Element of EARTSCOPE and USArray”, May 1, 2012 – September 30, 2012, G. Vlahovic, PI) this summer we were able to support two additional undergraduate students interested in seismology. After completing the mandatory training at the workshop held at the University of Pittsburg, PA (May 21-24) James Howard and Thomas Horne (undergraduate students) and Philip Martin (graduate) will use their geospatial analysis skills to select possible sites for the USArray stations, verify suitability of each site during the field visit and write and submit reconnaissance reports to the funding agency, Incorporated Research Institutions for Seismology.  In total, 25 sited will be identified and approved in North Carolina and Virginia before the end of the project.

c) Outreach Activities

On June 18, 2012, as part of the introduction to the University, Dr. Pierre Arroucau presented overview of the CREST seismology research program to high school students involved with Summer Ventures. 

During the summer of 2012 two graduate and one undergraduate student participating in the USArray siting project were trained to approach landowners and explain EarthScope project, USArray station requirements and seismology imaging techniques in understandable terms.  They were coached to answer questions about the goals of the project, specific USArray site requirements and landowners commitment and liability issues, general questions about earthquake activity in the eastern US, as well as contemporary concerns of the citizens.  For example, questions that frequently came up were about funding (if taxpayers’ money were being used to fund the project) and about fracking (if project had any connection with gas/oil exploration activities).  Students spent 10 weeks this summer visiting 25 potential sites and talking to more then a hundred North Carolina and Virginia landowners increasing their awareness about the EarthScope and seismology

d) Collaboration with other HBCUs

Over the past two years G. Vlahovic from NCCU and R. Malhotra from Fayetteville State University (FSU) have partnered on various initiatives, prominent among them the summer internship experience for FSU Geography undergraduates that was funded and promoted through NCCU. In Summer 2011 one student interned at Fort Bragg and two students interned at the City of Fayetteville. This summer (2012), one student interned at the Environmental Protection Agency and another one at the City of Fayetteville. Moreover, two undergraduate students from FSU applied and were accepted to the Earth Science master’s program at NCCU. Recently G. Vlahovic submitted a White Paper to National Geospatial Intelligence Agency titled "Talent Pipeline – Increasing Diversity and Enhancing Geospatial Programs". It outlines multipronged approach to development of lasting geospatial intelligence focused partnership and pipeline for minority talent between two universities.  The White Paper proposes that graduate programs (Computer Information Science, Earth Science, and Physics) at NCCU form a comprehensive geospatial intelligence focused partnership with undergraduate programs (Computer Science, Geography, Intelligence Studies, and Physics) at FSU to create a pipeline for developing future Intelligence Analysts. The partnership would provide undergraduate students pursuing geospatial intelligence studies at FSU early exposure to graduate school, a chance to participate in a summer internship/work programs at external partners such as Oak Ridge National Laboratory (ORNL), L-3 Communications, and Booze Allen Hamilton (BAH), and a study field trip to India. The proposed program, if funded, would over 4 years (2013–2017) expose FSU sophomores, juniors and seniors to graduate studies opportunities at NCCU and partnerships that lead to IC careers. It is anticipated that by the end of the grant 15 undergraduate students from FSU will join the graduate programs at NCCU every year.