S43A-1873
Recent Deformation in the Mississippi Embayment from High-Resolution Reflection Data
The New Madrid seismic zone (NMSZ), located in the Mississippi embayment (Central United States), is known for the highest rate of seismic activity east of the Rocky Mountains as well as the destructive sequence of earthquakes (> 7.5M) that occurred in New Madrid, Missouri during a three-month period in 1811-1812. Here we present preliminary results of a 300 km-long high-resolution marine seismic reflection survey conducted in June 2008 along the Mississippi river from Caruthersville, Missouri, to Helena, Arkansas. Paleoseismological observations in this area, just south of the NMSZ, indicate that seismic activity has occurred here that is not correlated with the present NMSZ active fault system. Our acquisition program was designed to image the sedimentary sequences from the southern part of the NMSZ through this area with the goal to identify and characterize concealed faults inferred from the paleoseismic evidence. The reflection data were acquired using a dual chamber 245/245 cm3 (15/15 in3) mini-GI airgun firing at 13.8 MPa (2000 psi) every 10-12 seconds and a 24 channel 75 m-long active streamer. The 4 m minimum offset, 3.125 m group interval and 1.6 m CMP spacing enabled us to successfully image and identify recent deformation in the sedimentary layers (< 1 km depth) with unprecedented resolution. An ecosounder (or CHIRP) was used in conjunction with the aforementioned seismic array to map near surface structure (< 15 m). Observed structures from the CHIRP data suggest small-scale faulting and geomorphologic forms. Reflections up to an approximate depth of 1 km allow us to map the Paleozoic, Cretaceous, Tertiary and Quaternary sequences throughout much of the profile, revealing that the deformation is accommodated within discrete zones where the unconsolidated sediments are folded and faulted. In particular, the data show pronounced folding of the sedimentary layers from the Paleozoic to the Eocene/Quaternary unconformity. Some of the most significant deformation is imaged near the presently seismically active southern arm of the NMSZ. Furthermore, Quaternary deformation is also visible approximately 30 km south of the Blytheville arch in an area where no seismicity is observed today.
S43A-1874
Reflection and Transmission Imaging of the Upper Crust Using Local Earthquake Seismograms
We utilize the characteristic features of primary P- and SH-wave coda, and Sp waveforms from local microearthquake data and perform pre-stack migration in an attempt to image prominent reflectors in the upper crust of the New Madrid Seismic Zone (NMSZ. Our case study uses data from a broadband station, PARM, of the Cooperative New Madrid seismic Network (CNMSN). Near-by exploration well log data of the Wilson 2-14 and Dow Chemical/Wilson #1 wells are used to constrain the upper 4 km of the velocity model in the region. A constant velocity of Vp 2.50 km/sec and Vs 0.70 km/sec for the unconsolidated sediments and Vp 6.0 km/sec and Vs 3.2 km/sec for the Paleozoic sedimentary rocks can represent the structure beneath PARM. Despite polarity differences among P-, SH-, and Sp waveforms, seismic images show consistent reflectors common among the three wave types. There are excellent correlations associated with the base of the upper Cretaceous/Holocene Mississippi Embayment Supergroup and the base of Knox group, which were also reported in a study of near-by seismic-reflection profiles by Hamilton and Zoback (1982). The Bonneterre Formation seems to be a prominent seismic stratigraphic marker associated with an interface displayed in the profiling. We find that earthquake event S-P times must be ~3 sec or more to resolve reflectors at about 4 km depth. The basement occurs at about 4 km and appears on the reflected P- and SH-wave images.
S43A-1875
The Effects of Mississippi Embayment Sediments on Local Earthquake Tomography Models for the New Madrid Seismic Zone
The effects of velocity and thickness variations in thick, unconsolidated Mississippi Embayment sediments on local earthquake tomography for the New Madrid seismic zone are investigated using normal station corrections, synthetic modeling, and application of a sediment correction based upon independent knowledge of sediment thickness and velocity structure. Synthetic modeling indicates that smearing from strong velocity perturbations in the unconsolidated sediments can influence the uppermost basement portion of the P wave velocity solution and can extend into deeper portions of the S wave velocity solution. However, there is little evidence for smearing due to velocity perturbations in the unconsolidated sediments in inversion results obtained using real arrival time data. The clustered distribution of hypocenters makes station corrections sensitive to velocity variations located deeper than the unconsolidated sediments and application of station corrections removes useful information about the basement from the inversion solution. The sediment correction also appears to remove useful information from the inversion solution but is an improvement for the P wave solution over removal of station corrections. There is no evidence that the sediment correction is removing smearing effects produced by velocity perturbations in the unconsolidated sediments. The unconsolidated sediments do not appear to be exerting a first order effect on local earthquake tomography solutions.
S43A-1876
P and S Wave Velocity Structure and Vp/Vs Ratios for the New Madrid Seismic Zone
Three dimensional P and S wave velocity models have been constructed for the New Madrid Seismic Zone (NMSZ) using double difference local earthquake tomography (tomoDD). TomoDD incorporates catalog arrival times with catalog and waveform cross correlation differential times to solve for P and S wave velocity and for high resolution earthquake locations. For the NMSZ, we utilized 101504 P wave differential times and 67811 S wave differential times from 1157 earthquakes recorded over the time period 2000 to 2007 by the Cooperative NMSZ Network. The NMSZ consists of three intersecting arms of seismicity located in the central United States. There are approximately 200 earthquakes a year in the NMSZ despite the absence of a major plate boundary. Most earthquakes occur along the central Reelfoot Fault leading to uneven source distribution. We use a finite difference travel time calculator combined with an irregular inversion grid of nodes spaced every 5 to 20 kilometers horizontally and 1 to 3 kilometers vertically. Model resolution was examined using chessboard and spike tests and indicated that resolution is highest close to the source region between depths of 5 to 12 kilometers. P and S wave models indicate that velocities close to the source region are slightly low relative to the 1D starting model. The decrease in velocities may be indicative of rock properties, such as increased fluid content and fracturing. A high P and S wave velocity anomaly located away from known faults is associated with a known mafic intrusion to the northwest of seismicity.
S43A-1877
Pn Velocity Distribution Beneath the New Madrid Seismic Zone, Central USA, Using Seismic Arrays
Regional earthquake data recorded by the Cooperative New Madrid Seismic Zone (NMSZ) network were analyzed to investigate spatial variation of Pn velocity beneath the NMSZ. Shallow regional earthquakes at epicenter distances from 220 km to 1100 km from different directions were selected to ensure a successful identification of Pn arrivals and a proper azimuthal coverage of seismic rays. Regional seismic network stations are sorted spatially into many overlapped small aperture seismic arrays. Apparent azimuth and apparent velocity of the incoming Pn waves for each array can thus be determined from array analysis of the Pn arrivals. The apparent Pn velocity determined for an array depicts the upper mantle Pn velocity for a region outside of the array along the incoming ray direction immediately beneath the earliest arrival station. Using regional events from the north, east, and west, the resultant Pn velocities beneath the NMSZ vary in the ranges from 7.91 ± 0.03km/sec to 8.22 ± 0.03km/sec. These observations are slightly different from the previous studies in the area using long seismic refraction lines by Mooney (1983) and Catchings (1999), which reported a constant Pn velocity of 8.1 km/sec and 8.25 km/sec, respectively, for the same region. In addition, Pn travel time residuals during least square fitting to determine apparent velocity reveal systematic spatial patterns that support further the observed lateral variations of Pn velocities in the region. Using the events from the south, the resultant apparent velocities of the first arrivals are ranging from 6.0 ± 0.05 km/sec to 7.0 ± 0.05 km/sec suggesting that the first arrivals are traveling first through a small section of oceanic crust before entering to mid to lower continental crust. The observed lateral Pn velocity variations will provide important constraints for receiver function structural study and other studies in the area.
S43A-1878
Upper Mantle Structure Beneath the Southeastern United States from Receiver Functions: Evidence for Small-scale Mantle Convection?
The Bermuda swell and hotspot have been proposed to be the result of small-scale convection driven by the step from the thick craton boundary to thinner passive margin beneath southeastern North America. If that assumption is correct, the cold downwelling limb of the convection cell should have higher seismic velocity than average mantle and the positive Clapeyron slope of the olivine to wadsleyite phase transformation implies a shallower 410-km discontinuity in colder than average mantle. We compute receiver functions using the permanent ANSS broadband seismic stations in the southeastern United States, in an attempt to resolve signals from the 410-km discontinuity. Five of these stations show coherent signals in the interval between 40 and 50 seconds, at approximately the time predicted by the IASPEI91 model. Three of these stations are in the Appalachian highlands (MCWV, BLA, LRAL) and two are in the Atlantic Coastal Plain (CBN, NHSC). The data are from magnitude M 6.2 and larger shocks occurring since 2001, in the distance range 30 to 102 degrees. Earthquakes to the south of the stations show the strongest signals from the upper mantle. The two stations nearest the coast (CBN and NHSC) appear to have signals that are early (CBN) and late (NHSC) compared to the stations to the west in the Appalachian highlands, which have very similar waveforms in good agreement with the time predicted by the IASPEI model. Work to determine whether these differences are due to crustal structure or actual variation in upper mantle structure is on-going.