Crustal studies, velocity inversions, and fault tectonics: Results from a microearthquake survey in the New Madrid Seismic Zone

Abstract

A dense network of portable seismographs was operated in the Ridgely area of western Tennessee in May of 1978. During 32 days of operation, 122 earthquakes were detected, of which 90 were located. The high quality P and S arrivals recorded exhibit travel time ratios (ts/tp) that decrease significantly with depth. This decrease is most rapid in the upper few kilometers and necessarily reflects similar changes in the velocity ratio (Vp/Vs). Based on laboratory and theoretical studies, such observations are consistent with a decrease in crack density with increasing confining pressure. A special study of Vp, Vs, and Vp/Vs revealed strong variations with depth that also correlate with rock type. Inversion of the data for both P and S velocity structures confirms the presence of a low‐velocity zone in the depth range 2 to 5 km; a high‐velocity lid, with high Vp/Vs and thus, high crack density; and a boundary at the near surface for S to P conversion, associated with the unconsolidated saturated sediments of the Mississippi embayment. Relocated hypocenters using independent velocity structures for P and S are systematically deeper and define a seismogenic layer between 4 and 13 km depth. The earthquakes near Ridgely show that most of the activity is occurring on a northwest striking reverse fault that dips to the southwest. Minor off‐fault activity within the hanging wall was also observed representing motion on a series of small normal faults dipping to the northeast. Other composite focal mechanisms were determined for nearby regions. These solutions are all in general agreement with a horizontal compressive stress field striking northeast to east, favoring right‐lateral strike‐slip motion on left‐stepping northeast trending faults, or reverse motion on faults trending northwest and north. These results are consistent with the stress field found in a large area interior to the United States, and imply a rotation of the regional stress field since the late Cretaceous when extensional faulting occurred in the area.

The relocation of microearthquakes in the northern Mississippi Embayment

“…But it is not clear why only faults along the rift axis and a small segment of the rift boundary are active, while other similarly oriented faults bounding the rift are apparently inactive. The first is that the focal mechanisms predominantly show a combination of strike-slip and reverse motions occurring either on northeast trending faults that parallel the axial seismicity trend or on northwest trending faults that are aligned with the seismicity pattern (O'Connell et al [1982], Nicholson et al ['1984], and this study). This hypothesis can be reasonably invoked to explain the concentration of seismicity between Ridgely and New Madrid because gravity and aeromagnetic data indicate a ring dike complex south of New Madrid [Hildenbrand et al, 1977[Hildenbrand et al, , 1982, seismic reflection profiles reveal lacolithic intrusions of estimated middle Eocene age in the Ridgely area [Zoback, 1979;Zoback et al, 1980], and seismic refraction studies suggest that the altered lower crustal layer, which is considered a zone of mantle intrusives, thickens beneath this area (Figure 3).…”

Relocation of Earthquakes in the New Madrid Seismic Zone: Estimation of 1D Velocity Structure and Geometry of a Seismogenic Fault

“…The prescribed velocity ranges used in this study with CMVEB (Nuttli et al, 1969), the P-wave velocity from inversion (Nicholson et al, 1984), AM85 (Andrews et al, 1985), and PANDA (Chiu et al, 1992). The model used in this study consists of 31 crustal layers; the prescribed ranges are shown with the upper (pink line) and the lower limit (blue line).…”

“…In this work, we used 100 initial models for computational efficiency. The prescribed P-wave velocity ranges used in this study, as well as those used with other velocity models (CMVEB [Nuttli et al, 1969], P-wave velocity from inversion [Nicholson et al, 1984], AM85 [Andrews et al, 1985], and PANDA [Chiu et al, 1992]) are shown in Figure 3. The S-wave velocities are taken as 1= 3 p of the P-wave velocities (Pujol et al, 1997;Kim et al, 2010).…”

“…Park et al (2007) reviewed 71 earthquakes to evaluate the stress field in the southern Korean Peninsula and demonstrated that R and D functions are useful for weak phases, especially Pn arrival. The correct phase identification is crucial to determine the hypocentral parameters, because two secondary arrivals, S-to-P (S P ) and P-to-S (P S ), which are converted from the bottom of the unconsolidated sediments in the Mississippi embayment, can be misidentified as S phase (Nicholson et al, 1984;Andrews et al, 1985).…”

“…Chiu et al (1992) first located earthquakes recorded by the PANDA network and imaged the associated fault planes using the so-called PANDA velocity model. Chiu et al (1992Chiu et al ( , 1997) showed the PANDA model provided a significant improvement in hypocenter determinations over previous studies that used other velocity models, such as those of Nuttli et al (1969), Mooney et al (1983), Nicholson et al (1984), and Andrews et al (1985). Pujol et al (1997) and Mueller and Pujol (2001) relocated earthquakes in the NMSZ using the joint hypocenter determination (JHD) technique, which partly accounted for lateral velocity variations.…”