Shear velocity structure beneath the central United States: implications for the origin of the Illinois Basin and intraplate seismicity

Abstract

We present new estimates of lithospheric shear velocities for the intraplate seismic zones and the Illinois Basin in the U.S. midcontinent by analyzing teleseismic Rayleigh waves. We find that relatively high crustal shear velocities (VS) characterize the southern Illinois Basin, while relatively low crustal velocities characterize the middle and lower crust of the central and northern Illinois Basin. The observed high crustal velocities may correspond to high‐density mafic intrusions emplaced into the crust during the development of the Reelfoot Rift, which may have contributed to the subsidence of the Illinois Basin. The low crustal VS beneath the central and northern basin follow the La Salle deformation belt. We also observe relatively low velocities in the mantle beneath the New Madrid seismic zone where VS decreases by about 7% compared to those outside of the rift. The low VS in the upper mantle also extends beneath the Wabash Valley and Ste. Genevieve seismic zones. Testing expected VS reductions based on plausible thermal heterogeneities for the midcontinent indicates that the 7% velocity reduction would not result from elevated temperatures alone. Instead this scale of anomaly requires a contribution from some combination of increased iron and water content. Both rifting and interaction with a mantle plume could introduce these compositional heterogeneities. Similar orientations for the NE‐SW low‐velocity zone and the Reelfoot Rift suggest a rift origin to the reduced velocities. The low VS upper mantle represents a weak region and the intraplate seismic zones would correspond to concentrated crustal deformation above weak mantle.

Detailed crustal thickness variations beneath the Illinois Basin area: Implications for crustal evolution of the midcontinent

“…In the region of the Reelfoot Rift, Mooney et al [] and Nelson and Zhang [] observed high‐velocity lower crust in seismic refraction profiles that they interpreted as a “rift pillow” created by magmatic differentiation during the rifting event. Liang and Langston [] and Chen et al [] confirmed the presence of high‐velocity lower crust using tomographic techniques. The result from our study and some other recent studies [ McGlannan and Gilbert , ; Shen and Ritzwoller , ] all show a relatively thin crust beneath the Reelfoot Rift and the Rough Creek Graben.…”

“…The St. Louis Arm, defined by Braile et al [, , ] as a rift extension following the strike of the Mississippi River between Missouri and Illinois, is coincident with the large Moho offset (5–10 km) discussed in section 5.2. Although similar velocity patterns are observed among these three arms in the lower crust and upper mantle [ Chen et al ., ; Savage et al ., ], our results on the spatial distribution of crustal velocity discontinuities contradict the concept of the St. Louis and Indiana Arms. Specifically, both the localized, shallow Moho (Figure ) and the confined, thin crystalline crust (Figure ) suggest that rifting of the crust directly impacted only the areas of the Reelfoot Rift and Rough Creek Graben.…”

“…They proposed a fossilized flat‐slab subduction in the uppermost mantle below about 90 km beneath the Illinois Basin. Shear velocities have revealed heterogeneities beneath the Illinois Basin area [ Chen et al ., ; Chu and Helmberger , ; Pollitz and Mooney , ]. Shen and Ritzwoller [] and McGlannan and Gilbert [] all observed relatively thick crust in the broad southern Illinois Basin area.…”

“…This study of the Illinois Basin and the regions around it benefits from the newly available data of the Ozark, Illinois, Indiana, and Kentucky (OIINK) EarthScope Flexible Array experiment [ Chen et al ., ; Yang et al ., ]. This paper builds upon a recent study by McGlannan and Gilbert [] that used P wave receiver functions from the EarthScope Transportable Array (TA) to produce a regional map of the crust‐mantle boundary similar to the one presented here.…”

Ridge asymmetry and deep aqueous alteration at the trench observed from Rayleigh wave tomography of the Juan de Fuca plate

“…Equalizing the rank on the finer grid increases the variance of the model parameters at long periods more than on the coarse grid, and we found that on the 0.8° grid, we were unable to usefully equalize the rank at 143 s, the longest period, so we dropped that period altogether in the second step. Rau and Forsyth [] and Chen et al [] used multiple steps, but we found that further refinement of the grid was of marginal value, given the noise level and scattering in the waveforms in this experiment.…”