Stratigraphic relations and radiocarbon ages of deposits exposed in several trenches and excavations help to establish the timing, sense of slip, and style of the deformation that resulted from late Holocene surface faulting on the Meers fault in southwestern Oklahoma. The eastern half of the scarp is formed on relatively ductile Permian Hennessey Shale and Quaternary alluvium, whereas the western half is formed on well-lithified, relatively brittle Permian Post Oak Conglomerate in the Slick Hills.
At Canyon Creek on the eastern half of the scarp, the shale and alluvium in two trenches are deformed mainly by monoclinal warping. These trenches contain stratigraphic evidence of one surface-faulting event that produced about 3 m of throw. At this site, the amount of throw in middle Holocene and middle Pleistocene deposits is similar. Lateral displacement is difficult to detect in these trenches, most likely because of plastic deformation in the shale and alluvium.
In contrast, trenches and excavations on the western half of the scarp show that the Holocene surface faulting produced at least as much lateral as vertical displacement. At two sites, the scarp has dammed small gullies and ponded fine-grained alluvium upslope from the scarp. The channels of the gullies at these ponded-alluvium sites have been separated 3-5 m left-laterally since they were dammed. The lateral displacement on the gullies is 3.3 to 1.6 times as much as the vertical displacement. In a pit excavated into the colluvium on the downthrown side of the scarp, subhorizontal striae on conglomerate clasts along the fault plane provide evidence of nearly pure strike-slip movement. The age of the striae is unknown, but they are believed to be Quaternary in age because it is unlikely that such delicate striae could be preserved in soluble carbonate rock in a near-surface weathering environment for many hundreds of thousands of years.
Multiple radiocarbon ages of soil-humus samples from the Canyon Creek trenches and the ponded-alluvium sites show that the last surface faulting occurred 1,200-1,300 yr ago. Limited geologic evidence, however, indicates a long-term recurrence interval on the order of 100,000 yr or more.
The youthful surface faulting compared to the apparently long recurrence interval presents a difficult problem for regional seismichazard assessments. Hazard assessments that rely on the long-term slip rate might seriously underestimate the hazard if the behavior of the fault is characterized by a temporal clustering of events, and if the late Holocene surface faulting signals the beginning of a period of frequent faulting. Conversely, if strain accumulates steadily on the Meers fault and is released regularly over time intervals of 100,000 yr or more, then the hazard may be low because much of the stored strain was released only about 1,000 yr ago. Improved earthquake-hazard assessments in much of the central United States and in stable intraplate settings worldwide require a better understanding of the long-term and short-term behavior of seismogenic intraplate faults.