Holocene Paleoseismology of the Steamboat Mountain Site: Evidence for Full-Length Rupture of the Teton Fault, Wyoming

Abstract

ABSTRACT The 72-km-long Teton fault in northwestern Wyoming is an ideal candidate for reconstructing the lateral extent of surface-rupturing earthquakes and testing models of normal-fault segmentation. To explore the history of earthquakes on the northern Teton fault, we hand-excavated two trenches at the Steamboat Mountain site, where the east-dipping Teton fault has vertically displaced west-sloping alluvial-fan surfaces. The trenches exposed glaciofluvial, alluvial-fan, and scarp-derived colluvial sediments and stratigraphic and structural evidence of two surface-rupturing earthquakes (SM1 and SM2). A Bayesian geochronologic model for the site includes three optically stimulated luminescence ages (∼12–17  ka) for the glaciofluvial units and 16 radiocarbon ages (∼1.2–8.6  ka) for the alluvial-fan and colluvial units and constrains SM1 and SM2 to 5.5±0.2  ka, 1σ (5.2–5.9 ka, 95%) and 9.7±0.9  ka, 1σ (8.5–11.5 ka, 95%), respectively. Structural, stratigraphic, and geomorphic relations yield vertical displacements for SM1 (2.0±0.6  m, 1σ) and SM2 (2.0±1.0  m, 1σ). The Steamboat Mountain paleoseismic chronology overlaps temporally with earthquakes interpreted from previous terrestrial and lacustrine paleoseismic data along the fault. Integrating these data, we infer that the youngest Teton fault rupture occurred at ∼5.3  ka, generated 1.7±1.0  m, 1σ of vertical displacement along 51–70 km of the fault, and had a moment magnitude (Mw) of ∼7.0–7.2. This rupture was apparently unimpeded by structural complexities along the Teton fault. The integrated chronology permits a previous full-length rupture at ∼10  ka and possible partial ruptures of the fault at ∼8–9  ka. To reconcile conflicting terrestrial and lacustrine paleoseismic data, we propose a hypothesis of alternating full- and partial-length ruptures of the Teton fault, including Mw∼6.5–7.2 earthquakes every ∼1.2  ky. Additional paleoseismic data for the northern and central sections of the fault would serve to test this bimodal rupture hypothesis.