FOLD FORM AND FAULT GEOMETRY AT UMTANUM RIDGE, YAKIMA FOLD BELT, WASHINGTON

The US Pacific Northwest contains abundant evidence for geologically recent deformation on upper crustal faults, which pose a seismic hazard to major energy and industrial facilities throughout the region. Located in the backarc of Cascadia, the Yakima Folds in central Washington record Miocene-to-present deformation of the Columbia River Basalt Group (CRBG). Several lines of evidence suggest that deformation of the folds is ongoing and the underlying faults may pose a seismic hazard. Prior work estimates maximum earthquake magnitudes on the underlying faults to be in the range M­­­_w5.2 to M_w7.8, based on probabilistic seismic hazard analyses. The large range stems, in part, from the uncertainty in fault geometry. In order to better understand the seismic hazard posed by the Yakima folds in central Washington, we present a new geologic map incorporating interpretation of high-resolution LiDAR DEMs for the area surrounding the intersection of Umtanum Ridge and Yakima Canyon. Using these data to constrain the geometry of the anticline comprising Umtanum Ridge, we test a range of two-dimensional kinematic models to investigate how fold form relates to fault geometry beneath the structure. The best-fit model combines fault-bend and trishear fault-propagation folding to match measured flow top attitudes along a line of section through the ridge. The preferred model is produced by a steeply-dipping reverse fault extending through the CRBG and into the underlying Paleogene sedimentary rocks. This model implies that up to 520 m of dip slip has occurred since 15.6 Ma on a fault soling in a detachment ~ 4 km deep. Such an interpretation, if correct, would imply that the seismic hazard posed by Umtanum Ridge, and perhaps other Yakima folds, tends towards the upper end of previous estimates. The preferred model suggests long-term horizontal contraction of the fold of ~0.01 mm/yr and uplift of the crest of ~0.03 mm/yr, about half the previously published rates derived from geodetic and geomorphic data, respectively. The discrepancy implies that current rates of deformation may be higher than the long-term average. Our two-dimensional approach captures only dip-slip and does not address the possibility of significant strike slip deformation proposed recently by other workers.