Soil catenas to estimate ages of movements on normal fault scarps, with an example from the Wasatch fault zone, Utah, USA

Abstract

Trenches excavated across fault scarps on a 13 ka delta surface at Brigham City, Utah, USA, document four types of soil-catena phenomena. First, the relict summit soil is eroded at the scarp crest, and its components incorporated into the accumulating colluvial wedge at the scarp base. Second, the pre-faulting soil weakens laterally as it is traced scarpward beneath the colluvial wedge from a relict position beyond the scarp toe. Third, soils developed on individual colluvial wedges and on the underlying, pre-faulting surface merge downslope to form a cumulative soil profile at the scarp toeslope. Fourth, soil horizons thicken downslope, probably due to the combined affect of more available moisture, water infiltration, and sediment accumulation. Two soil catenas were used to evaluate the applicability of a continuity approach to paleoearthquake dating, by which percentages of total development time represented by each soil at a fault scarp footslope could be estimated. The continuity approach sets total soil development at the scarp footslope, including any buried pre-faulting soil plus the sum development of all colluvial wedge soils corrected by a slope factor ( F), equal to total soil development at the scarp summit. Soil development was characterized by Profile Development Index (PDI) and pedogenic clay parameters, using facies models to estimate parent material values for genetically-different colluvial units, and recognizing inherited soil components in colluvial wedges as parent material properties. The F slope factor, used to correct colluvial wedge soils for development attributable to slope position rather than age, was checked by using a nearby fluvial scarp as an independent control. Our soil data, when compared to independent age estimates, imply that rates of soil development have not been constant over the last 13 ka at Brigham City, but rather were considerably higher than average between 13 ka and 8.3 ka. Higher rates are probably due to a relatively high influx of eolian dust during the late Pleistocene to early Holocene following the lowering of Lake Bonneville, a large paleolake in the western USA. These variable rates of soil development preclude using the continuity approach for paleoearthquake dating at Brigham City. However, promising results showing that generally consistent F factors can be calculated for fault and fluvial scarps leave open the possibility that in areas where soil development rates have been relatively constant over the time period in question, the continuity approach may be a broadly applicable method for paleoearthquake dating.