THE NATURE AND TRANSPORT OF THE FINE-GRAINED COMPONENT OF SWIFT CREEK LANDSLIDE, NW WASHINGTON

Extreme sedimentation in Swift Creek, located in the Cascades foothills in NW Washington (48°54'45”, 122°16') is due to erosion from the oversteepened, unvegetated toe of a large (0.55 km^2) active earthflow. The amount of bedload in the creek has necessitated several mitigation projects in the channel including annual dredging and temporary sediment traps in an attempt to reduce the risk of flooding and damage to manmade structures downstream. The amount of suspended sediment transported by Swift Creek poses a threat to downstream ecosystems in the Sumas River due to the turbidity that it introduces into this fish-producing river. One objective of this study was to measure the annual suspended sediment yield in Swift Creek and to see if precipitation and/or discharge could be used as predictors for suspended sediment concentrations. Since March of 2005, the suspended sediment concentrations ranged from 0.02 g/L to 41.6 g/L.

The nature of suspended sediment in the creek is a direct result of the weathering process of the serpentinitic bedrock in which the landslide is rooted. In order to better understand the transport of the fine-grained component in Swift Creek, the second objective of this study used XRD and SEM analyses to decipher the weathering process of the serpentinite: from the bedrock, to the fine-grained surface component, to its transport in Swift Creek.

The serpentinite is weathering to chrysotile asbestos with minor amounts of chlorite, illite and hydrotalcite, all of which occur as clay seeps on the unvegetated surface of the landslide. By volume, chrysotile makes up at least 50% of the suspended load transported in Swift Creek. The chemical makeup of the suspended load therefore poses health concerns wherever it is transported to. Modeling is in progress to estimate the total suspended sediment flux from detailed precipitation records. An estimate of the total suspended sediment flux, known as a result of this study, coupled with the already ascertained estimates of the bedload, can be used in the design of future mitigation projects to control the effect of the landslide evident downstream.