Updates to the suspended sediment SPARROW model developed for western Oregon and northwestern California

A SPARROW (SPAtially Related Regressions On Watershed attributes) model that was previously developed for western Oregon and northwestern California was updated using advancements in the SPARROW software and refinements to the input data. As was the case for the original model calibration, the updated models used the NHD Plus Version 2 as a hydrologic framework and relied on the same estimates of long-term mean suspended-sediment loads and watershed attributes. The updated calibration results indicated that two different SPARROW models were possible—one model from which sediment sources were represented by local lithology and one from which sediment sources were represented by generalized land-cover classes; precipitation, catchment slope, wildfire disturbance, and sediment loss in impoundments were significantly correlated with suspended-sediment loads in both models. The updated models also included a method to compensate for the bias introduced by using total suspended solids to represent suspended sediment in the calibration dataset—a feature that was not available during the original model calibration. The effect of this feature was an overall increase in estimated suspended-sediment loads. Although the lithology- and the land-cover based models used different landscape properties to describe sediment sources, each could be useful in specific applications. The lithology-based model provides more accurate estimates of suspended-sediment load, but the land-cover based model allows water-quality managers to estimate how much in-stream suspended-sediment load originates in areas with extensive development compared to the load that originates in areas with relatively little human impact.