Road Sediment Models: Strengths and Necessary Improvements

Managed forests require a road network engineered for heavy loads of equipment and products. Due to the extent of forest roads, directly measuring road surface erosion is impractical, leading  to development of models. We test three models (WEPP, GRAIP and SEDMODL2) in watersheds across the US by comparing measured sediment production from road segments to model predictions. Best management practices (BMPs) reduce sediment from erosion of cutslopes, ditches, and road delivery to streams.  Empirical models estimate erosion using relationships between erosion and erodibility and infiltration of underlying geology/soil, precipitation, road prism components, tread surfacing, cut/fill slope cover, ditch conditions, disturbance history, road drainage patterns and area, and interception of groundwater by cut slopes.  Physically based models (e.g. WEPP) use equations governing the physics of surface erosion to calculate runoff, surface erosion, and transport.  Results suggest models can predict percentages of improvement, but are fair to poor predictors of actual amounts of sediment generated on road surfaces unless calibrated. Better predictive capabilities of erosion variations due to climate and soil differences and better estimates of fractional sediment delivery from road segments would improve model utility for broad spatial scales and wider applications to road related water quality and biological concerns.