A Methodology for Planning Road Best Management Practices Combining WEPP: Road Erosion Modeling and Simulated Annealing Optimization

Erosion from forest roads is a known problem in mountainous terrain. To abate these negative consequences, physical Best Management Practices (BMPs) are implemented, sometimes with no knowledge of erosion hot spots. With the need to minimize water quality impacts while at the same time accounting for multiple considerations and constraints, road BMP planning at the watershed scale is a difficult task. To assist in this planning process, a methodology is presented here that combines WEPP: Road erosion predictions with simulated annealing optimization. Under this methodology, erosion predictions associated with BMP options for a segment comprise the objective function of an optimization problem. This methodology was tested on a watershed in the Lake Tahoe Basin. WEPP: Road input data was gathered through road surveys. Modeling results predicted relatively little sediment leaving the forest buffer, as a result of numerous well-maintained BMPs and the dry climate found in the watershed. A sensitivity analysis for all WEPP: Road input parameters is presented, which provides insight into the general applicability of these erosion estimates as well as the relative importance of each input parameter. After evaluating erosion risk across the entire watershed, applicable BMPs were assigned to problem road segments and WEPP: Road was used to predict change in sediment leaving the buffer with BMP implementation at a given site. These predictions, combined with budget constraints as well as equipment scheduling considerations, were incorporated into an algorithm using simulated annealing as its optimization engine. Three modeled scenarios demonstrate the viability of this methodology in reducing total sediment leaving the road buffer over a planning horizon. Of the 173 segments surveyed, 38 segments could be treated using generic BMPs. For all three scenarios, BMP-SA reduced sediment leaving the buffer by as much as 70% over the course of a 20-year planning horizon. For the 38 segments treated with BMPs, sediment was reduced by greater than 90% over the planning horizon. This methodology is a viable approach for streamlining watershed-scale road network BMP planning, despite its heavy reliance on road erosion estimates