The response of forest roads on steep, landslide-prone terrain in western Oregon to the February 1996 storm

Following the February 5-9,1996 storm in western Oregon, a large,infrequent landslide-producing storm, twoinventories of road-related landslides wereundertaken to evaluate the response ofcontemporary forest road networks to largestorms. A total of 386 miles of forest roadwere driven or walked and 128 road-relatedlandslides were inventoried on federal, state,and private land in both the Oregon Coastand Cascade Ranges. Both landscapevariables, such as road age and topographicposition, and site-specific road variables,such as road grade, length, cutslope height,and fill depth, among others, were collected.When the results of the inventories of road-related landslides following the February1996 storm were compared with the resultsof previous inventories of road-relatedlandslides on the same or similar terrain,there was no obvious difference in landslidedensity values either on a road length orroad prism area basis. This indicates thatthere is a cost of doing business when roadsare constructed in steep, landslide-proneterrain. However, the data also showed thatthe average landslide volume had decreased,in some cases significantly, andcontemporary road densities were less thanactual or predicted road densities fromprevious decades. This means that while thedensity of road-related landslides hasn'tdecreased, with fewer roads and, on average,smaller landslides there should have beenless road-related erosion generated per unitarea of forest in the 1996 storm. On a site-specific scale, approximately half of theshallow, translational landslides thatinitiated in the road fill were associated withthe outfall of surface road drainagestructures, which means that road drainageis a feature of forest road systems thatsimply must be accounted for. Also,cutslope heights greater than 15 feet werepositively correlated with both theoccurrence and volume of road-related landslides. On a landscape scale, most of theroad-related erosion came from roadsconstructed during or before the 1960'srather than roads constructed since the1960's. Also, the majority of the road-related erosion was associated with roadsoccupying a rnidslope topographic positionversus ridge or valley bottom roads.
KEY WORDS--Slope stability, landslides,landslide inventory, road-related landslides,forest roads, forest road drainage, February1996 storm