Evaluating post‐wildfire logging‐slash cover treatment to reduce hillslope erosion after salvage logging using ground measurements and remote sensing

• Published in: Hydrological processes 2020-11, Vol.34 (23), p.4431-4445
• Main Authors: Robichaud, Peter R., Lewis, Sarah A., Brown, Robert E., Bone, Edwin D., Brooks, Erin S.
• Format: Article
• Language: English
• Subjects: Dead wood; Debris flow; Evaluation; Experiments; Field tests; Fires; Flooding; Flow rates; Flow velocity; Ground cover; Imagery; Logging; post‐fire; Rain; Rainfall; recovery; Reduction; Remote sensing; rill erosion; Salvage; salvage logging; Salvaging; Satellite imagery; Sediment; Sediment yield; Sediments; Silt; silt fence; skid trail; Soil; Soil erosion; Spaceborne remote sensing; Trails; Vegetation index; Watersheds; Wildfires; WorldView
• ISSN: 0885-6087; 1099-1085
• DOI: 10.1002/hyp.13882

Continuing long and extensive wildfire seasons in the Western US emphasize the need for better understanding of wildfire impacts including post‐fire management scenarios. Advancements in our understanding of post‐fire hillslope erosion and watershed response such as flooding, sediment yield, and debris flows have recently received considerable attention. The potential impacts of removing dead trees, called salvage logging, has been studied, however the use of remotely sensed imagery after salvage logging to evaluate spatial patterns and recovery is novel. The 2015 North Star Fire provided an opportunity to evaluate hillslope erosion reduction using two field experiments and coincidental remotely sensed imagery over 3 years. Simulated rill experiments with four flow rates were used to quantify hillslope erosion on skidder trails with and without added logging slash compared with a burned‐only control. Seven replicated hillslope silt fence plots with the same treatments were also evaluated for natural rainfall events. WorldView‐2 satellite imagery was used to relate ground cover and erodible bare soil between the two experiments using multi‐temporal Normalized Differenced Vegetation Index (NDVI) values. Results indicate that the skid trails produced significantly more sediment (0.70 g s−1) than either the slash treated skid trail (0.34 g s−1) or controls (0.04 g s−1) with the simulated rill experiment. Similarly, under natural rainfall conditions sediment yield from hillslope silt fence plots was significantly greater for the skid trail (3.42 Mg ha−1) than either the slash treated skid trail (0.18 Mg ha−1) or controls (0 Mg ha−1). An NDVI value of 0.32 on all plots over all years corresponded to a ground cover of about 60% which is an established threshold for erosion reduction. Significant relationships between NDVI, ground cover, and sediment values suggest that NDVI may help managers evaluate ground cover and erosion potential remotely after disturbances such as a wildfire or salvage logging. Post‐wildfire salvage logging can create a highly disturbed landscape that is susceptible to runoff and erosion events following precipitation because of the increased soil exposure and compaction. Significant relationships between field ground cover measures, sediment yield values from two experiments, and NDVI (Normalized Difference Vegetation Index) values from satellite images suggest that NDVI may help managers evaluate ground cover, and in turn estimate erosion