Estimating merchantable and non-merchantable wood volume in slash walls using terrestrial and airborne LiDAR

Abstract

Slash walls are an effective forest management practice to prevent deer herbivory and promote forest regeneration following a timber harvest. However, the material cost, i.e. amount of wood material, needed for quality slash walls is unknown or at best based on highly uncertain empirical estimation. Quantifying how total and merchantable timber wood volume varies with width and height across slash walls will facilitate future planning and application of slash walls. In this study, we estimated wood volume per unit length (∼30.48 m) from total stem cross-sectional area in 40 randomly selected cut-through passages cut into 14 slash walls located at Cornell University’s Arnot Teaching and Research Forest (ATRF) in south-central New York. Within each passage we used Terrestrial Laser Scanning (TLS) to identify individual stems and obtain cross-sectional surface area. We found that TLS derived stem diameters were highly consistent with manual stem diameter measurements for 10 selected wall passages (N = 308, R² = 0.933, RMSE = 1.881 cm). However, TLS tends to underestimate total cross-sectional area due to omission of small stem cross-sections (diameter < 5.08 cm). Cross-sections of small stems often have a low number of total points even when point density is high, especially when they are not facing the scanner. They are also more vulnerable to occlusion (i.e., lack of points due to blocking by other stems). TLS-based wood volume estimates were similar across both short and tall slash walls in the study with an average volume of 21.34 m³/30 m for short slash walls and 21.50 m³/30 m for tall slash walls, corresponding to 16.9 t dry biomass /30 m and 17.0 t dry biomass /30 m respectively. The small difference in woody volume was due to tall slash walls achieving additional height mainly through non-compacted small diameter crown and brush tops that have little woody volume. Average merchantable (diameter > 15.24 cm) and non-merchantable (diameter < 15.24 cm) timber fraction was 68.75 % and 31.35 % for short slash walls and 65.04 % and 34.96 % for tall slash walls. We further assessed slash wall height and width using Airborne Laser Scanning (ALS) for eight other tall slash walls. We found TLS measurements are representative of whole slash wall height and width variability based on ALS measurements. Our results provide novel data and methodology to estimate the construction cost of slash walls which are critical for optimizing slash wall applications.