Estimation of aerodynamic roughness of a harvested Douglas-fir forest using airborne LiDAR

- Digital terrain models were derived from pre- and post-harvest LiDAR datasets.
- These DTMs were subtracted to quantify a post-harvest surface roughness layer.
- The aerodynamic roughness is estimated from the variability in roughness heights.
- LiDAR-derived aerodynamic roughness agreed with micrometeorological measurements.

The aerodynamic roughness length (z0) is a key variable for the parameterization of momentum, mass and heat exchanges between land surfaces and the atmosphere. Its estimation however is complicated due to the large number of input variables such as height and arrangement of roughness elements on the surface, and its measurement relies on complex micrometeorological instrumentation that is typically unavailable. One remote sensing technology well suited to measuring the height of objects is light detection and ranging (LiDAR). This study demonstrates the use of pre- and post-harvest LiDAR data to quantify the aerodynamic roughness of a post-harvest forest surface. LiDAR data were acquired before and after clearcut harvesting of a 77-ha Douglas-fir dominated site on Vancouver Island, for which a micrometeorological tower provided direct year-long measurements of shear or Reynolds stress (i.e., momentum flux) and wind speed, thus permitting the independent assessment of z0 using the logarithmic wind profile equation.The LiDAR data were used to estimate z0 based on the standard deviation of roughness element heights within the source areas of the micrometeorological tower. Estimated z0 from the LiDAR analysis compared well to z0 calculated using the micrometeorological measurements. The standard deviation of roughness element height estimated from the LiDAR analysis resulted in z0 = 0.13 ± 0.01 m (mean ± SD) for neutral atmospheric stability conditions and z0 = 0.13 ± 0.01 m for all stability conditions. The value of z0 calculated using the logarithmic wind profile equation was 0.13 ± 0.13 m for neutral conditions and 0.12 ± 0.30 m for all stability conditions after applying diabatic profile corrections. The results from this study demonstrate the potential of using LiDAR data to estimate z0 across large areas and in complex situations where direct measurements of z0 are impossible.