Quantifying downed coarse woody material and residual forest basal area following retention harvesting in northeastern Minnesota using Landsat sensor data

Abstract

Retention harvesting shows great promise for restoring and maintaining forest structural and compositional diversity. However, economical, comprehensive monitoring is needed to advance understanding of the effectiveness of these management strategies through time. We investigate multitemporal winter Landsat sensor data (capturing snow ground cover at 7.6 cm and 106.7 cm depths) as a tool for discriminating between and providing regional estimates of both residual forest basal area (BA) and downed coarse woody material (DCWM) volume following retention harvesting in Minnesota, USA. Measurements from 34 ground plots were used with Landsat predictor variables to estimate these two biophysical forest parameters. According to similar studies, results for DCWM volume estimation are considered adequate, with an R2adj = 0.54 and absolute RMSE (RMSEa) = 19.02 m3·ha−1. Residual forest BA estimates were similar: total BA R2adj = 0.55 (RMSEa = 1.85 m2·ha−1), hardwood BA R2adj = 0.67 (RMSEa = 1.23 m2·ha−1), and conifer BA R2adj = 0.52 (RMSEa = 0.94 m2·ha−1). Use of winter Landsat imagery was key to quantifying these important forest biophysical parameters — a tool that carries the potential to transform our understanding of the impact of human and natural disturbance regimes on northern forest ecosystems.