Forecasting forest inventory for spruce plantations using airborne laser scanning data

Abstract

This thesis focuses on the imputation of tree-level inventory for light detection and ranging (LiDAR) grid cells across 83 000 ha of spruce (Picea sp.) plantations to permit forecasting forest inventory using a tree-list growth model. Grid cells with LiDAR-derived inventory predictions were matched with one of over 5500 spruce plantation sample plot measurements based on planted species and minimum sum of squared difference between total and merchantable basal area, total and merchantable volume, top height, and merchantable quadratic mean diameter. Imputed tree lists resulted in inventory variable values that were highly correlated with observed values and statistically equivalent basal area distributions. When input into a locally calibrated tree-list growth model, variable increments predicted using imputed tree lists were strongly correlated with those using measured tree lists, resulting in predicted volumes that were similar to observed volumes. Using forecasted inventory variables, annual commercial thinning treatments were planned at grid-cell resolution from 2018–2020.