Q6 – S. Lamb

Improving pre-harvest ground verification of commercial thinning using LiDAR-derived forest inventory

Sean’s work under AWARE examined methods to impute tree lists for individual light detection and ranging (LiDAR) grid cells (20 × 20 m)  in spruce (Picea sp.) plantations, in order to forecast inventory variables using a locally calibrated tree-list growth model. Sean evaluated a method to match stand structural variables estimated from LiDAR to those in a library of over 5500 sample plot measurements to impute tree lists for LiDAR grid cells across 83 000 ha of spruce plantations in J.D. Irving, Limited’s Black Brook District in northwestern New Brunswick. Matches were determined 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. Forest inventory variables obtained by the plot matches were highly correlated (r = 0.91–0.99) with those measured on 98 validation plots. Basal area distributions derived from plot matching were statistically equivalent to those observed on the validation plots 86% of the time (α = 0.05). When the predictions for all validation plots were aggregated, there was minimal difference between predicted and actual basal area distributions by planted species and species compositions were similar. Overall, results indicated that tree-level inventory could be imputed for LiDAR grid cells using existing forest inventory plot measurements and suggested that the tree-level inventory was acceptably accurate for updating future forest inventory using a tree-list growth model.

To evaluate the accuracy of tree-level inventory projections, Sean then compared inventory variable increments predicted with a locally calibrated tree-list growth model (Open Stand Model) using tree lists imputed by plot matching with those using measured tree lists from 98 validation plots. Sean found that total and merchantable basal area, total and merchantable volume, Lorey’s height, and quadratic mean diameter increments were highly correlated (0.75–0.86) with percent root mean squared error ranging from 13%–49%. For validation plots with ≤ 10% LiDAR-derived error for all plot-matched variables, percent root mean squared error was much lower (5%–13%). When compared with volumes from 15 blocks harvested 3–5 years after LiDAR acquisition, average forecasted volume differed by only 1.5% and was not significantly different (p = 0.54). In contrast, unforecasted LiDAR-derived volume was significantly different (p = 0.032) from actual harvested volumes, which showed added value in forecasting LiDAR-derived inventory using the plot-matching method.

To demonstrate the novel application of this method for operational management decisions, annual commercial thinning was planned at grid-cell resolution from 2018–2020 using forecasted inventory variables and commercial thinning eligibility rules.

 

Commercial thinning (CT) ranking for LiDAR grid cells in (a) Black Brook District and (b) one sample stand using inventory variables derived from tree lists imputed by plot matching and forecasted to 2018 using Open Stand Model. Cell ranking was determined using J.D. Irving, Limited commercial thinning eligibility rules for first commercial thinning, which consider species, age, gross merchantable volume, mean tree gross merchantable volume, mean annual height increment, and live crown ratio (Lamb et al. 2018).

 

Publications

Lamb, S.M., MacLean, D.A., Hennigar, C.R., and Pitt, D.G. 2017. Imputing tree lists for New Brunswick spruce plantations through nearest-neighbor matching of airborne laser scan and inventory plot data. Can. J. Remote Sens. 43(3): 269–285.

Lamb, S.M., MacLean, D.A., Hennigar, C.R., and Pitt, D.G. 2018. Forecasting forest inventory using imputed tree lists for LiDAR grid cells and a tree-list growth model. Forests 9(4): 167.

Presentations

Lamb, S.M. 2017. Enhanced forest planning. Oral presentation at New Brunswick Enhanced Inventory Project – Team Meeting. March 21. Fredericton, NB, Canada.

Lamb, S.M. 2017. LiDAR-derived forest inventory. Oral presentation at Remsoft User’s Group. June 14. Fredericton, NB, Canada.

Lamb, S.M. 2017. Improving pre-harvest ground verification of commercial thinning in spruce plantations using a nearest neighbor tree-list imputation method. Oral presentation at AWARE Annual General Meeting. May 17. Edmundston, NB, Canada.

Lamb, S.M. 2017. Improving pre-harvest ground verification of commercial thinning in spruce plantations using a nearest neighbor tree-list imputation method. Oral presentation at New Brunswick Enhanced Inventory Project – Team Meeting. January 11. Fredericton, NB, Canada.

Lamb, S.M. 2016. Tree-list imputation of LiDAR cells in New Brunswick softwood plantations. Oral presentation at AWARE Annual General Meeting. May 25. Cornerbrook, NL, Canada.

Lamb, S.M. 2016. Forecasting development of planted stands in New Brunswick using LiDAR. Oral presentation at New Brunswick Enhanced Inventory Project – Team Meeting. January 19. Fredericton, NB, Canada.

Lamb, S.M. 2015. Forecasting development of planted stands in New Brunswick using LiDAR. Oral presentation at New Brunswick Enhanced Inventory Project – Team Meeting. June 18. Edmundston, NB, Canada.