Operational implementation of a LiDAR inventory in Boreal Ontario

Abstract

An existing Light Detection and Ranging (LiDAR) data set captured on the Romeo Malette Forest near Timmins, Ontario, was used to explore and demonstrate the feasibility of such data to enrich existing strategic forest-level resource inventory data. Despite suboptimal calibration data, stand inventory variables such as top height, average height, basal area, gross total volume, gross merchantable volume, and above-ground biomass were estimated from 136 calibration plots and validated on 138 independent plots, with root mean square errors generally less than 20% of mean values. Stand densities (trees per ha) were estimated with less precision (30%). These relationships were used as regression estimators to predict the suite of variables for each 400-m 2 tile on the 630 000-ha forest, with predictions capable of being aggregated in any user-defined manner-for a stand, block, or forest-with appropriate estimates of statistical precision. This pilot study demonstrated that LiDAR data may satisfy growing needs for inventory data to scale operational/tactical, through strategic needs, as well as provide spatial detail for planning and the optimization of forest management activities.Key words: forest inventory, Light Detection and Ranging (LiDAR), models, Seemingly Unrelated Regression RÉSUMÉ Un ensemble de données LiDAR (télédétection par laser) recueillies pour la Forêt Roméo Malette près de Timmins en Ontario, a été utilisé pour étudier et démontrer la possibilité d'utiliser ces données pour enrichir les données existantes d'inventaire des ressources forestières de premier plan. Malgré une calibration des données inférieure à ce qui était souhaité, les variables d'inventaire des peuplements comme la hauteur moyenne supérieure, la hauteur moyenne, la surface terrière, le volume brut total, le volume marchand total et la biomasse au-dessus du sol ont été estimées à partir de 136 parcelles de calibration et validées pour 138 parcelles indépendantes, avec une erreur quadratique moyenne géné-ralement inférieure à 20 % des valeurs moyennes. La densité des peuplements (arbres par hectare) a été estimée avec moins de précision (30 %). Ces relations ont été utilisées comme estimateurs de régressions utilisées pour générer une série de variables pour chaque unité de 400 m 2 de la forêt de 630 000 ha, avec des prédictions cumulables selon la requête de l'utilisateur-pour un peuplement, pour un bloc ou pour la forêt-avec des estimations appropriées d'une précision statistique. Ce projet pilote a démontré que les données LiDAR pourraient répondre aux besoins sans cesse croissants en matière de données d'inventaire pour définir les plans opérationnels/tactiques, bien que stratégiques, ainsi que pour établir les détails spatiaux requis pour la planification et l' optimisation des activités d'aménagement forestier.

A Comparison of Airborne Laser Scanning and Image Point Cloud Derived Tree Size Class Distribution Models in Boreal Ontario

“…Ground sampling is expensive and plot size is generally balanced by the number of samples. Prediction unit size is a function of the spatial resolution desired for the inventory attributes, and ~400 m 2 has been found to be a suitable size for both ALS [2] and IPC [3] predictions in the boreal forest. Larger prediction units may be used, but as the size increases, so does the cost of calibration and validation.…”

“…The majority of the ground data were collected on the Hearst Forest during the summer of 2010, according to a previously documented field protocol [2]. A total of 446 circular, 400-m 2 temporary sample plots were established throughout the range of development stages within eight forest types ( Table 1).…”

“…The ALS predictor variables (Table 5) were derived from point-cloud statistics following previously described methods [2]. Veteran trees generally result in a few high ALS returns that have a large influence on some of the measures of spread, such as standard deviation of ALS returns (STD_DEV) and vertical complexity index (VCI).…”

“…ALS predictor variables were derived from point-cloud statistics (Table 5) following the methods described by Woods et al [2]. Table 5.…”

“…Recent studies have shown that these data may be augmented with area-based estimates of growing stock (basal area, volume) and average tree size (height, diameter, volume) derived from either Airborne Laser Scanning (ALS) [1,2] or stereo image point clouds (IPC) [3], to facilitate tactical (5-year) and operational (1-year) planning of forest operations. Since harvesting operations have become more mechanized and processing facilities are increasingly optimized for particular products and sizes of raw materials, there is considerable interest in adding information to inventories on the size assortment of the stems.…”