Over the past two decades there has been an abundance of research demonstrating the utility of airborne light detection and ranging (LiDAR) for predicting forest biophysical/inventory variables at the plot and stand levels. However, to date there has been little effort to develop a set of protocols for data acquisition and processing that would move governments or the forest industry towards cost-effective implementation of this technology for strategic and tactical (i.e., operational) forest resource inventories. The goal of this paper is to initiate this process by examining the significance of LiDAR data acquisition (i.e., point density) for modeling forest inventory variables for the range of species and stand conditions representing much of Ontario, Canada. Field data for approximately 200 plots, sampling a broad range of forest types and conditions across Ontario, were collected for three study sites. Airborne LiDAR data, characterized by a mean density of 3.2 pulses m⁻² were systematically decimated to produce additional datasets with densities of approximately 1.6 and 0.5 pulses m⁻². Stepwise regression models, incorporating LiDAR height and density metrics, were developed for each of the three LiDAR datasets across a range of forest types to estimate the following forest inventory variables: (1) average height (R²(adj) = 0.75–0.95); (2) top height (R²(adj) = 0.74–0.98); (3) quadratic mean diameter (R²(adj) = 0.55–0.85); (4) basal area (R²(adj) = 0.22–0.93); (5) gross total volume (R²(adj) = 0.42–0.94); (6) gross merchantable volume (R²(adj) = 0.35–0.93); (7) total aboveground biomass (R²(adj) = 0.23–0.93); and (8) stem density (R²(adj) = 0.17–0.86). Aside from a few cases (i.e., average height and density for some stand types), no decimation effect was observed with respect to the precision of the prediction of the majority of forest variables, which suggests that a mean density of 0.5 pulses m⁻² is sufficient for plot and stand level modeling under these diverse forest conditions across Ontario.

Leaf Area Index (LAI) is an important input variable for forest ecosystem modeling as it is a factor in predicting productivity and biomass, two key aspects of forest health. Current in situ methods of determining LAI are sometimes destructive and generally very time consuming. Other LAI derivation methods, mainly satellite-based in nature, do not provide sufficient spatial resolution or the precision required by forest managers for tactical planning. This paper focuses on estimating LAI from: (i) height and density metrics derived from Light Detection and Ranging (LiDAR); (ii) spectral vegetation indices (SVIs), in particular the Normalized Difference Vegetation Index (NDVI); and (iii) a combination of these methods. For the Hearst Forest of Northern Ontario, in situ measurements of LAI were derived from digital hemispherical photographs (DHPs) while remote sensing variables were derived from low density LiDAR (i.e., 1 m⁻²) and high spatial resolution WorldView-2 data (2 m). Multiple Linear Regression (MLR) models were generated using these variables. Results from these analyses demonstrate: (i) moderate explanatory power (i.e., R² = 0.53) for LiDAR height and density metrics that have proven to be related to canopy structure; (ii) no relationship when using SVIs; and (iii) no significant improvement of LiDAR models when combining them with SVI variables. The results suggest that LiDAR models in boreal forest environments provide satisfactory estimations of LAI, even with narrow ranges of LAI for model calibration. Models derived from low point density LiDAR in a mixedwood boreal environment seem to offer a reliable method of estimating LAI at high spatial resolution for decision makers in the forestry community. This method can be easily incorporated into simultaneous modeling efforts for forest inventory variables using LiDAR.

As a result of the warming observed at high latitudes, there is significant potential for the balance of ecosystem processes to change, i.e., the balance between carbon sequestration and respiration may be altered, giving rise to the release of soil carbon through elevated ecosystem respiration. Gross ecosystem productivity and ecosystem respiration vary in relation to the pattern of vegetation community type and associated biophysical traits (e.g., percent cover, biomass, chlorophyll concentration, etc.). In an arctic environment where vegetation is highly variable across the landscape, the use of high spatial resolution imagery can assist in discerning complex patterns of vegetation and biophysical variables. The research presented here examines the relationship between ecological and spectral variables in order to generate an ecologically meaningful vegetation classification from high spatial resolution remote sensing data. Our methodology integrates ordination and image classifications techniques for two non-overlapping Arctic sites across a 5° latitudinal gradient (approximately 70° to 75°N). Ordination techniques were applied to determine the arrangement of sample sites, in relation to environmental variables, followed by cluster analysis to create ecological classes. The derived classes were then used to classify high spatial resolution IKONOS multispectral data. The results demonstrate moderate levels of success. Classifications had overall accuracies between 69%–79% and Kappa values of 0.54–0.69. Vegetation classes were generally distinct at each site with the exception of sedge wetlands. Based on the results presented here, the combination of ecological and remote sensing techniques can produce classifications that have ecological meaning and are spectrally separable in an arctic environment. These classification schemes are critical for modeling ecosystem processes.

This study examined vegetation greening at two arctic sites: the Apex River Watershed (ARW), Baffin Island, Nunavut (a Low Arctic site) and the Cape Bounty Arctic Watershed Observatory (CBAWO), Melville Island, Nunavut (a High Arctic site). The vegetation at both study sites was characterized using a supervised land-cover classification approach using high spatial resolution satellite remote sensing data (i.e., IKONOS [4 m] and WorldView-2 [2 m]). Meanwhile, Normalized Difference Vegetation Index (NDVI) data spanning the past 30 years were derived from intermediate spatial resolution data (i.e., Landsat TM/ETM/OLI [30 m]). The land-cover classifications were used to partition the Landsat NDVI time series by vegetation type. Climate variables (i.e., temperature, precipitation, and growing season length [GSL]) were examined to explore potential relationships of NDVI to climate warming trends. The results of the land-cover classifications demonstrated inherent trends of vegetation types along elevat...

Airborne laser scanning (ALS) systems tuned to the near-infrared (NIR; 1064 nm) wavelength have become the best available data source for characterizing vegetation structure. Proliferation of multi-spectral ALS (M-ALS) data with lasers tuned at two additional wavelengths (commonly 532 nm; green, and 1550 nm; short-wave infrared (SWIR)) has promoted interest in the benefit of additional wavelengths for forest inventory modelling. In this study, structural and intensity based M-ALS metrics were derived from wavelengths independently and combined to assess their value for modelling forest inventory attributes (Lorey’s height (HL), gross volume (V), and basal area (BA)) and overstorey species diversity (Shannon index (H), Simpson index (D), and species richness (R)) in a diverse mixed-wood forest in Ontario, Canada. The area-based approach (ABA) to forest attribute modelling was used, where structural- and intensity-based metrics were calculated and used as inputs for random forest models. Structural metrics from the SWIR channel (SWIR_struc) were found to be the most accurate for H and R (%RMSE = 14.3 and 14.9), and NIR_struc were most accurate for V (%RMSE = 20.4). The addition of intensity metrics marginally increased the accuracy of HL models for SWIR and combined channels (%RMSE = 7.5). Additionally, a multi-resolution (0.5, 1, 2 m) voxel analysis was performed, where intensity data were used to calculate a suite of spectral indices. Plot-level summaries of spectral indices from each voxel resolution alone, as well as combined with structural metrics from the NIR wavelength, were used as random forest predictors. The addition of structural metrics from the NIR band reduced %RMSE for all models with HL, BA, and V realizing the largest improvements. Intensity metrics were found to be important variables in the 1 m and 2 m voxel models for D and H. Overall, results indicated that structural metrics were the most appropriate. However, the inclusion of intensity metrics, and continued testing of their potential for modelling diversity indices is warranted, given minor improvements when included. Continued analyses using M-ALS intensity metrics and voxel-based indices would help to better understand the value of these data, and their future role in forest management.