Voxelization is an effective approach to reduce the computational cost of processing Light Detection and Ranging (LiDAR) data, yet it results in a loss of fine-scale structural information. This study explores whether low-level voxel content information, specifically target occupancy percentage within a voxel, can be inferred from high-level voxelized LiDAR point cloud data collected from Digital Imaging and remote Sensing Image Generation (DIRSIG) software. In our study, the targets include bark, leaf, soil, and miscellaneous materials. We propose a multi-target regression approach in the context of imbalanced learning using Kernel Point Convolutions (KPConv). Our research leverages cost-sensitive learning to address class imbalance called density-based relevance (DBR). We employ weighted Mean Saquared Erorr (MSE), Focal Regression (FocalR), and regularization to improve the optimization of KPConv. This study performs a sensitivity analysis on the voxel size (0.25 - 2 meters) to evaluate the effect of various grid representations in capturing the nuances of the forest. This sensitivity analysis reveals that larger voxel sizes (e.g., 2 meters) result in lower errors due to reduced variability, while smaller voxel sizes (e.g., 0.25 or 0.5 meter) exhibit higher errors, particularly within the canopy, where variability is greatest. For bark and leaf targets, error values at smaller voxel size datasets (0.25 and 0.5 meter) were significantly higher than those in larger voxel size datasets (2 meters), highlighting the difficulty in accurately estimating within-canopy voxel content at fine resolutions. This suggests that the choice of voxel size is application-dependent. Our work fills the gap in deep imbalance learning models for multi-target regression and simulated datasets for 3D LiDAR point clouds of forests.

Forests and the ecosystem services they provide are increasingly threatened by exotic insect pests. Among these, hemlock woolly adelgid (HWA; Adelges tsugae) causes the decline and mortality of hemlocks (Tsuga spp.) in eastern North America, threatening to extirpate hemlock throughout much of its range. Effects of HWA on forest structure vary spatially and temporally and the resulting consequences for net primary production (NPP) are poorly understood. Here, we investigated the effects of HWA on canopy structural complexity (CSC), wood NPP, and their relationship in a temperate mixed forest. We hypothesized that CSC would be greater (H1) and wood production lower (H2) in more severely disturbed plots, and that effects of CSC would mediate disturbanceproduction relationships (H3). We found that higher levels of HWA infestation were broadly - but not unanimously - associated with greater structural complexity across multiple canopy strata (H1), as quantified by novel unmanned aerial system (UAS)-lidar metrics. In contrast to our expectations, tree mortality during the two-year measurement window was relatively low and did not directly reduce wood production (H2). Rather, wood production was independently coupled with mid- and whole-canopy CSC (R2 = 0.96), driven in part by HWA. However, we were unable to conclude from our analysis that significant relationships between CSC and wood NPP were mediated by HWA's effect on canopy structure (H3). Results overall suggested that HWA may generate structural complexity in the near-term through tree damage and mortality, but that the relationship between CSC and NPP may not be, or is only weakly, mediated by HWA, at least when mortality rates are low. Furthermore, the high degree of variance in wood NPP explained by CSC metrics derived from airborne lidar suggests they have potential utility for modeling and monitoring NPP in temperate forests.

Accurate and scalable estimation of forest production is essential for quantifying carbon sequestration, forecasting timber yields, and guiding climate change mitigation strategies. While prior studies established a positive linkage between net primary production (NPP) and canopy structural complexity (CSC) metrics derived from terrestrial LiDAR, the spatial coverage of ground-based surveys is limited. Airborne laser scanning (ALS) could offer a rapid and spatially extensive alternative to terrestrial scanning, but the predictive capacity of ALS-derived CSC metrics for estimating forest production remains insufficiently explored. To address this gap, we derived a suite of three-dimensional (3D) CSC metrics from small-footprint, high-density ALS data collected by the National Ecological Observatory Network’s Airborne Observation Platform. We evaluated relationships between CSC metrics and the NPP of plots nested within seven deciduous and evergreen temperate forests. Optimal metric combinations for predicting NPP within and across forest types were identified using partial least squares regression coupled with recursive feature elimination. ALS-derived CSC metrics explained 77% (RMSE = 11%) and 76% (RMSE = 13%) of the variance in deciduous and evergreen forest plot NPP, respectively. Our findings demonstrate that 3D CSC metrics derived from high-density ALS are robust predictors of plot-level NPP, offering performance comparable to terrestrial scanners while enabling greater scalability and more efficient data acquisition.

The measurement of forest structure has evolved steadily due to advances in technology, methodology, and theory. Such advances have greatly increased our capacity to describe key forest structural elements and resulted in a range of measurement approaches from traditional analog tools such as measurement tapes to highly derived and computationally intensive methods such as advanced remote sensing tools (e.g., lidar, radar). This assortment of measurement approaches results in structural metrics unique to each method, with the caveat that metrics may be biased or constrained by the measurement approach taken. While forest structural diversity (FSD) metrics foster novel research opportunities, understanding how they are measured or derived, limitations of the measurement approach taken, as well as their biological interpretation is crucial for proper application. We review the measurement of forest structure and structural diversity-an umbrella term that includes quantification of the distribution of functional and biotic components of forests. We consider how and where these approaches can be used, the role of technology in measuring structure, how measurement impacts extend beyond research, and current limitations and potential opportunities for future research.

Recent expansion in data sharing has created unprecedented opportunities to explore structure-function linkages in ecosystems across spatial and temporal scales. However, characteristics of the same data product, such as resolution, can change over time or spatial locations, as protocols are adapted to new technology or conditions, which may impact the data’s potential utility and accuracy for addressing end user scientific questions. The National Ecological Observatory Network (NEON) provides data products for users from 81 sites and over a planned 30-year time frame, including discrete return Light Detection and Range (LiDAR) from an airborne observatory platform. LiDAR is a well-established and increasingly available remote sensing technology for measuring three-dimensional (3D) characteristics of ecosystem and landscape structure, including forest structural diversity. The LiDAR product that NEON provides can vary in point density from 2 – 25+ points/m2 depending on instrument and acquisition date. We used NEON LiDAR from five forested sites to (1) identify the minimum point density at which structural diversity metrics can be robustly estimated across forested sites from different ecoclimatic zones in the USA and (2) to test the effects of variable point density on the estimation of a suite of structural diversity metrics and multivariate structural complexity types within and across forested sites. Twelve out of sixteen structural diversity metrics were sensitive to LiDAR point density in at least one of the five NEON forested sites. The minimum point density to reliably estimate the metrics ranged from 2.0 to 7.5 pt/m2, but our results indicate that point densities above 7-8 pt/m2 should provide robust measurements of structural diversity in forests for temporal or spatial comparisons. The delineation of multivariate structural complexity types from a suite of 16 structural diversity metrics was robust within sites and across forest types for a LiDAR point density of 4 pt/m2 and above. This study shows that different metrics of structural diversity can vary in their sensitivity to the resolution of LiDAR data and users of these open-source data products should consider the point density of their data and use caution in metric selection when making spatial or temporal comparisons from these datasets.

Supporting information, datasets, and R and JavaScript code for the the National Ecological Observatory Network’s Airborne Observation Platform (AOP) sampling design and publication, "Spanning Scales: The Airborne Spatial and Temporal Sampling Design of the National Ecological Observatory Network"

Establishing linkages between light detection and ranging (lidar) data, produced from interrogating forest canopies, to the highly complex forest structures, composition, and traits that such forests contain, remains an extremely difficult problem. Radiative transfer models have been developed to help solve this problem and test new sensor platforms in a virtual environment. Many forest canopy studies include the major assumption of isotropic (Lambertian) reflecting and transmitting leaves or non-transmitting leaves. Here, we study when these assumptions may be valid and evaluate their associated impacts/effects on the lidar waveform, as well as its dependence on wavelength, lidar footprint, view angle, and leaf angle distribution (LAD), by using the Digital Imaging and Remote Sensing Image Generation (DIRSIG) remote sensing radiative transfer simulation model. The largest effects of Lambertian assumptions on the waveform are observed at visible wavelengths, small footprints, and oblique interrogation angles relative to the mean leaf angle. For example, a 77% increase in return signal was observed with a configuration of a 550 nm wavelength, 10 cm footprint, and 45° interrogation angle to planophile leaves. These effects are attributed to (i) the bidirectional scattering distribution function (BSDF) becoming almost purely specular in the visible, (ii) small footprints having fewer leaf angles to integrate over, and (iii) oblique angles causing diminished backscatter due to forward scattering. Non-transmitting leaf assumptions have the greatest error for large footprints at near-infrared (NIR) wavelengths. Regardless of leaf angle distribution, all simulations with non-transmitting leaves with a 5 m footprint and 1064 nm wavelength saw around a 15% reduction in return signal. We attribute the signal reduction to the increased multiscatter contribution for larger fields of view, and increased transmission at NIR wavelengths. Armed with the knowledge from this study, researchers will be able to select appropriate sensor configurations to account for or limit BSDF effects in forest lidar data.

The hemlock woolly adelgid (HWA; Adelges tsugae) is an invasive insect infestation that is spreading into the forests of the northeastern United States, driven by the warmer winter temperatures associated with climate change. The initial stages of this disturbance are difficult to detect with passive optical remote sensing, since the insect often causes its host species, eastern hemlock trees (Tsuga canadensis), to defoliate in the midstory and understory before showing impacts in the overstory. New active remote sensing technologies—such as the recently launched NASA Global Ecosystem Dynamics Investigation (GEDI) spaceborne lidar—can address this limitation by penetrating canopy gaps and recording lower canopy structural changes. This study explores new opportunities for monitoring the HWA infestation with airborne lidar scanning (ALS) and GEDI spaceborne lidar data. GEDI waveforms were simulated using airborne lidar datasets from an HWA-infested forest plot at the Harvard Forest ForestGEO site in central Massachusetts. Two airborne lidar instruments, the NASA G-LiHT and the NEON AOP, overflew the site in 2012 and 2016. GEDI waveforms were simulated from each airborne lidar dataset, and the change in waveform metrics from 2012 to 2016 was compared to field-derived hemlock mortality at the ForestGEO site. Hemlock plots were shown to be undergoing dynamic changes as a result of the HWA infestation, losing substantial plant area in the middle canopy, while still growing in the upper canopy. Changes in midstory plant area (PAI 11–12 m above ground) and overall canopy permeability (indicated by RH10) accounted for 60% of the variation in hemlock mortality in a logistic regression model. The robustness of these structure-condition relationships held even when simulated waveforms were treated as real GEDI data with added noise and sparse spatial coverage. These results show promise for future disturbance monitoring studies with ALS and GEDI lidar data.

Full waveform (FW) LiDAR holds great potential for retrieving vegetation structure parameters at a high level of detail, but this prospect is constrained by practical factors such as the lack of available handy processing tools and the technical intricacy of waveform processing. This study introduces a new product named the Hyper Point Cloud (HPC), derived from FW LiDAR data, and explores its potential applications, such as tree crown delineation using the HPC-based intensity and percentile height (PH) surfaces, which shows promise as a solution to the constraints of using FW LiDAR data. The results of the HPC present a new direction for handling FW LiDAR data and offer prospects for studying the mid-story and understory of vegetation with high point density (~182 points/m2). The intensity-derived digital surface model (DSM) generated from the HPC shows that the ground region has higher maximum intensity (MAXI) and mean intensity (MI) than the vegetation region, while having lower total intensity (TI) and number of intensities (NI) at a given grid cell. Our analysis of intensity distribution contours at the individual tree level exhibit similar patterns, indicating that the MAXI and MI decrease from the tree crown center to the tree boundary, while a rising trend is observed for TI and NI. These intensity variable contours provide a theoretical justification for using HPC-based intensity surfaces to segment tree crowns and exploit their potential for extracting tree attributes. The HPC-based intensity surfaces and the HPC-based PH Canopy Height Models (CHM) demonstrate promising tree segmentation results comparable to the LiDAR-derived CHM for estimating tree attributes such as tree locations, crown widths and tree heights. We envision that products such as the HPC and the HPC-based intensity and height surfaces introduced in this study can open new perspectives for the use of FW LiDAR data and alleviate the technical barrier of exploring FW LiDAR data for detailed vegetation structure characterization.

Waveform Light Detection and Ranging (LiDAR) data have advantages over discrete-return LiDAR data in accurately characterizing vegetation structure. However, we lack a comprehensive understanding of waveform data processing approaches under different topography and vegetation conditions. The objective of this paper is to highlight a novel deconvolution algorithm, the Gold algorithm, for processing waveform LiDAR data with optimal deconvolution parameters. Further, we present a comparative study of waveform processing methods to provide insight into selecting an approach for a given combination of vegetation and terrain characteristics. We employed two waveform processing methods: (1) direct decomposition, (2) deconvolution and decomposition. In method two, we utilized two deconvolution algorithms – the Richardson-Lucy (RL) algorithm and the Gold algorithm. The comprehensive and quantitative comparisons were conducted in terms of the number of detected echoes, position accuracy, the bias of the end products (such as digital terrain model (DTM) and canopy height model (CHM)) from the corresponding reference data, along with parameter uncertainty for these end products obtained from different methods. This study was conducted at three study sites that include diverse ecological regions, vegetation and elevation gradients. Results demonstrate that two deconvolution algorithms are sensitive to the pre-processing steps of input data. The deconvolution and decomposition method is more capable of detecting hidden echoes with a lower false echo detection rate, especially for the Gold algorithm. Compared to the reference data, all approaches generate satisfactory accuracy assessment results with small mean spatial difference (<1.22m for DTMs, <0.77m for CHMs) and root mean square error (RMSE) (<1.26m for DTMs, <1.93m for CHMs). More specifically, the Gold algorithm is superior to others with smaller root mean square error (RMSE) (<1.01m), while the direct decomposition approach works better in terms of the percentage of spatial difference within 0.5 and 1m. The parameter uncertainty analysis demonstrates that the Gold algorithm outperforms other approaches in dense vegetation areas, with the smallest RMSE, and the RL algorithm performs better in sparse vegetation areas in terms of RMSE. Additionally, the high level of uncertainty occurs more on areas with high slope and high vegetation. This study provides an alternative and innovative approach for waveform processing that will benefit high fidelity processing of waveform LiDAR data to characterize vegetation structures.

Monitoring forest productivity and health is key to sustainable ecosystem management and informed decision making. A key parameter used in monitoring forest resources is the leaf area index (LAI), which is defined as the one-sided leaf area per unit ground area and is used to describe the canopy radiation regime, among other forest biophysical dynamics. Traditional optics-based methods to estimate LAI rely on the measurement of canopy transmission and foliage clumping. Extending optical methods to LiDAR data has been challenging and studies have reported effective LAI assessments, with no further quantification of foliage clumping. This study investigates the use of the box-counting method to assess the fractal dimension of point cloud data for contrasting forest types and along a gradient of foliage dispersal. We demonstrate the box-counting method on simulated 'range-to-hit', as well as acquired airborne discrete LiDAR data. Coherent results obtained from the different test cases hint at the potential of the box-counting fractal dimension to characterize foliage clumping and bode well for the use of clumping assessments in support of airborne, wall-to-wall estimates of LAI.

In a typical waveform light detection and ranging (lidar) system, the received pulse can be represented by the convolution of the system impulse response, the outgoing pulse, and the underlying signal representing actual target interactions. Deconvolution is the process of removing the contribution of the system impulse response and outgoing pulse from the received signal, so that the true interactions may be seen. In many examples, deconvolution has been shown to expose fine structure within the waveform, which may be used to improve accuracy when estimating the vertical location of certain features. For instance, the exact location of the ground may be more accurately determined by separating the response of the ground from that of understory vegetation or vegetative ground cover. However, in order for the deconvolution to be successful, the impulse response and outgoing pulse must be known, and many deconvolution methods are sensitive to small errors in the estimation of these inputs. In this study, we propose a deconvolution method that uses a flat target response in place of the impulse response and outgoing pulse.