LiDAR-derived Above-ground Biomass Research Articles

Active Remote Sensing Assessment of Biomass Productivity and Canopy Structure of Short-Rotation Coppice American Sycamore (Platanus occidentalis L.)

The short-rotation coppice (SRC) culture of trees provides a sustainable form of renewable biomass energy, while simultaneously sequestering carbon and contributing to the regional carbon feedstock balance. To understand the role of SRC in carbon feedstock balances, field inventories with selective destructive tree sampling are commonly used to estimate aboveground biomass (AGB) and canopy structure dynamics. However, these methods are resource intensive and spatially limited. To address these constraints, we examined the utility of publicly available airborne Light Detection and Ranging (LiDAR) data and easily accessible imagery from Unmanned Aerial Systems (UASs) to estimate the AGB and canopy structure of an American sycamore SRC in the piedmont region of North Carolina, USA. We compared LiDAR-derived AGB estimates to field estimates from 2015, and UAS-derived AGB estimates to field estimates from 2022 across four planting densities (10,000, 5000, 2500, and 1250 trees per hectare (tph)). The results showed significant effects of planting density treatments on LIDAR- and UAS-derived canopy metrics and significant relationships between these canopy metrics and AGB. In the 10,000 tph, the field-estimated AGB in 2015 (7.00 ± 1.56 Mg ha−1) and LiDAR-derived AGB (7.19 ± 0.13 Mg ha−1) were comparable. On the other hand, the UAS-derived AGB was overestimated in the 10,000 tph planting density and underestimated in the 1250 tph compared to the 2022 field-estimated AGB. This study demonstrates that the remote sensing-derived estimates are within an acceptable level of error for biomass estimation when compared to precise field estimates, thereby showing the potential for increasing the use of accessible remote-sensing technology to estimate AGB of SRC plantations.

Mapping historical forest biomass for stock-change assessments at parcel to landscape scales

Understanding historical forest dynamics, specifically changes in forest biomass and carbon stocks, has become critical for assessing current forest climate benefits and projecting future benefits under various policy, regulatory, and stewardship scenarios. Carbon accounting frameworks based exclusively on national forest inventories are limited to broad-scale estimates, but model-based approaches that combine these inventories with remotely sensed data can yield contiguous fine-resolution maps of forest biomass and carbon stocks across landscapes over time. Here we describe a fundamental step in building a map-based stock-change framework: mapping historical forest biomass at fine temporal and spatial resolution (annual, 30 m) across all of New York State (USA) from 1990 to 2019, using freely available data and open-source tools. Using Landsat imagery, US Forest Service Forest Inventory and Analysis (FIA) data, and off-the-shelf LiDAR collections we developed three modeling approaches for mapping historical forest aboveground biomass (AGB): training on FIA plot-level AGB estimates (direct), training on LiDAR-derived AGB maps (indirect), and an ensemble averaging predictions from the direct and indirect models. Model prediction surfaces (maps) were tested against FIA estimates at multiple scales. All three approaches produced viable outputs, yet tradeoffs were evident in terms of model complexity, map accuracy, saturation, and fine-scale pattern representation. The resulting map products can help identify where, when, and how forest carbon stocks are changing as a result of both anthropogenic and natural drivers alike. These products can thus serve as inputs to a wide range of applications including stock-change assessments, monitoring reporting and verification frameworks, and prioritizing parcels for protection or enrollment in improved management programs.

Decision tree-based machine learning models for above-ground biomass estimation using multi-source remote sensing data and object-based image analysis

Forest above-ground biomass (AGB) estimation provides valuable information about the carbon cycle. Thus, the overall goal of this paper is to present an approach to enhance the accuracy of the AGB estimation. The main objectives are to: 1) investigate the performance of remote sensing data sources, including airborne light detection and ranging (LiDAR), optical, SAR, and their combination to improve the AGB predictions, 2) examine the capability of tree-based machine learning models, and 3) compare the performance of pixel-based and object-based image analysis (OBIA). To investigate the performance of machine learning models, multiple tree-based algorithms were fitted to predictors derived from airborne LiDAR data, Landsat, Sentinel-2, Sentinel-1, and PALSAR-2/PALSAR SAR data collected within New York’s Adirondack Park. Combining remote sensing data from multiple sources improved the model accuracy (RMSE: 52.14 Mg ha−1 and R2: 0.49). There was no significant difference among gradient boosting machine (GBM), random forest (RF), and extreme gradient boosting (XGBoost) models. In addition, pixel-based and object-based models were compared using the airborne LiDAR-derived AGB raster as a training/testing sample. The OBIA provided the best results with the RMSE of 33.77 Mg ha−1 and R2 of 0.81 for the combination of optical and SAR data in the GBM model.

Estimating above-ground biomass of subtropical forest using airborne LiDAR in Hong Kong

Seventy-percent of the terrestrial area of Hong Kong is covered by vegetation and 40% is protected as the Country Park. The above-ground biomass (AGB) acts as reliable source of carbon sink and while Hong Kong has recognized the importance of carbon sink in forest and urged for forest protection in the latest strategic plan, yet no study has been conducted on assessing the baseline of terrestrial AGB and its carbon storage. This study compared and estimated the AGB by the traditional allometric modeling and the Light Detection and Ranging (LiDAR) plot metrics at plot-level in a subtropical forest of Hong Kong. The study has tested five allometric models which were developed from pantropical regions, subtropical areas and locally. The best model was then selected as the dependent variable to develop the LiDAR-derived AGB model. The raw LiDAR point cloud was pre-processed to normalized height point cloud and hence generating the LiDAR metric as independent variables for the model development. Regression models were used to estimate AGB at various plot sizes (i.e., in 10-m, 5-m and 2.5-m radius). The models were then evaluated statistically and validated by bootstrapping and leave-one-out cross validation (LOOCV). The results indicated the LiDAR metric derived from larger plot size outperformed the smaller plot size, with model R2 of 0.864 and root-mean-square-error (RMSE) of 37.75 kg/ha. It also found that pantropical model was comparable to a site-specific model when including the bioclimatic variable in subtropical forests. This study provides the approach for delineating the baseline of terrestrial above-ground biomass and carbon stock in subtropical forests upon an appropriate plot size is being deployed.

Estimating urban above ground biomass with multi-scale LiDAR

BackgroundUrban trees have long been valued for providing ecosystem services (mitigation of the “heat island” effect, suppression of air pollution, etc.); more recently the potential of urban forests to store significant above ground biomass (AGB) has also be recognised. However, urban areas pose particular challenges when assessing AGB due to plasticity of tree form, high species diversity as well as heterogeneous and complex land cover. Remote sensing, in particular light detection and ranging (LiDAR), provide a unique opportunity to assess urban AGB by directly measuring tree structure. In this study, terrestrial LiDAR measurements were used to derive new allometry for the London Borough of Camden, that incorporates the wide range of tree structures typical of an urban setting. Using a wall-to-wall airborne LiDAR dataset, individual trees were then identified across the Borough with a new individual tree detection (ITD) method. The new allometry was subsequently applied to the identified trees, generating a Borough-wide estimate of AGB.ResultsCamden has an estimated median AGB density of 51.6 Mg ha–1 where maximum AGB density is found in pockets of woodland; terrestrial LiDAR-derived AGB estimates suggest these areas are comparable to temperate and tropical forest. Multiple linear regression of terrestrial LiDAR-derived maximum height and projected crown area explained 93% of variance in tree volume, highlighting the utility of these metrics to characterise diverse tree structure. Locally derived allometry provided accurate estimates of tree volume whereas a Borough-wide allometry tended to overestimate AGB in woodland areas. The new ITD method successfully identified individual trees; however, AGB was underestimated by ≤ 25% when compared to terrestrial LiDAR, owing to the inability of ITD to resolve crown overlap. A Monte Carlo uncertainty analysis identified assigning wood density values as the largest source of uncertainty when estimating AGB.ConclusionOver the coming century global populations are predicted to become increasingly urbanised, leading to an unprecedented expansion of urban land cover. Urban areas will become more important as carbon sinks and effective tools to assess carbon densities in these areas are therefore required. Using multi-scale LiDAR presents an opportunity to achieve this, providing a spatially explicit map of urban forest structure and AGB.

Development of a Regional Lidar-Derived Above-Ground Biomass Model with Bayesian Model Averaging for Use in Ponderosa Pine and Mixed Conifer Forests in Arizona and New Mexico, USA

Historical forest management practices in the southwestern US have left forests prone to high-severity, stand-replacement fires. Reducing the cost of forest-fire management and reintroducing fire to the landscape without negative impact depends on detailed knowledge of stand composition, in particular, above-ground biomass (AGB). Lidar-based modeling techniques provide opportunities to increase ability of managers to monitor AGB and other forest metrics at reduced cost. We developed a regional lidar-based statistical model to estimate AGB for Ponderosa pine and mixed conifer forest systems of the southwestern USA, using previously collected field data. Model selection was performed using Bayesian model averaging (BMA) to reduce researcher bias, fully explore the model space, and avoid overfitting. The selected model includes measures of canopy height, canopy density, and height distribution. The model selected with BMA explains 71% of the variability in field-estimates of AGB, and the RMSE of the two independent validation data sets are 23.25 and 32.82 Mg/ha. The regional model is structured in accordance with previously described local models, and performs equivalently to these smaller scale models. We have demonstrated the effectiveness of lidar for developing cost-effective, robust regional AGB models for monitoring and planning adaptively at the landscape scale.

Estimation of forest aboveground biomass and uncertainties by integration of field measurements, airborne LiDAR, and SAR and optical satellite data in Mexico

BackgroundInformation on the spatial distribution of aboveground biomass (AGB) over large areas is needed for understanding and managing processes involved in the carbon cycle and supporting international policies for climate change mitigation and adaption. Furthermore, these products provide important baseline data for the development of sustainable management strategies to local stakeholders. The use of remote sensing data can provide spatially explicit information of AGB from local to global scales. In this study, we mapped national Mexican forest AGB using satellite remote sensing data and a machine learning approach. We modelled AGB using two scenarios: (1) extensive national forest inventory (NFI), and (2) airborne Light Detection and Ranging (LiDAR) as reference data. Finally, we propagated uncertainties from field measurements to LiDAR-derived AGB and to the national wall-to-wall forest AGB map.ResultsThe estimated AGB maps (NFI- and LiDAR-calibrated) showed similar goodness-of-fit statistics (R2, Root Mean Square Error (RMSE)) at three different scales compared to the independent validation data set. We observed different spatial patterns of AGB in tropical dense forests, where no or limited number of NFI data were available, with higher AGB values in the LiDAR-calibrated map. We estimated much higher uncertainties in the AGB maps based on two-stage up-scaling method (i.e., from field measurements to LiDAR and from LiDAR-based estimates to satellite imagery) compared to the traditional field to satellite up-scaling. By removing LiDAR-based AGB pixels with high uncertainties, it was possible to estimate national forest AGB with similar uncertainties as calibrated with NFI data only.ConclusionsSince LiDAR data can be acquired much faster and for much larger areas compared to field inventory data, LiDAR is attractive for repetitive large scale AGB mapping. In this study, we showed that two-stage up-scaling methods for AGB estimation over large areas need to be analyzed and validated with great care. The uncertainties in the LiDAR-estimated AGB propagate further in the wall-to-wall map and can be up to 150%. Thus, when a two-stage up-scaling method is applied, it is crucial to characterize the uncertainties at all stages in order to generate robust results. Considering the findings mentioned above LiDAR can be used as an extension to NFI for example for areas that are difficult or not possible to access.

Stacked Sparse Autoencoder Modeling Using the Synergy of Airborne LiDAR and Satellite Optical and SAR Data to Map Forest Above-Ground Biomass

Timely, spatially complete, and reliable forest above-ground biomass (AGB) data are a prerequisite to support forest management and policy formulation. Traditionally, forest AGB is spatially estimated by integrating satellite images, in particular, optical data, with field plots from forest inventory programs. However, field data are limited in remote and unmanaged areas. In addition, optical reflectance usually saturates at high-density biomass level and is subject to cloud contaminations. Thus, this study aimed to develop a deep learning based workflow for mapping forest AGB by integrating Landsat 8 and Sentinel-1A images with airborne light detection and ranging (LiDAR) data. A reference AGB map was derived from the wall-to-wall LiDAR data and field measurements. The LiDAR plots-stratified random samples of forest biomass extracted from the LiDAR simulated strips in the reference map-were adopted as a surrogate for traditional field plots. In addition to the deep learning model, i.e., stacked sparse autoencoder network (SSAE), five different prediction techniques including multiple stepwise linear regressions, K-nearest neighbor, support vector machine, back propagation neural networks, and random forest were individually used to establish the relationship between LiDAR-derived forest biomass and the satellite predictors. Optical variables (Landsat 8 OLI), SAR variables (Sentinel-1A), and their combined variables were individually input to the six prediction models. Results showed that the SSAE model had the best performance for the forest biomass estimation. The combined optical and microwave dataset as explanatory variables improved the modeling performance compared to either the optical-only or microwave-only data, regardless of prediction algorithms. The best mapping accuracy was obtained by the SSAE model with inputs of optical and microwave integrated metrics that yielded R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of 0.812, root mean squared error (RMSE) of 21.753 Mg/ha, and relative RMSE (RMSEr) of 14.457%. Overall, the SSAE model with inputs of combined Landsat 8 OLI and Sentinel-1A information could result in accurate estimation of forest biomass by using the stratification-sampled and LiDAR-derived AGB as ground reference data. The modeling workflow has the potential to promote future forest growth monitoring and carbon stock assessment across large areas.

Lidar detection of individual tree size in tropical forests

Abstract Characterization of tropical forest trees has been limited to field-based techniques focused on measurement of diameter of the cylindrical part of the bole, with large uncertainty in measuring large trees with irregular shapes, and other size attributes such as total tree height and the crown size. Here, we introduce a methodology to decompose lidar point cloud data into 3D clusters corresponding to individual tree crowns (ITC) that enables the estimation of many biophysical variables of tropical forests such as tree height, crown area, crown volume, and tree number density. The ITC-based approach was tested using airborne high-resolution lidar data collected over the 50-ha Center for Tropical Forest Science (CTFS) plot in the Barro Colorado Island, Panama. The lack of tree height and crown size measurements in the field prohibits the direct validation of the ITC metrics. We assess the reliability of our method by comparing the aboveground biomass (AGB) estimated using ground and lidar individual tree measurements at multiple spatial scales, namely 1 ha, 2.25 ha, 4 ha, and 6.25 ha. We examined four different lidar-derived AGB models, with three based on individual tree height, crown volume, and crown area, and one with mean top canopy height (TCH) calculated at the plot level using the lidar canopy height model. Results show that the predictive power of all models based on ITC size and TCH increases with decreasing spatial resolution from 16.9% at 1 ha for the worst model to 5.0% at 6.25 ha for the best model. The TCH-based model performed slightly better than ITC-based models except at higher spatial scales (~ 4 ha) and when errors due to edge effects associated with tree crowns were reduced. Unlike the TCH models that change regionally depending on forest type and structure allometry, the ITC-based models are derived as a function of individual tree allometry and can be extended globally to all tropical forests. The method for lidar detection of individual crown size overcome some limitations of ground-based inventories such as 1) it is able to access crowns of large trees and 2) it enables the assessment of directional changes in tree density, canopy architecture and forest dynamics over large and inaccessible areas to support robust tropical ecological studies.

Evaluating Lidar Point Densities for Effective Estimation of Aboveground Biomass

The U.S. Geological Survey (USGS) 3D Elevation Program (3DEP) was recently established to provide airborne lidar data coverage on a national scale. As part of a broader research effort of the USGS to develop an effective remote sensing-based methodology for the creation of an operational biomass Essential Climate Variable (Biomass ECV) data product, we evaluated the performance of airborne lidar data at various pulse densities against Landsat 8 satellite imagery in estimating above ground biomass for forests and woodlands in a study area in east-central Arizona, U.S. High point density airborne lidar data, were randomly sampled to produce five lidar datasets with reduced densities ranging from 0.5 to 8 point(s)/m 2 , corresponding to the point density range of 3DEP to provide national lidar coverage over time. Lidar-derived aboveground biomass estimate errors showed an overall decreasing trend as lidar point density increased from 0.5 to 8 points/m 2 . Landsat 8-based aboveground biomass estimates produced errors larger than the lowest lidar point density of 0.5 point/m 2 , and therefore Landsat 8 observations alone were ineffective relative to airborne lidar for generating a Biomass ECV product, at least for the forest and woodland vegetation types of the Southwestern U.S. While a national Biomass ECV product with optimal accuracy could potentially be achieved with 3DEP data at 8 points/m 2 , our results indicate that even lower density lidar data could be sufficient to provide a national Biomass ECV product with accuracies significantly higher than that from Landsat observations alone.

Integrating airborne LiDAR and space-borne radar via multivariate kriging to estimate above-ground biomass

Understanding and investigating synergies between LiDAR (light detection and ranging) and SAR (synthetic aperture radar) provide new and innovative opportunities to characterize above-ground biomass. We demonstrate a spatial modeling framework that integrates above-ground biomass transects, derived from plot-based field data and small-footprint discrete return LiDAR, with complete wall-to-wall spaceborne L-band and C-band SAR to predict biomass over a larger area. Transect intervals of 2000m, 1000m, and 500m were tested. Co-kriging, regression kriging, and regression co-kriging were used to extend the LiDAR-derived biomass transects. LiDAR-derived above-ground biomass and L-band backscatter (HV polarization) were moderately correlated, with a maximum semivariance distance between the LiDAR-derived biomass and SAR data of 374m. Regression kriging at a sample interval of 500m showed the smallest root mean squared error (RMSE) and mean absolute error (MAE) at 203.9Mgha−1 and 131.6Mgha−1, respectively. The mean error (ME) showed an average bias of −14.0Mgha−1. Predictions using regression co-kriging at a sample interval of 2000m resulted in the highest RMSE and MAE values at 238.2Mgha−1 and 164.6Mgha−1, respectively. ME also was highest, averaging −37.4Mgha−1. Regardless of the spatial modeling technique employed, lower errors in predicted above-ground biomass were associated with smaller transect intervals. Moderate correlations between the LiDAR-derived above-ground biomass and the radar data impacted the predictive accuracy of the spatial models; however, overall variation in above-ground biomass in the study area was well represented. This study demonstrated that a sampling framework integrating LiDAR data with space-borne radar data using a spatial modeling approach can provide spatially-explicit above-ground biomass estimates for large areas. Such a sampling framework can be used in combination with ground plot and land cover data to assess carbon stocks under conditions where more common optical remote sensing approaches are difficult to implement.

Estimating stem volume and biomass of Pinus koraiensis using LiDAR data

The objective of this study was to estimate the stem volume and biomass of individual trees using the crown geometric volume (CGV), which was extracted from small-footprint light detection and ranging (LiDAR) data. Attempts were made to analyze the stem volume and biomass of Korean Pine stands (Pinus koraiensis Sieb. et Zucc.) for three classes of tree density: low (240 N/ha), medium (370 N/ha), and high (1,340 N/ha). To delineate individual trees, extended maxima transformation and watershed segmentation of image processing methods were applied, as in one of our previous studies. As the next step, the crown base height (CBH) of individual trees has to be determined; information for this was found in the LiDAR point cloud data using k-means clustering. The LiDAR-derived CGV and stem volume can be estimated on the basis of the proportional relationship between the CGV and stem volume. As a result, low tree-density plots had the best performance for LiDAR-derived CBH, CGV, and stem volume (R (2) = 0.67, 0.57, and 0.68, respectively) and accuracy was lowest for high tree-density plots (R (2) = 0.48, 0.36, and 0.44, respectively). In the case of medium tree-density plots accuracy was R (2) = 0.51, 0.52, and 0.62, respectively. The LiDAR-derived stem biomass can be predicted from the stem volume using the wood basic density of coniferous trees (0.48 g/cm(3)), and the LiDAR-derived above-ground biomass can then be estimated from the stem volume using the biomass conversion and expansion factors (BCEF, 1.29) proposed by the Korea Forest Research Institute (KFRI).