Table 2 - uploaded by Joanne C. White
Content may be subject to copyright.
Source publication
The ability to use digital remotely sensed data for forest inventory is often limited by the nature of the measures, which, with the exception of multi-angular or stereo observations, are largely insensitive to vertically distributed attributes. As a result, empirical estimates are typically made to characterize attributes such as height, volume, o...
Contexts in source publication
Context 1
... a sampling and transect collection perspective, Table 1 also shows the in-built limitation in swath width as a function of scan angle. The instruments presented in Table 1, while commonly used, have been superseded by newer instruments and models (Table 2). In Table 2 instruments from different companies representing the model years 2010/11 show the impact of increased scan angles. ...
Context 2
... instruments presented in Table 1, while commonly used, have been superseded by newer instruments and models (Table 2). In Table 2 instruments from different companies representing the model years 2010/11 show the impact of increased scan angles. For instance, for the Optech and Leica instruments presented, the in- crease in scan angles with the newer generation instruments has ef- fectively doubled the swath width possible from a similar flying height. ...
Context 3
... infor- mation from each of these technologies using accurate time referen- cing, the absolute position of a reflecting surface can be solved ( Lefsky et al., 2002). Indeed, it is the parallel advances of these tech- nologies that have provided the impetus for the increasing number of applications for lidar technology ( Lim et al., 2003). Table 1 Typical parameterization of a lidar survey for forest applications. Platform Fixed wing aircraft Fixed wing aircraft Helicopter Sensor Optech ALTM 3100C Leica ALS50-II Riegl LMS-Q140i-60 Sensor model year 2004 2004 Not reported Maximum number of returns per emitted ...
Similar publications
Forest inventory data are often summarized in tabular format where rows and columns represent levels of categorical variables. It is generally desirable to constrain tables based on different categories to sum to common totals. It may also be desirable to develop a system that incorporates stratification. Weights can be assigned to inventory plots...
Citations
... Forest ecosystems are often characterized in terms of structure, composition, and functions [1]. Light Detection and Ranging (LiDAR) remote sensing (RS) has substantially improved our understanding of forest structure around the world in recent decades [2][3][4][5]. LiDAR instruments provide explicit three-dimensional (3D) data that have enabled measurements of forest structure parameters such as canopy height, leaf area index, and diameter at breast height across different scales with unprecedented accuracy [6][7][8]. ...
Purpose of the Review
Many LiDAR remote sensing studies over the past decade promised data fusion as a potential avenue to increase accuracy, spatial-temporal resolution, and information extraction in the final data products. Here, we performed a structured literature review to analyze relevant studies on these topics published in the last decade and the main motivations and applications for fusion, and the methods used. We discuss the findings with a panel of experts and report important lessons, main challenges, and future directions.
Recent Findings
LiDAR fusion with other datasets, including multispectral, hyperspectral, and radar, is found to be useful for a variety of applications in the literature, both at individual tree level and at area level, for tree/crown segmentation, aboveground biomass assessments, canopy height, tree species identification, structural parameters, and fuel load assessments etc. In most cases, gains are achieved in improving the accuracy (e.g. better tree species classifications), and spatial-temporal resolution (e.g. for canopy height). However, questions remain regarding whether the marginal improvements reported in a range of studies are worth the extra investment, specifically from an operational point of view. We also provide a clear definition of “data fusion” to inform the scientific community on data fusion, combination, and integration.
Summary
This review provides a positive outlook for LiDAR fusion applications in the decade to come, while raising questions about the trade-off between benefits versus the time and effort needed for collecting and combining multiple datasets.
... These strategies have demonstrated tremendous potential to reduce field plot installation costs and improve wall-to-wall AGB estimate accuracy, which could provide solutions for forest data collection in forest inventory-plagued regions such as the Miombo ecoregion. A study by Wulder et al. [31] presented a complete review of employing lidar sampling to allow large-area forest characterizations, in which lidar samples were utilized in a way comparable to field samples. However, their review focused on airborne, which are still expensive to acquire in the Miombo region. ...
To date, only a limited number of studies have utilized remote sensing imagery to estimate aboveground biomass (AGB) in the Miombo ecoregion using wall-to-wall medium resolution optical satellite imagery (Sentinel-2 and Landsat), localized airborne light detection and ranging (lidar), or localized unmanned aerial systems (UAS) images. On the one hand, the optical satellite imagery is suitable for wall-to-wall coverage, but the AGB estimates based on such imagery lack precision for local or stand-level sustainable forest management and international reporting mechanisms. On the other hand, the AGB estimates based on airborne lidar and UAS imagery have the precision required for sustainable forest management at a local level and international reporting requirements but lack capacity for wall-to-wall coverage. Therefore, the main aim of this study was to investigate the use of UAS-lidar as a sampling tool for satellite-based AGB estimation in the Miombo woodlands of Zambia. In order to bridge the spatial data gap, this study employed a two-phase sampling approach, utilizing Sentinel-2 imagery, partial-coverage UAS-lidar data, and field plot data to estimate AGB in the 8094-hectare Miengwe Forest, Miombo Woodlands, Zambia, where UAS-lidar estimated AGB was used as reference data for estimating AGB using Sentinel-2 image metrics. The findings showed that utilizing UAS-lidar as reference data for predicting AGB using Sentinel-2 image metrics yielded superior results (Adj-R² = 0.70, RMSE = 27.97) than using direct field estimated AGB and Sentinel-2 image metrics (R² = 0.55, RMSE = 38.10). The quality of AGB estimates obtained from this approach, coupled with the ongoing advancement and cost-cutting of UAS-lidar technology as well as the continuous availability of wall-to-wall optical imagery such as Sentinel-2, provides much-needed direction for future forest structural attribute estimation for efficient management of the Miombo woodlands.
... Los métodos manuales para medir la relación entre el matorral y los elementos del dosel son complejos e imperfectos. En contrapartida, los métodos indirectos de medición de la estructura vertical de la vegetación basados en LiDAR (Light Detection and Ranging) permiten estimar variables como el índice de área foliar (LAI, Leaf Area Index), el índice de área vegetal o la cubierta del dosel, así como caracterizar la estructura vertical o la vegetación del sotobosque (Lefsky et al., 2002;Wulder et al., 2012;Crespo-Peremarch et al., 2018). ...
The potential to quantify and monitor understory dynamics, especially the connectivity between the understory and the overstory strata, plays a key role in the study of the spread of wildfires. The connectivity between forest strata is challenging to measure in Mediterranean forests due to the complexity and resources involved. The present study proposes a method to identify the understory and overstory strata and to quantify changes in connectivity between the strata from terrestrial laser scanner (TLS) data. The method consists of dividing TLS point clouds into columns of data, extracting height return distribution profiles and applying Gaussian decomposition to analyze the evolution of connectivity between strata in the period from 2015 to 2022. Results show that the Pinus pinaster has a lower connectivity with the understory compared to the P. halepensis. In addition, it is shown the effect of self-pruning and, consequently, the reduction of fuel material in the stems over time. In general terms, the shrub cover and its height decreased, reducing connectivity between strata over time. Results observed in the overstory stratum are probably a consequence of the effect of tree growth without silvicultural treatment, and the negative variation in the understory is probably related to the climatology of the period analyzed. This approach allows us to deepen our understanding of the vertical dynamics of the Mediterranean forest, effectively discern between understory and overstory strata, and quantify the change in structure, key elements in silvicultural management, and forest wildfire prevention.
... This technology allows for the direct estimation of 3D structural parameters such as height, stem density, and canopy cover, providing invaluable insights into forest characteristics and dynamics. Therefore, LiDAR suffers little from saturation effects (Wulder et al., 2012;Duncanson et al., 2022), although it cannot penetrate through clouds or haze. With the launch of GEDI, the high-resolution LiDAR instrument which is designed to measure vegetation structure , it is possible to quantify forest AGB in distributed 25m diameter footprints, a footprint size chosen to minimize slope effects compared to the previous spaceborne LiDAR the Ice Cloud and Land Elevation Satellite (ICESat) with its 65m footprints, over very large areas. ...
Large-scale high spatial resolution aboveground biomass (AGB) maps play a crucial role in determining forest carbon stocks and how they are changing, which is instrumental in understanding the global carbon cycle, and implementing policy to mitigate climate change. The advent of the new space-borne LiDAR sensor, NASA's GEDI instrument, provides unparalleled possibilities for the accurate and unbiased estimation of forest AGB at high resolution, particularly in dense and tall forests, where Synthetic Aperture Radar (SAR) and passive optical data exhibit saturation. However, GEDI is a sampling instrument, collecting dispersed footprints, and its data must be combined with that from other continuous cover satellites to create high-resolution maps, using local machine learning methods. In this study, we developed local models to estimate forest AGB from GEDI L2A data, as the models used to create GEDI L4 AGB data incorporated minimal field data from China. We then applied LightGBM and random forest regression to generate wall-to-wall AGB maps at 25 m resolution, using extensive GEDI footprints as well as Sentinel-1 data, ALOS-2 PALSAR-2 and Sentinel-2 optical data. Through a 5-fold cross-validation, LightGBM demonstrated a slightly better performance than Random Forest across two contrasting regions. However, in both regions, the computation speed of LightGBM is substantially faster than that of the random forest model, requiring roughly one-third of the time to compute on the same hardware. Through the validation against field data, the 25 m resolution AGB maps generated using the local models developed in this study exhibited higher accuracy compared to the GEDI L4B AGB data. We found in both regions an increase in error as slope increased. The trained models were tested on nearby but different regions and exhibited good performance.
... SWBs also foster unique and diverse ecosystems [6], providing extensive ecosystem services, such as food and water This is mainly attributed to the following limitations of remote sensing imagery: (1) Temporal resolution: Not all remote sensing satellites can provide the necessary temporal resolution to capture changes in lake area each month, and cloud cover and atmospheric conditions can limit the ability to obtain clear images [19]. (2) Cloud cover and atmospheric conditions: one of the main limitations of optical remote sensing is cloud obstruction, which complicates the accurate identification of lake boundaries [20]. (3) Image processing and data gaps: extracting water bodies from satellite imagery requires complex processing steps, and technical issues can lead to data gaps, affecting the creation of continuous monthly time series [21]. ...
The surface area changes of 151 natural lakes over 37 months in the Yellow River Basin, based on remote sensing data and 21 meteorological indicators, employing spatial distribution feature analysis, principal component analysis (PCA), correlation analysis, and multiple regression analysis, identify key meteorological factors influencing these variations and their interrelationships. During the study period, lake area averages were from 0.009 km2 to 506.497 km2, with standard deviations ranging from 0.003 km2 to 184.372 km2. The coefficient of variation spans from 3.043 to 217.436, indicating considerable variability in lake area stability. Six primary meteorological factors were determined to have a significant impact on lake surface area fluctuations: 24 h precipitation, maximum daily precipitation, hours of sunshine, maximum wind speed, minimum relative humidity, and lakes in the source region of the Yellow River generally showed a significant positive correlation. For maximum wind speed (m/s), 28 lakes showed significant correlations, with five positive and twenty-three negative correlations, correlation coefficients ranging from −0.34 to −0.63, average −0.47, indicating an overall negative correlation between lake surface area and maximum wind speed. For maximum daily precipitation (mm), 36 lakes had 21 showing a positive correlation, indicating a positive correlation between lake surface area and daily precipitation in larger lakes. Furthermore, of the 117 lakes with sufficient data to model, the predictive capabilities of various models for lake surface area changes showcased distinct advantages, with the random forest model outperforming others in a dataset of 65 lakes, Ridge regression is best for 28 lakes, Lasso regression performs best for 20 lakes, Linear model is only best for 4 cases. The random forest model provides the best fit due to its ability to handle a large number of feature variables and consider their interactions, thereby offering the best fitting effect. These insights are crucial for understanding the influence of meteorological factors on lake surface area changes within the Yellow River Basin and are instrumental in developing predictive models based on meteorological data.
... LiDAR sensor uses laser ranging, which divides the time between the emission of a pulse and the detection of a reflected return to calculate the distance between the sensor and the target (Baltsavias, 1999). Time-of-flight (TOF) measurements are used in LiDAR sensors to gather information on the distance (x, y, and z) and intensity (reflectivity) of various places in a scene (Golparvar-Fard et al., 2011), and this sensor can simultaneously measure both vertical and horizontal structures with precise details and measurements (Wulder et al., 2012). ...
Unmanned aerial vehicles (UAVs), building information modeling (BIM), and game engines are evolving technologies that are rapidly being adopted to enhance construction safety management and create safety training platforms. For construction safety improvements, utilizing game engines plays a crucial role where a 3D model is required. 3D models of a real construction site can be produced using UAV photogrammetry. High-quality UAV photogrammetry-derived 3D models have the potential to be integrated with game engines and support construction safety. However, the qualities of the 3D models from the photogrammetry techniques vary due to several factors, such as flight altitude, image overlapping percentages, and structure from motion (SfM) algorithms of post-processing tools. Hence, this study aims to evaluate the qualities of photogrammetric products (point cloud and 3D models) by employing several novel methods for more efficient integration with game engines. Furthermore, the study's goal is to ascertain whether construction safety improvement can benefit from integrating game engines with 3D models generated from UAV photogrammetry. A game is developed to provide virtual instructions to workers and safety associates based on OSHA regulations through the integration of a UAV-derived 3D model and game engine. On the other hand, BIM is another source of 3D models, and in this study, the potentialities and limitations of BIM technology in improving safety management are discussed. In addition, a comprehensive framework is developed to integrate BIM data and a game engine. Finally, two case studies are conducted on real-life scaffolding accident simulation and emergency evacuation modeling following the framework.
... Building maps of the environment from sensor data requires addressing the SLAM problem [11], [37] involving several subtasks such as incremental pose estimation [40], place recognition [41], loop closing [12], and optmization [2], [32]. Using laser scanning to map and study forests has been extensively studied [30], [44], with considerable attention given to airborne laser scanner [5], [18], [34], [38] and TLS data [1], [6], [8], [19], [20], [26], [42], [43]. Mobile laser scanning has only recently become a more viable approach, usually in the form of handheld or UAV platforms [3], [7], [22], [30]. ...
Forests play a crucial role in our ecosystems, functioning as carbon sinks, climate stabilizers, biodiversity hubs, and sources of wood. By the very nature of their scale, monitoring and maintaining forests is a challenging task. Robotics in forestry can have the potential for substantial automation toward efficient and sustainable foresting practices. In this paper, we address the problem of automatically producing a forest inventory by exploiting LiDAR data collected by a mobile platform. To construct an inventory, we first extract tree instances from point clouds. Then, we process each instance to extract forestry inventory information. Our approach provides the per-tree geometric trait of "diameter at breast height" together with the individual tree locations in a plot. We validate our results against manual measurements collected by foresters during field trials. Our experiments show strong segmentation and tree trait estimation performance, underlining the potential for automating forestry services. Results furthermore show a superior performance compared to the popular baseline methods used in this domain.
... Bahru and Ding [12] investigated the effects of forest attributes including stand density, canopy leaf area, and DBH on species litter production. Some studies used remote sensing or LiDAR-based data to quantify forest attributes, monitor forest inventory, and subsequently simulate fire environment and fire risk based on forest attributes and forest fuel [13,14]. However, less attention has been paid to modeling and linking forest attributes with comprehensive fire environments based on fine-scale field data and explaining forest attributes from a fire environment perspective. ...
Forest ecosystem attributes and their spatial variation across the landscape have the potential to subsequently influence variations in fire behavior. Understanding this variation is critical to fire managers in their ability to predict fire behavior and rate of spread. However, a fine-scale description of fuel patterns and their relationship with overstory and understory attributes for north-central Appalachia is lacking due to the complicated quantification of variations in topography, forest attributes, and their interactions. To better understand the fire environment in north-central Appalachia and provide a comprehensive evaluation based on fine-scale topography, ninety-four plots were established across different aspects and slope positions within an oak–hickory forest located in southeast Ohio, USA, which historically fell within fire regime group I with a fire return interval ranging from 7 to 26 years. The data collected from these plots were analyzed by four components of the fire environment, which include the overstory, understory, shrub and herbaceous layers, surface fuels, and fuel conditions. The results reveal that fuel bed composition changed across aspects and slope position, and it is a primary factor that influences the environment where fire occurs. Specifically, the oak fuel load was highest on south-facing slopes and in upper slope positions, while maple fuel loads were similar across all aspects and slope positions. Oak and maple basal areas were the most significant factors in predicting the oak and maple fuel load, respectively. In the shrub and undergrowth layers, woody plant coverage was higher in upper slope positions compared to lower slope positions. Overstory canopy closure displayed a significant negative correlation with understory trees/ha and woody plant variables. The findings in this study can provide a better understanding of fine-scale fuel bed and vegetation characteristics, which can subsequently feed into fire behavior modeling research in north-central Appalachia based on the different characterizations of the fire environment by landscape position.
... As a result, in order to overcome these limitations, alternative or complementary methods have been explored actively (Hyyppä et al., 2000;Williams et al., 1994). In this regards, remote sensing technology has emerged as an effective technique for obtaining multidimensional and multi-temporal information about Earth's surface (White et al., 2016) has successfully been employed to extract the structural characteristics of forests (Wulder, 1998;Wulder et al., 2012). ...
... The LiDAR point cloud data enables the estimation of the three-dimensional structure of tree canopies in different canopy layers with high accuracy, particularly during harvest periods (Lefsky et al., 2002;Zhang et al., 2017). By taking advantage of LiDAR data characteristics, this technology has been extensively applied in forest-related applications and studies over the past two decades (Hyyppä et al., 2012;Popescu et al., 2003;Wulder et al., 2012). ...
Currently, the information on the structural attributes of forests, such as the diameter at breast height (DBH) and the aboveground biomass (AGB), is being used widely in various disciplines. In this study, we first proposed a novel tree detection algorithm called multi‐scale individual tree detection (MSITD) algorithm, which combines the strengths of raster‐based and point‐based approaches in order to detect individual trees from LiDAR data accurately. After tree detection, the DBH and AGB attributes were estimated using the ground control data and metrics extracted from LiDAR data, adopting the safe semi‐supervised regression (SAFER) algorithm specifically designed for addressing regression problems with limited sample data. The performances of these algorithms were evaluated within a 10‐fold nested cross‐validation approach, utilizing the LiDAR data available in the NEWFOR project. The evaluation of the obtained results revealed that both the MSITD algorithm and the SAFER algorithm demonstrate substantial superiority compared to the benchmark algorithms in tree detection, especially for the understory trees, and forest structural attributes estimation, respectively. On average, the MSITD algorithm exhibited a 13% better performance in terms of extraction rate and an 11% better performance in terms of matching rate compared to the benchmark individual tree detection algorithms. For forest structural attributes estimation, the SAFER algorithm provided superior predictions compared to the benchmark ML algorithms, with the average RMSE of 3.38 cm, MAE of 2.84 cm, and R² of 0.59 for DBH and the average RMSE of 75.79 kg, MAE of 70.02 kg, and R² of 0.56 for AGB.
... The estimation methods of vegetation carbon storage are mainly divided into three types [41]: (1) The first is survey-based estimation, namely estimation using regional forest survey data, and this method usually gives the most accurate results but with extremely high labor and time costs [42]. (2) The second is remote-sensing-based estimation, for which commonly used data include optical remote sensing data, LiDAR data and synthetic aperture radar (SAR) satellite data [43]. For instance, Vincent et al. used WorldView-2 and LiDAR data to perform a fine estimation of vegetation carbon storage in Auckland, New Zealand, and the accuracy reached up to 95.9% [44]. ...
Forest and its dynamics are of great significance for accurately estimating regional carbon sequestration, emissions and carbon sink capacity. In this work, an efficient framework that integrates remote sensing, deep learning and statistical modeling was proposed to extract forest change information and then derive forest carbon storage dynamics during the period 2017 to 2020 in Jiangning District, Nanjing, Eastern China. Firstly, the panchromatic band and multi-spectral bands of GF-1 images were fused by using four different methods; Secondly, an improved Mask-RCNN integrated with Swin Transformer was devised to extract forest distribution information in 2020. Finally, by using the substitution strategy of space for time in the 2017 Forest Management and Planning Inventory (FMPI) data, local carbon density allometric growth equations were fitted by coniferous forest and broad-leaved forest types and compared, and the optimal fitting was accordingly determined, followed by the measurements of forest-change-induced carbon storage dynamics. The results indicated that the improved Mask-RCNN synergizing with the Swin Transformer gained an overall accuracy of 93.9% when mapping the local forest types. The carbon storage of forest standing woods was calculated at 1,449,400 tons in 2020, increased by 14.59% relative to that of 2017. This analysis provides a technical reference for monitoring forest change and lays a data foundation for local agencies to formulate forest management policies in the process of achieving dual-carbon goals.
