8. Subset Selection in Regression (Monographs on Statistics and Applied Probability, no. 40). By A. J. Miller. ISBN 0 412 35380 6. Chapman and Hall, London, 1990. 240 pp. £25.00.

Subset Selection in Regression

lidR: An R package for analysis of Airborne Laser Scanning (ALS) data

https://pure.mpg.de/pubman/item/item_3033361_3/component/file_3033369/1-s2.0-S0034425719300689-main.pdf

Historical background and current developments for mapping burned area from satellite Earth observation

Estimating forest structural attributes in planted forests is crucial for sustainably management of forests and helps to understand the contributions of forests to global carbon storage. The Unmanned Aerial Vehicle-Light Detecting and Ranging (UAV-LiDAR) has become a promising technology and attempts to be used for forest management, due to its capacity to provide highly accurate estimations of three-dimensional (3D) forest structural information with a lower cost, higher flexibility and finer resolution than airborne LiDAR. In this study, the effectiveness of plot-level metrics (i.e., distributional, canopy volume and Weibull-fitted metrics) and individual-tree-summarized metrics (i.e., maximum, minimum and mean height of trees and the number of trees from the individual tree detection (ITD) results) derived from UAV-LiDAR point clouds were assessed, then these metrics were used to fit estimation models of six forest structural attributes by parametric (i.e., partial least squares (PLS)) and non-parametric (i.e., k-Nearest Neighbors (k-NN) and Random Forest (RF)) approaches, within a Ginkgo plantation in east China. In addition, we assessed the effects of UAV-LiDAR point cloud density on the derived metrics and individual tree segmentation results, and evaluated the correlations of these metrics with aboveground biomass (AGB) by a sensitivity analysis. The results showed that, in general, models based on both plot-level and individual-tree-summarized metrics (CV-R2 = 0.66–0.97, rRMSE = 2.83–23.35%) performed better than models based on the plot-level metrics only (CV-R2 = 0.62–0.97, rRMSE = 3.81–27.64%). PLS had a relatively high prediction accuracy for Lorey’s mean height (CV-R2 = 0.97, rRMSE = 2.83%), whereas k-NN performed well for predicting volume (CV-R2 = 0.94, rRMSE = 8.95%) and AGB (CV-R2 = 0.95, rRMSE = 8.81%). For the point cloud density sensitivity analysis, the canopy volume metrics showed a higher dependence on point cloud density than other metrics. ITD results showed a relatively high accuracy (F1-score > 74.93%) when the point cloud density was higher than 10% (16 pts·m−2). The correlations between AGB and the metrics of height percentiles, lower height level of canopy return densities and canopy cover appeared stable across different point cloud densities when the point cloud density was reduced from 50% (80 pts·m−2) to 5% (8 pts·m−2).

Estimating forest structural attributes using UAV-LiDAR data in Ginkgo plantations

Point cloud filtering on UAV based point cloud

Light detection and ranging (LiDAR) is an emerging remote-sensing technology with potential to assist in mapping, monitoring, and assessment of forest resources. Despite a growing body of peer-reviewed literature documenting the filtering methods of LiDAR data, there seems to be little information about qualitative and quantitative assessment of filtering methods to select the most appropriate to create digital elevation models with the final objective of normalizing the point cloud in forestry applications. Furthermore, most algorithms are proprietary and have high purchase costs, while a few are openly available and supported by published results. This paper compares the accuracy of seven discrete return LiDAR filtering methods, implemented in nonproprietary tools and software in classification of the point clouds provided by the Spanish National Plan for Aerial Orthophotography (PNOA). Two test sites in moderate to steep slopes and various land cover types were selected. The classification accuracy of each algorithm was assessed using 424 points classified by hand and located in different terrain slopes, cover types, point cloud densities, and scan angles. MCC filter presented the best overall performance with an 83.3% of success rate and a Kappa index of 0.67. Compared to other filters, MCC and LAStools balanced quite well the error rates. Sprouted scrub with abandoned logs, stumps, and woody debris and terrain slopes over 15° were the most problematic cover types in filtering. However, the influence of point density and scan-angle variables in filtering is lower, as morphological methods are less sensitive to them.

/pdf/a-comparison-of-open-source-lidar-filtering-algorithms-in-a-4lx3aoo05q.pdf

A Comparison of Open-Source LiDAR Filtering Algorithms in a Mediterranean Forest Environment

Airborne Laser Scanning (ALS) is capable of estimating a variety of forest parameters using different metrics extracted from the normalized heights of the point cloud using a Digital Elevation Model (DEM). In this study, six interpolation routines were tested over a range of land cover and terrain roughness in order to generate a collection of DEMs with spatial resolution of 1 and 2 m. The accuracy of the DEMs was assessed twice, first using a test sample extracted from the ALS point cloud, second using a set of 55 ground control points collected with a high precision Global Positioning System (GPS). The effects of terrain slope, land cover, ground point density and pulse penetration on the interpolation error were examined stratifying the study area with these variables. In addition, a Classification and Regression Tree (CART) analysis allowed the development of a prediction uncertainty map to identify in which areas DEMs and Airborne Light Detection and Ranging (LiDAR) derived products may be of low quality. The Triangulated Irregular Network (TIN) to raster interpolation method produced the best result in the validation process with the training data set while the Inverse Distance Weighted (IDW) routine was the best in the validation with GPS (RMSE of 2.68 cm and RMSE of 37.10 cm, respectively).

/pdf/interpolation-routines-assessment-in-als-derived-digital-33kuv04cx3.pdf

Interpolation routines assessment in ALS-derived Digital Elevation Models for forestry applications

Mediterranean pine forests in Spain experience wildland fire events with different frequencies, intensities, and severities which result in diverse socio-ecological consequences. In order to predict fire severity, spectral indices derived from remotely sensed images have been used extensively. Such spectral indices are usually used in combination with ground sampling to relate detected radiometric changes to actual fire effects. However, the potential of the tridimensional information captured by Airborne Laser Scanners (ALS) to severity mapping has been less explored. With the objective of addressing this question, in this paper, explanatory variables extracted from ALS point clouds are related to field estimations of the Composite Burn Index collected in four fires located in Aragon (Spain). Logistic regression models were developed and statistically tested and validated to map fire severity with up to 85.5% accuracy. The canopy relief ratio and the percentage of all returns above one meter height were the most significant variables and were therefore used to create a continuous map of severity levels.

Forest Fire Severity Assessment Using ALS Data in a Mediterranean Environment

Use of low point density ALS data to estimate stand-level structural variables in Mediterranean Aleppo pine forest

The account of total biomass can assist with the evaluation of climate regulation policies from local to global scales. This study estimates total biomass (TB), including tree and shrub biomass fractions, in Pinus halepensis Miller forest stands located in the Aragon Region (Spain) using Airborne Laser Scanning (ALS) data and fieldwork. A comparison of five selection methods and five regression models was performed to relate the TB, estimated in 83 field plots through allometric equations, to several independent variables extracted from ALS point cloud. A height threshold was used to include returns above 0.2 m when calculating ALS variables. The sample was divided into training and test sets composed of 62 and 21 plots, respectively. The model with the lower root mean square error (15.14 tons/ha) after validation was the multiple linear regression model including three ALS variables: the 25th percentile of the return heights, the variance, and the percentage of first returns above the mean. This study confirms the usefulness of low-density ALS data to accurately estimate total biomass, and thus better assess the availability of biomass and carbon content in a Mediterranean Aleppo pine forest.

/pdf/estimation-of-total-biomass-in-aleppo-pine-forest-stands-3liocvyn9p.pdf

Antonio Luis Montealegre

Papers

A Comparison of Open-Source LiDAR Filtering Algorithms in a Mediterranean Forest Environment

Interpolation routines assessment in ALS-derived Digital Elevation Models for forestry applications

Forest Fire Severity Assessment Using ALS Data in a Mediterranean Environment

Use of low point density ALS data to estimate stand-level structural variables in Mediterranean Aleppo pine forest

Estimation of Total Biomass in Aleppo Pine Forest Stands Applying Parametric and Nonparametric Methods to Low-Density Airborne Laser Scanning Data