1 Introduction

Forest biomass is essential for estimating carbon stocks, guiding management, and monitoring disturbances. Field measurements are accurate but limited to small plots, making large-scale assessments costly and labour-intensive. Airborne Laser Scanning (ALS) helps overcome these limits by scaling plot data to wall-to-wall maps. The area-based approach (ABA) is a common method for estimating forest attributes from ALS data. This workshop guides participants through the ABA process for mapping aboveground biomass (AGB), including processing LiDAR data, integrating field measurements, and creating large-scale maps. It emphasizes understanding ABA’s assumptions, strengths, limitations, and practical uses in biomass and carbon assessment. By the end, participants will gain practical skills for ALS-based forest mapping for inventories, monitoring, and sustainable management.

2 Aim of the Workshop

The primary aim of this workshop is to produce forest attributes—such as canopy height and canopy cover—as well as wall-to-wall maps of aboveground biomass (AGB; m³ ha⁻¹) from ALS data for a selected area of interest (AOI) using AI-based modelling approaches. To achieve this, the entire processing workflow is implemented in R, providing a fully reproducible and open-source environment. Participants will work with a suite of R packages designed to process, analyse, and model airborne LiDAR data for forest and ecological applications.

3 Target audience

Students, researchers, and professionals in forestry, ecology, remote sensing, or environmental sciences interested in local tools for forest monitoring. Only basic R coding knowledge is required.

4 Expected outcomes

Participants will:

  • Access and process ALS datasets for forest applications.

  • Develop and apply an area-based upscaling model to generate wall-to-wall stem volume or aboveground biomass maps.

  • Visualize, interpret, and compare stem volume patterns across different forest types and land covers.

  • Integrate ALS-derived metrics with field measurements to improve biomass and carbon stock estimates.

  • Confidently use R and QGIS for geospatial data analysis and forest monitoring workflows.

5 Requirements

Computer with R and QGIS software.

6 Schedule

February 9, 9:00 AM – 2:00 PM (EST)

7 Essential R packages

  • lidR: Tools for reading, processing, visualising, and extracting metrics from airborne LiDAR (ALS) point cloud data (Roussel et al., 2015).
  • rLiDAR: Set of tools for reading, processing and visualizing small set of LiDAR (Light Detection and Ranging) data for forest inventory applications (Silva et al., 2015).
  • RCSF: Cloth Simulation Filter (CSF) ground filtering algorithm based on cloth simulation (Zhang et al., 2016).
  • raster: Reading, writing, and manipulating raster spatial data (Hijmans et al., 2015).
  • terra: Modern framework for handling large raster and vector datasets (Hijmans et al., 2022).
  • sf: Vector spatial data using simple features.
  • dplyr: Data manipulation (filtering, grouping, summarising).
  • ggplot2: High-quality data visualisations.
  • plot3D: Static 3D visualisations for point clouds.
  • ranger: Fast Random Forest implementation (Wright et al., 2019).
  • xgboost: Gradient boosting decision trees (Chen et al., 2019).
  • usdm:Uncertainty Analysis for Species Distribution Models (Naimi et al., 2017).