The Ultimate Dictionary: Aerial Inventory and Forest Data Analysis with Drones
Table of Contents
Navigate the technical world of modern forestry with confidence. This comprehensive glossary is your guide to all the terms and concepts in airborne inventory and forest data analysis with drones. Whether you are a forest owner, consultant, student or just curious, this resource will help you understand everything from basic GIS concepts to advanced LiDAR processing and its applications in precision forestry.
Basic concepts and technology
- Remote sensing: Technology and science for collecting, processing and interpreting information about the Earth's surface and objects on it without being in direct physical contact.
- Geographical Information System (GIS): A computer system for capturing, storing, controlling, integrating, analyzing and visualizing spatially referenced data.
- Grid data: A data model that represents a surface as a grid of pixels. Each pixel has a value representing an attribute (e.g. color, height). Examples: orthophotos, elevation models
- Vector data: A data model that uses points, lines and polygons to represent objects with discrete boundaries. Examples: tree positions (points), stand boundaries (polygons).
- Coordinate system: A reference system to uniquely define positions. In Sweden it is most often used SWEREF 99 TM. Globally used WGS 84.
- Geodata: Data with an explicit geographic component linking it to a specific location on Earth.
- GNSS (Global Navigation Satellite System): The collective name for satellite-based navigation systems, including GPS (USA), GLONASS (Russia), Galileo (Europe) and BeiDou (China).
- IMU (Inertial Measurement Unit): A sensor that measures an object's angular velocity and orientation, critical for calculating the exact position of the drone and sensor at any given moment.
- Spatial resolution: A measure of the smallest detail that can be discerned in a dataset. For raster data, this is usually the same as the pixel size (GSD).
Planning, Rules and Preparation
Air missions: The specific, planned data collection mission, including purpose, area and methodology.
Area of Interest (AOI): The precisely defined geographical area to be mapped.
Flight planning: The process of defining in software the flight routes, altitude, speed and image overlap to ensure complete data coverage.
Authorization: Legal approval from Transport Agency for specific flight operations, such as flight beyond visual line of sight (BVLOS).
UAS operator: The person or organization legally responsible for drone operations.
Remote pilot: The certified person operating the drone.
Risk analysis (SORA): A standardized methodology (Specific Operations Risk Assessment) to analyze and mitigate the risks of a drone mission.
Control Zone (CTR): A controlled airspace around an airport where air traffic control authorization is required.
Soil dissolution (GSD): The actual size of a single pixel on the ground. Lower GSD means higher detail.
Support Point / Pass Point (GCP): A point on the ground with precisely known coordinates, measured with high precision, used to improve the absolute accuracy of the final products.
Checkpoint: A measured point used to independently evaluate the geometric accuracy of the finished result.
Reference station / Base station: A fixed GNSS receiver at a known position transmitting real-time correction data (RTK) or logs data for post-processing (PPK).
Hardware: Drones and Sensors
Drones / UAV (Unmanned Aerial Vehicle): The unmanned, flying ship itself.
UAS (Unmanned Aircraft System): The complete system, including drone, ground station, data link and payload.
Multirotor / Copter: Multi-rotor drones (quad, hexa, octocopter). Stable and flexible for detailed surveys.
Fixed-wing drones: Drone with wings like an airplane. Energy efficient and suitable for covering large areas quickly.
VTOL (Vertical Take-Off and Landing): A hybrid between fixed-wing and multi-rotor, capable of vertical lift and efficient horizontal flight.
Payload: The equipment (mainly sensors) that the drone carries to perform its mission.
Gimbal / Cardan suspension: A motorized, gyro-stabilized suspension that keeps the sensor perfectly stable regardless of the drone's movements.
RGB camera: A standard camera that captures light in the visible spectrum (Red, Green, Blue).
Multispectral camera: An advanced camera that captures data in several specific wavelength bands, including Near-infrared (NIR) and Red Edge (Red Edge), which are sensitive to the health of vegetation.
Hyperspectral camera: An extremely advanced camera that captures data in hundreds of wavelength bands, providing a detailed spectral signature for each pixel.
Thermal camera / Thermal imaging camera: A sensor that measures thermal radiation (heat) and creates images based on the temperature of surfaces.
LiDAR (Light Detection and Ranging): An active sensor technology that sends out laser pulses and measures the time of reflection. This creates a highly detailed three-dimensional point cloud of the environment.
RTK (Real-Time Kinematic): A GNSS technology that uses real-time correction data to achieve centimeter accuracy in drone positioning.
PPK (Post-Processed Kinematic): A technique where the GNSS data of the drone and a reference station are combined after the flight to calculate an accurate flight path, often more robust than RTK.
Flight and Data Collection
Image overlap: The percentage that an image overlaps its neighboring images. Necessary for photogrammetry. Divided into frontal overlap (in the direction of flight) and lateral overlap (between airlines).
Raw data: Raw, primary data directly from the sensor (e.g. RAW images, .las files from LiDAR).
Laser pulse: A single, short-lived pulse of laser light emitted from a LiDAR sensor.
Return / Eco: A detected reflection of a laser pulse. One pulse can generate multiple returns (first return from canopy, last return from ground), which is crucial to see through vegetation.
Intensity (LiDAR): The strength of the reflected laser return. The value can provide information about the material that reflected the pulse.
Line of sight (VLOS): Regulations that require the remote pilot to be in direct visual contact with the drone at all times.
Out of sight (BVLOS): Flight operations where the drone flies outside the pilot's direct field of vision, which requires special authorizations.
Data Treatment and Processing
Photogrammetry: The science of extracting 3D information from 2D images.
Structure from Motion (SfM): A key photogrammetric technique that calculates the 3D structure of a scene from a series of overlapping images.
Point clouds: A large collection of points in a 3D coordinate system (X, Y, Z) representing the outer surface of objects and terrain.
Classification (of point clouds): The process of automatically assigning a category to each point (e.g. Land, Vegetation, Building, Noise).
Normalization (of point clouds): The process of converting the height of points above sea level to height above ground level. This creates a point cloud that shows the true height of trees and objects.
Georeferencing: The process of assigning data to a real geographic coordinate system, so that it can be displayed and analyzed with other geographic data.
Orthomosaic / Orthophoto: A geometrically corrected and composite aerial image with a uniform scale, like a map.
Digital Surface Model (DSM): An elevation model representing the surface including all objects, such as buildings and the canopy of trees.
Digital Terrain Model (DTM): An elevation model that represents only the "bare" ground surface, after filtering out vegetation and buildings.
Canopy Height Model (CHM): A model showing the height of trees above the ground. It is created by subtracting the terrain model from the surface model: CHM = DSM - DTM.
Forestry Analysis, Metrics and Variables
Stocks: A delimited forest area that is relatively uniform and forms a basic unit for forest management planning.
Individual tree segmentation (ITS): The process of automatically identifying and delimiting individual tree crowns using algorithms.
Tree height: The height of a tree, from ground level to the highest point of the tree, one of the most accurate variables from drone data.
Crown diameter: The average width of a tree's crown, an important parameter for estimating other tree variables.
Diameter at breast height (dbh): Tree trunk diameter measured at 1.3 meters height. Cannot be measured directly from the air but is estimated with high accuracy by correlation with tree height and crown diameter.
Base area: The sum of the trunk cross-sectional areas of all trees at breast height per hectare (m²/ha). A key value to describe the density of a stand.
Wood storage / Volume: The total volume of tree stems in a stand, usually expressed in forest cubic meters per hectare (m³sk/ha).
Density / Stem count: The number of trees (stems) per unit area.
Tree species identification: The process of determining which species a tree or stand belongs to, often using multispectral data.
Vegetation index (e.g. NDVI, NDRE): Mathematical combinations of spectral bands to highlight the health and vigor of vegetation.
Applications and Deliverables
Basic forestry data: The set of key variables (tree height, volume, basal area, etc.) that describe the condition of the forest.
Digital forest management plan: A modern, data-driven forest management plan where maps and stand records are based on high-resolution drone data.
Proposed action: Data-driven recommendations for forest management, such as clearing needs, thinning needs or identification of mature for final felling stock.
Forest pest inventory: Effective mapping and quantification of damage caused by e.g. Spruce Bark Beetle, windfalls (windfalls), drought or fire.
Nature value assessment: Identification and mapping of sites with high nature value, such as old trees, dead wood and gaps.
Follow-up of planting (Revegetation inventory): Control of survival and growth of newly planted seedlings.
Precision forestry: A forest management concept that uses high-resolution data to adapt management actions to the local variations within a stand.
Advanced concepts and related fields
Machine learning: Use of algorithms trained on data to recognize patterns, e.g. for tree species identification or damage classification.
Deep Learning: A subcategory of machine learning that uses deep neural networks, particularly effective for image analysis.
Change Detection: Comparison of data collected at different points in time to quantify changes such as growth or deforestation.
Data Fusion: The process of combining data from multiple sensors (e.g. LiDAR and hyperspectral data) to create a richer analysis.
Spectral signature: The unique reflectance profile of a material across different wavelengths, which is used to identify, for example, different tree species.
Do you have questions or want to know how airborne inventory can streamline your forestry operations? Contact us today for a consultation and discover the potential of modern forest data analysis.