station logo

LIDAR Basics


Home
USDA Forest Service, Pacific Northwest Research Station


FUSION splash
                      screen

The FUSION/LDV analysis and visualization software is a set of public domain tools to help you understand and use LIDAR data.



LIDAR collection system uses a powerful laser sensor comprised of a transmitter and receiver, a geodetic-quality Global Positioning System (GPS) receiver and an Inertial Measurement Unit (IMU). The laser sensor is precision mounted to the underside of an aircraft. Once airborne, the sensor emits rapid pulses of infrared laser light, which are used to determine ranges to points on the terrain below. Combined with the GPS and IMU data, these ranges are converted into XYZ measurements that represent all objects and surfaces visible from the aircraft.





Points only
points
                  and tree objects
rendered trees

The lidar point cloud includes direct measurements of various canopy elements including stems, branches, and foliage. This series of images shows the lidar data for above-ground vegetation, tree objects inferred from the data, and a visual simulation of the tree objects and terrain rendered with PNW Research Station’s EnVision software.


LIDAR Overview

What is LIDAR?

Lidar uses laser light to measure distances. It is used in many ways, from estimating atmospheric aerosols by shooting a laser skyward to catching speeders in freeway traffic with a handheld laser-speed detector. Airborne laser-scanning technology is a specialized, aircraft-based type of lidar that provides extremely accurate, detailed 3-D measurements of the ground, vegetation, and buildings. Developed in just the last 20 years, one of lidar’s first commercial uses in the United States was to survey powerline corridors to identify encroaching vegetation. Additional uses include mapping landforms and coastal areas. In open, flat areas, ground contours can be recorded from an aircraft flying overhead providing accuracy within 6 inches of actual elevation. In steep, forested areas accuracy is typically in the range of 1 to 2 feet and depends on many factors, including density of canopy cover and the spacing of laser shots. The speed and accuracy of lidar made it feasible to map large areas with the kind of detail that before had only been possible with time-consuming and expensive ground survey crews.

Federal agencies such as the Federal Emergency Management Administration (FEMA) and U.S. Geological Survey (USGS), along with county and state agencies, began using lidar to map the terrain in flood plains and earthquake hazard zones. The Puget Sound Lidar Consortium, an informal group of agencies, used lidar in the Puget Sound area and found previously undetected earthquake faults and large, deep-seated, old landslides. In other parts of the country, lidar was used to map highly detailed contours across large flood plains, which could be used to pinpoint areas of high risk. In some areas, entire states have been flown with lidar to produce more accurate digital terrain data for emergency planning and response. Lidar mapping of terrain uses a technique called “bareearth filtering.” Laser scan data about trees and buildings are stripped away, leaving just the bare-ground data. The VMaRS team first began their lidar research involving application in forests in 1997 to find out how much accuracy was lost in lidar flights over areas with heavy forest cover. They wanted to better understand the level of error in lidar mapping of the ground through forest canopy, to be used in analyzing terrain maps in forested areas. The team and their University of Washington collaborators found that the data thrown away by geologists were a rich source of information for foresters, a finding that has been well corroborated by lidar forestry research groups around the world.

How Does LIDAR work?

The use of lasers has become commonplace, from laser printers to laser surgery. In airborne-laser-mapping lidar, lasers are taken into the sky. Instruments are mounted on a single- or twin-engine plane or a helicopter. Airborne lidar technology uses four major pieces of equipment (see figureat left ). These are a laser emitter-receiver scanning unit attached to the aircraft; global positioning system (GPS) units on the aircraft and on the ground; an inertial measurement unit (IMU) attached to the scanner, which measures roll, pitch, and yaw of the aircraft; and a computer to control the system and store data. Several types of airborne lidar systems have been developed; commercial systems commonly used in forestry are discrete-return, small-footprint systems. “Small footprint” means that the laser beam diameter at ground level is typically in the range of 6 inches to 3 feet. The current generration of laser scanners send up to 1,000,000 pulses of light per second to the ground and measures how long it takes each pulse to reflect back to the unit. These times are used to compute the distance each pulse traveled from scanner to ground. The GPS and IMU units determine the precise location and attitude of the laser scanner as the pulses are emitted, and an exact coordinate is calculated for each point. The laser scanner uses an oscillating mirror or rotating prism (depending on the sensor model), so that the light pulses sweep across a swath of landscape below the aircraft. Large areas are surveyed with a series of parallel flight lines. The laser pulses used are safe for people and all living things. Because the system emits its own light, flights can be done day or night, as long as the skies are clear.

Thus, with distance and location information accurately determined, the laser pulses yield direct, 3-D measurements of the ground surface, vegetation, roads, and buildings. Millions of data points are recorded, so many that lidar creates a 3-D data cloud. After the flight, software calculates the final data points by using the location information and laser data. Final results are typically produced in weeks, whereas traditional ground-based mapping methods took months or years. The first acre of a lidar flight is expensive, owing to the costs of the aircraft, equipment, and personnel. But when large areas are covered, the costs can drop to less than $1 per acre. The technology is commercially available through a number of sources.

Early Research (summary)

Early research conducted by the team to evaluate LIDAR was conducted at the Blue Ridge study site, a roughly 2-square- mile area within Capitol State Forest. Capitol State Forest covers 90,000 acres near Olympia, Washington, and is managed by the Washington Department of Natural Resources (DNR). The Blue Ridge site was the location of a large study on silvicultural regimes, being conducted under a partnership between the Silviculture and Forest Models Team and local DNR managers. Designed to study silvicultural options for young growth Douglas-fir forests, the study treatments included commercial thinning, patch cuts, group selection, creation of two-age stands, clearcut harvest, and uncut stands. Most of the area was covered with 70-year-old Douglas-fir stands in 1997 before the study. The larger trees were mostly Douglas-fir with some hemlock and western redcedar; the largest trees were reaching 40 inches diameter at breast height. A few pole-sized trees and western hemlock seedlings were scattered throughout. For the silvicultural options study, data were gathered on the stands 1 year before the treatments, 1 year after treatments, and every 5 years thereafter. The Blue Ridge study area was an ideal area for testing lidar in forested terrain. A detailed set of field plots, treatments that would create varying canopy densities, and post treatment monitoring create controlled, well-documented conditions for testing lidar. The first lidar flight over the site was in 1998, a month before the treatments began; the second lidar flight was in early 1999, after treatments were completed and before the growing season began; and the third was in late 2003, at the end of the growing season. The time between the latter two flights covered a five-growing-season span, and they also gave scientists a chance to compare a leaf-off scan, done before deciduous trees leafed out for the year, with a leaf-on scan, done while leaves were still on the deciduous trees. Thus scientists were able to test lidar results under this variation in conditions.

The first objective for lidar was to test its accuracy in mapping the ground surface under varying canopy densities. The research team didn’t expect lidar to map the terrain very well through heavy cover. They thought there’d be big differences in accuracy between the clearcut and the uncut 70-year-old stand. But it turned out that the average error was about the height of a standard field boot, or 9 inches (Read more). Moreover, accuracy was statistically the same in the clearcut and the 70-year-old stand. This highly accurate lidar ground surface was generated by filtering out only those points that hit the ground by using a bare-earth filtering process. Common practice at the time was to discard all non-ground data points and only use the bare-earth points to produce a digital elevation model. However the team of scientists felt that the measurements from the foliage and stems could tell a more important story regarding forest structure.

The team found that lidar had enormous potential for supplying information about the forests themselves, including data on stand and individual tree characteristics. Lidar produced accurate estimates of individual tree heights in the overstory, with results within 3 feet of the photogrammetrically-measured results. It also estimated crown diameter. With lidar data on individual tree crowns, scientists could generate a detailed canopy surface model. Average stand height could also be estimated from the data.

Next the scientists worked on estimating stand characteristics with lidar, which measures the location of reflecting surfaces within the stand, including foliage and branches. They found strong relations between lidar data and field-measured data for dominant height, stem volume, biomass, and basal area. (Basal area is defined as the cross-sectional area of all trees in a stand, as measured at breast height; the pretreatment basal area for the Blue Ridge stands ranged from 227 to 255 square feet per acre.) Also, intensity data, or the reflectance strength, from each lidar pulse, combined with information on canopy height, showed potential for distinguishing hardwood and conifer canopy areas. Thus lidar looked promising for estimating stand characteristics. The team compared the lidar data clouds from early 1999 and late 2003 and found that lidar successfully estimated the differences in growth among the treatment units. Lidar data, which were confirmed by traditional ground-based measurements, found that the unthinned control unit of mature trees had a total height growth of 3 to 9 feet over the five growing seasons, and remaining dominant trees in the heavily thinned stand had less height growth but greater expansion in their crowns. The 35-year-old stand in the study area had total height growth of 9 to 15 feet, considerably more height gain than the mature stands. Lidar’s accuracy, then, was comparable to stand inventory results from traditional methods. But labor-intensive traditional methods can be done on only selected plots or stands, because of the expense. Lidar, once fully validated, could potentially provide detailed, accurate assessments of stand characteristics and growth across an entire forest. The research team quickly realized that lidar would be a revolutionary technology in forestry.

This page was last updated on June 11, 2018 by Bob McGaughey