Mapping the Earth at the Speed of Light
|A DSM is a collection of 3D coordinates for certain points on the ground’s surface or objects covering those points.|
LiDAR technology provides a giant leap forward in our ability to map the Earth's surface using nonphotographic-based methods.
Most critical human activities require vast knowledge of the surrounding geography and environment. Understanding the effects of events such as sea level rise, global warming and hurricanes, and developmental activities such as precision farming and solar and wind site design all require accurate topographical maps. Unfortunately, most regions of the world lack an accurate representation of their topography, partly because of the high costs associated with conducting a precise survey of the land, and partly due to a lack of awareness of the importance of having accurate terrain models.
Since the world wars, aerial surveying has become widely accepted as an efficient way to map the Earth’s topography. However, photography-based aerial surveying remains unaffordable for many nations around the world. In the last several decades, mapping scientists and technology leaders have invested a tremendous amount of resources in the search for alternatives. One result of these efforts is LiDAR technology, which collects highly accurate information at the speed of light--or very nearly so, with the exception of the post processing that is required. Today, LiDAR is providing a revolution and a giant leap forward in our ability to map the Earth’s surface using non-photographic-based methods.
The term “laser,” which is an acronym for light amplification by stimulated emission of radiation, is defined as “a device that emits light (electromagnetic radiation) through a process of optical amplification based on the stimulated emission of photons.”* In other words, the laser is energized light (visible or not visible) represented by a flux of photons traveling in space at the known speed of light. The light becomes amplified (energized) through a process called pumping, which involves subjecting the light to an electric current or to a light with different wavelengths.
As energized light in the form of a laser travels in space or vacuum, it backscatters (reflects photons) when it collides with an object on its way. Some of the backscattered energy (photons) returns in the direction of the emitted device, where it is collected by a receiver.
|An intensity image is a measure of the strength of the backscatter as recorded by the system detector.|
With the advances in electronic timing devices, the emitted energized light is accurately timed as it leaves the device and as the backscattered energy arrives at the receiver. Knowing the time it took the energized light to travel back and forth to the object and the speed of light in the air, one can accurately calculate the distance between the light source and the collided object using the simple equation of speed:
Range = 1/2 (Time x Speed of Light)
Given that the time of travel back and forth and the speed of light are known, the distance between the laser source and the reflective object can be easily computed. Such distance is also called range.
Mapping scientists and technologists who were aware of the ability to measure the range from an aerial sensor to a ground feature used this laser-based range finding method to design a mapping sensor called LiDAR (light detection and ranging) to map the Earth’s topography. Range on its own is not sufficient to compute the required three-dimensional coordinates; other parameters also need to be measured. In addition to the laser source, LiDAR contains auxiliary systems to measure the position and orientation of the light-emitting device; an instrument to measure the rotational angles of the emitting device, called inertial measurement unit (IMU); and a survey-grade GPS receiver. The system also contains a scanning mirror or rotating prism to scatter the laser shots on the ground (see Figures 1 and 2). As the energized light is produced as a flux of photons at a high repetition rate by the laser source, it is directed to the oscillating mirror or the spinning prism to be scattered on the ground along the flight path of the aircraft.
When a laser pulse is generated and directed to the mirror or the prism, it eventually reaches the ground. The time of the pulse departure is noted accurately. It is worth mentioning here that as the pulse travels through space, it spreads out to form a conical shape similar to the light generated by a flashlight. Therefore, when the pulse reaches the ground, it covers an area rather than a point. The footprint of the pulse on the ground is a function of the flying altitude and the aperture size of the source laser.
The laser pulse collides with the ground object on its way and reflects back partially or entirely. If the pulse hits a hard surface, such as the ground or a building, there will be only one reflection (or return). However, if the pulse hits a tree branch, the pulse (flux of photons) may split so that part of the pulse returns to the system while the rest of the pulse continues its travel between the branches until it hits another branch or the ground. That part of the pulse will then be reflected back to the system, where its arrival will be acknowledged and timed.
LiDAR technology has advanced rapidly over the last decade. Today’s aerial LiDAR systems offer a wide range of design and capabilities. The main operational factors that distinguish one system from another are the maximum flying altitude and the laser repetition rate.
The flying altitude plays a significant role in determining the efficiency of the system since the productivity rate or the ground coverage per hour is directly related to the altitude of operation. A lower altitude means reduced productivity (assuming all other factors remain the same). Most systems available today operate efficiently from an altitude of 10,000 feet or lower. Very few systems are capable of operating from an altitude exceeding 15,000 feet.
The laser repetition rate is the rate at which the laser pulses are generated and directed to the oscillating mirror or the spinning prism. A higher pulse rate provides higher productivity since it allows better operating parameters in term of flying speed, flying altitude and better ground point density. Most of the aerial LiDAR systems available today operate within a range of 150 to 500 KHz (or 150,000 to 500,000 pulses per second). Some other systems are capable of surveying millions of ground points per second.
Parameters such as the mirror oscillation speed (or prism spinning speed) and the field of view at which the laser shots are distributed around both sides of the flying path also play a role in distributing the laser pulses (shots) on the ground. However, these are not considered to be a discriminating factor in favoring one system over another since most systems have similar capabilities.
The laser source is unique in its characteristic with the exception of the portion of the light it uses and the power of the laser. However, there are a few main designs when it comes to the integration of an airborne LiDAR system. These include the linear system, the Gieger and photon counting system, and the flash LiDAR system.
The linear system is the most widely used system in the mapping profession. The principal of operation is similar to what is illustrated in Figures 1 and 2 and described in the previous section. Modern technologies enable linear systems to survey the ground at a rate of up to 500,000 points per second. This technology is mature enough to be used by the vast majority of LiDAR data providers.
|Figure 1. The laser shots are scattered along the flight path of the aircraft.|
The Geiger and photon counting system is a relative newcomer to the mapping community, and only a few of these systems, if any, are in commercial use today. However, it is expected to be widely used in the near future. The main difference between this type and the linear system is in the way it uses the flux of photons on the ground (Figure 1). While a linear system is a single detector or single receiver system, the Geiger and photon counting system uses an array of detectors similar to the CCD array of a digital camera. In another words, the single laser pulse is divided into an array of subpulses using a diffractive optical element (DOE) (see Figure 3). When these subpulses return to the sensor, they are received by an equal number of segmented anodes (Figure 4), resulting in a dense ground point collection or survey.
The flash LiDAR system, as its name implies, illuminates an area on the ground and samples it with an array of sensors in a similar fashion to the Geiger and photon counting system. This also results in an array of points on the ground instead of a single point.
The deliverables generated by LiDAR systems largely depend on the organization that is managing the data. However, most LiDAR systems are capable of producing digital surface models (DSMs), intensity images or full waveform digitization.
A DSM, which is the main product produced by all LiDAR systems, is a collection of three-dimensional coordinates (X,Y,Z) for certain points on the ground’s surface or objects covering those points, such as trees and buildings. The aggregation of these points builds the point cloud or the surface model that shapes the terrain. DSMs vary in density, and therefore, in the amount of detail they reveal about the ground surface. The image on page 18 is an example of a DSM presented as a hill shade.
An intensity image is not a digital surface model. Rather, it is a measure of the strength of the backscatter as recorded by the system detector. When this signal strength is presented in a raster format with a dynamic range of 0 to 255, the result is an image resembling a black-and-white photograph with a different meaning. Such an image is useful for certain mapping applications since it has the same positional accuracy of a DSM. The inset image on page 18 shows an example of an intensity image.
Full waveform digitization may be considered as another product, but it is really a recording of the same backscatter with higher frequency (in the range of nanoseconds). Still relatively new to the mapping community, waveform analysis has great potential in ground terrain analysis, especially in the analysis of natural resources, such as forests.
LiDAR-derived data is mainly used to model the ground terrain. However, digital surface models derived from LiDAR data are used for different applications. For engineers and surveyors, LiDAR-derived terrain models enhance both the accuracy and quality of their daily activities. The high density of LiDAR-derived ground points cannot be matched by any other surveying method, including photogrammetric methods.
The latest advances in LiDAR system design have made it possible to collect data with a density of 8 points per square meter from an airborne platform and hundreds, if not thousands, of points per square meter from a mobile ground platform. Figure 5 illustrates an example of an aerial data collection with a density of about 3.5 points per square meter. Such highly accurate and detailed terrain models are used for road building as well as route planning and engineering. LiDAR data is also used in mapping water bodies and analyzing the flow of surface water (see Figure 6).
Other applications for LiDAR data include disaster response, contour generation, land use/land cover, aviation obstruction analysis, 3D city modeling, transportation mapping, view shed analysis, biomass and forest management, hydrographic and coastline modeling, and change detection.
As new advances are made in LiDAR technologies, the cost of terrain mapping using laser-based sensors will continue to inch downward. The combination of improved benefits at lower costs will drive an increased use of laser-based scanning systems for terrain mapping through the rest of the decade--or at least until more efficient technologies emerge. Until then, we will continue mapping the Earth’s terrain at the speed of light.
Visit www.geodatapoint.com for a new series of technology-focused articles by Qassim A. Abdullah, PhD, CP, PRM, RPP.