Feature: Lidar Fact and Fiction
Professional Surveyor Magazine - April 2008
Todd A. Stennett and Sandra Wade-Grusky
While it may seem new, lidar (light detection and ranging) has actually been around for over 30 years now. The technology works similarly to radar (radio detection and ranging), but transmits and receives light from lasers rather than radio waves from radio signals. Its use goes all the way back to the 1970s with the introduction of systems by NASA. But through the 1970s and 80s and the first half of the 90s, the technology was far too cumbersome and temperamental to be deployed commercially. Its integrators worked hard to reduce weight and size, increase accuracy, and increase reliability.
The process took decades, but by the late 1990s, we saw the first commercial sensors deployed. The technology, as well as computing power that increases lidar speed and accuracy, has progressed to make lidar a preferred alternative for developing digital elevation models (DEM) of the Earth's surface.
Today, about 150 sensors are installed in aircraft throughout the world, with about half of these active in the private sector. The rest are retired or used by governments or educational institutions. As owners and operators of four of these sensors, we frequently come across a few myths about lidar. Let's take a look at a few that impact professional surveyors.
Myth #1: Lidar activity is not classified as land surveying and therefore is not governed and regulated accordingly
Well, the industry frequently operates under that assumption, but it's dangerous to do so. It varies by state and specific activity. To get to the truth, we sent certified letters to every state board of registrars. As expected, some simply cited the applicable references and code numbers, sidestepping a direct response. Proudly, some 22 states stepped up and rendered their positions on the matter.
We asked each state a series of questions on this topic. Did you know that in one or more states:
- The act of collecting lidar data is deemed to be regulated under state licensure law, including the operation of a lidar sensor.
- The act of computation of collected lidar data is deemed to be regulated under state licensure law.
- The act of occupying geodetic control points with GPS receivers (to position the aircraft) is deemed to be regulated under state licensure law.
- The act of measuring ground elevations with GPS receivers (used internally by a firm to check their lidar measurements) is deemed to be regulated under state licensure law.
The states vary widely on their experience and education requirements for Professional or Registered Land Surveyor licensure. And some states do not consider ANY of the above items to require a license.
I think we'd all agree that at some point in the process, coordinates are computed, and that those coordinates may be used for a variety of purposes. Many of these purposes are governed by licensure laws, and generally for good reason.
In general, it's wise to ensure coordination and legitimate oversight with a PLS on all lidar projects, and we find that local knowledge is usually efficient. (If you're experienced with or interested in lidar quality control, we're always interested to have more friends across the nation and world. Give us a shout, and express your interest.)
Myth #2: Lidar is cost prohibitive on most small projects
The minimum breakeven scope of work for some lidar firms is 20 to 40 acres of typical terrain, smaller if vegetation or access becomes an issue. While many firms don't wish to perform projects smaller than 25 to 50 square kilometers, others routinely perform high volumes of small sites. When performing on behalf of licensed professionals who get involved in the survey's groundwork, we price any state in the Southwest at about $3,500 plus $1.49 per acre, if you can provide lead time. Once the acreage reaches a few hundred, that per-acre fee can fall—a long way. The marginal cost to collect one more acre, for most sensors out there, is less than 10 cents. Our price for these marginal acres has gone as low as $0.15 per acre when we get into the largest projects, typically greater than 1,000 square miles. The above pricing applies to data in support of 2-foot contouring to ASPRS Class 1 standards. It covers mass points, cleaned to bald earth, along with the "removed features" file, which contains all the structures, vegetation, birds, aircraft, and other non-ground features we pick up while flying.
The trick is to take advantage of that low marginal cost and pool a few nearby projects, even with competitors that are cooperative. Also, national road shows are offered where firms fly predetermined areas around the country and sign up customers. With this, firms can take advantage of lidar for smaller sites anywhere in the United States and avoid having to meet a $25,000 or higher minimum project size. It just takes a bit of searching on your favorite search engine and gathering a few references.
Myth #3: Lidar can't support one-foot contouring
Yes, lidar can support 1-foot contouring, even to the most rigorous standards, but it may not be cost effective to deploy it for this purpose in certain conditions. We've run more photogrammetry firms through their own assessments than I can count, and the results are consistent. First, disbelief. Then, testing and approval at 2-foot interval mapping. Then, tighter flight constraints (especially with regard to GPS, where the majority of error is introduced) and testing at 1-foot intervals. Lidar is far better for these applications in some terrain and vegetation types and far better for pure volumetric projects such as landfills, earthmoving, etc.
On the other hand, lidar should NOT be used to support final, large, design-scale mapping of highly intricate terrain and features where breaklines and planimetrics are the vital component, unless you have a skilled team for the photogrammetric editing processes. While lidar will clearly show the crown of roadways and cul-de-sacs, it won't deliver a perfect face-of-curb. It is, after all, discrete point data and not continuous line data.
Some lidar providers are claiming accuracies up to 7 and 8 centimeters at 2-sigma. We believe it can be achieved, but typically promise only to 15 centimeters at this confidence interval, because that's easily repeatable.
Myth #4: Lidar can see through trees
Well, that would be cool, for sure. Unfortunately, it's only true in one sense. If you walk into an area of dense canopy and take a look at the ground, you can see that a certain percentage of the ground is hit by direct sunlight. That percentage very closely correlates to the percentage of your laser shots that will make it to ground. If you have 80 percent shadow on ground, then about one in five shots will make it to ground. With repetition rates now above 150,000 shots per second, a dense array of lidar points isn't difficult. We're shooting dozens of areas right now at around 10 shots per square meter, and sometimes triple that for our helicopter work. But the percentage opening in canopy is the percentage of these shots that will provide a reliable ground surface.
We've been called to survey a farmer's land to help eliminate his gopher dens in one case and to help locate buried treasure in another. Not with lidar. We've been called to locate, rescue, and recover people who've fallen into icy waters. Not with lidar. But for most terrain needs between 1-foot and 1-meter vertical accuracies, lidar is a more cost-effective choice. That's why the industry's largest photogrammetry firms have all jumped on board. Because they don't want competitors to offer a more cost-effective solution to their own clients, many firms have decided to get in the game, whether they choose to outsource or bring the capabilities in-house.
Think You Can't Afford Lidar? Guess Again
To surveyors who haven't yet used the technology, lidar sounds like some futuristic Star Trek invention that costs a fortune to use. While lidar does represent the cutting edge of mapping techniques, the average aerial lidar project actually doesn't cost more than one completed with traditional photogrammetric methods. In the surveying community, however, the benefits of lidar over traditional aerial photogrammetry, including cost, are not well understood.
The basic components of a lidar airborne laser mapping system include a laser scanner and cooling system, GPS receivers, and an inertial navigation system (INS). The laser scanner mounts within a properly outfitted aircraft and emits infrared laser beams at a high frequency. The scanner records the difference in time between the emission of the laser pulses and the reception of the reflected signal. A mirror mounted in front of the laser rotates and causes the laser pulses to sweep at an angle, back and forth along a line. The position and orientation of the aircraft is determined using phase-differenced kinematic GPS. GPS systems are located in the aircraft and at several ground stations within the area to be mapped. The orientation of the aircraft is then controlled and determined by the INS.
The round trip travel time of the laser pulses from the aircraft to the ground are measured and recorded, along with the position and orientation of the aircraft at the time of transmission of each pulse. After the flight, the vectors from the aircraft to the ground are combined with the aircraft position at the time of each measurement, and the 3D XYZ coordinates of each ground point are computed.
The system can be operated at various scan frequencies and altitudes, depending on the measurement accuracy dictated by project requirements and the regulated eye-safe range of the laser. By accurately timing the round trip travel time of the light pulses to the surface, it is possible to determine the distance from the laser to the ground, typically with a precision of 10 to 25 centimeters. Typical operating specifications permit flying speeds of 50 to 200 knots, flying heights of 100 to 5,000 meters, scanning angles up to plus 20 degrees, and pulse rates of 2,000 to more than 100,000 pulses per second. These parameters yield enough data points to create a highly accurate digital terrain model (DTM). Typical users of this technology have achieved accuracies of roughly 15 centimeters at up to 95 percent confidence interval vertically and 1 foot or 30 centimeters horizontally.
Post-flight processing combines precise aircraft trajectories developed from differential GPS solutions with the corrected laser ranging data and aircraft roll, pitch, and heading information. Integration of this data produces a precise horizontal position and vertical elevation for each laser pulse. Each data point can be identified by type, i.e. ground, vegetation, building, power line, or other object. Once classified, it is simple to manipulate data, remove layers of data points, and create DTMs.
Today, the entire process of airborne laser mapping is highly automated from flight planning to data acquisition to the generation of digital terrain models. Airborne laser mapping instruments are active sensor systems, as opposed to passive imagery such as cameras. Consequently, they offer unique capabilities and benefits compared to traditional photogrammetry.
The absolute accuracy of the elevation data is 15 centimeters; relative accuracy can be less than 5 cm. Absolute accuracy of the XY data is dependent on operating parameters such as flight altitude, but accuracy in the range of tens of centimeters to one meter can usually be achieved. The elevation data is generated at thousands of points per second, resulting in elevation point densities far greater than traditional ground survey methods. One hour of data collection can result in over 10,000,000 individually geo-referenced elevation points. With these high sampling rates, it is possible to rapidly complete a large topographic survey and still generate DTMs with a grid spacing of one meter or less.
But accuracy is not the only benefit of airborne laser mapping. The technology allows for extremely rapid rates of topographic data collection, resulting in exceptionally fast data delivery.
As the old commercials used to ask, "Now how much would you pay?" But wait … lidar outperforms traditional photogrammetry in these other areas:
- Lidar sensors can be operated in any weather and at low sun angles that would prevent an aerial photography survey.
- Rural and remote areas can be surveyed easily and quickly because each XYZ point is individually geo-referenced, and aerial triangulation or orthorectification of data is not required.
- Photogrammetric methods for DTM generation are very time consuming and labor intensive compared to airborne laser mapping.
- While satisfactory results in zones with limited contrast, such as coasts, beaches, and wetlands, are difficult to achieve with traditional photogrammetry, lidar genare traditional generally gives good results in these areas.
- Lidar performs well in forest areas where vegetation cover prevents visibility of the ground in aerial photographs.
- Lidar is better for road, pipeline, or power line planning for narrow corridor mapping.
- Lidar works well for open-pit mining operations where the final data is needed within a few hours of collection.
You might think that the myriad benefits of lidar would make it much more expensive, but studies have shown that lidar requires only 25 to 33 percent of the budget needed for photogrammetric compilation (see Petzold et al., 1999), possibly due to project delays caused by less-than-ideal environmental conditions or by the time-consuming, expensive processes required by traditional photogrammetry. With all the benefits and cost savings, the serious surveyor can't afford to not use aerial lidar.
About the Authors
Todd Stennett is the founder and CEO of Airborne 1 Corporation, based in El Segundo, California. The firm offers turnkey service, training for data and sensor operations, sensor rentals, and fractional sales of lidar assets.
Sandra Wade-Grusky, marketing manager for Airborne 1, is responsible for all internet and print marketing activities for the company.
» Back to our April 2008 Issue