From the Ground Up: LiDAR planning.
The decisions made in the planning process have a significant impact on the accuracy and usefulness of the elevation surface. These decisions must be made carefully given the inverse relationship between accuracy, point density and cost. Decisions that increase accuracy and point density also increase the cost of acquisition and post processing of the data. Like all good planning exercises, it is important to find a balance among these considerations that meets the requirements and provides the most bang for the project cost. Effective communication between the client and LiDAR provider is very important to the project’s success.
Significant diversity in projects seems to be the rule rather than the exception these days. Very few projects are alike in terms of the amount of relief, density of development, amount and type of vegetation (dense coniferous vegetation or marsh grasses provide significant challenges compared to leaf-off deciduous acquisition), time of year for the acquisition, required horizontal and vertical accuracies, nominal data postings or point-density requirements, and the required digital formats and downstream uses of the data.
Once the project requirements, challenges and downstream uses of the data are understood, the project planning can begin in earnest. A number of decisions regarding the operation of the sensor are critical to ensuring a successful project in terms of accuracy and posting requirements. However, the two decisions that generally have the most significant impact on both project accuracy and cost are the selection of the flying height and the field of view, which is the angular swing of the laser scan under the aircraft. Lower flying heights produce more-accurate elevation points on the ground and allow for increased point density. But lower flying heights also result in narrower swaths, which increase the acquisition time and therefore increase costs. Typical flying heights for 1-foot contour surfaces range from 3,500 to 4,500 feet. Similarly, heights for 2-foot surfaces range from 4,500 to 6,500 feet.
A similar discussion applies to the field of view. Many units in use today have maximum fields of view around 75 degrees. This is equivalent to a scan of 37.5 degrees each side of nadir (the point directly below the aircraft). Increased fields of view increase the swath width, which decreases the time for acquisition but also reduces the accuracy at increased angles and the ability to penetrate vegetation. Large fields of view also work against the collection of elevation models in highly urbanized areas due to more obstruction from buildings at increased scan angles. The maximum field of view is infrequently selected during data collection; most collection takes place with a full angle field of view between 20 and 45 degrees (10 to 22.5 degrees each side of nadir).
Once the decision has been made for the flying height and the field of view, decisions for the laser repetition rate, scan frequency and ground speed of the aircraft must be made. Most LiDAR units in use today have maximum laser repetition rates ranging from 150 to 200 kHz. These rates are equivalent to acquisitions ranging from 150,000 to 200,000 elevation points per second of flight. These maximum rates, however, are only achieved at low flying heights. Reduced laser rates are used with increased heights.
The scan rate is the speed at which the scan mirror is moved back and forth (directing the laser pulses from side to side below the aircraft) during acquisition. Increased laser rates and aircraft speeds require higher scan frequencies to “square up” the elevation points on the ground. For the typical project, it is highly desirable to have along-track and cross-track postings that are equal to one another. With typical aircraft ground speeds ranging from 90 to 150 knots or more, the scan mirror must be moved quickly to achieve a consistent ground pattern.
The GPS configuration during flight has a significant impact on the accuracy of the LiDAR elevation surface. Uncertainties in the 3D position of the sensor determined from GPS are one of the largest components in the error budget. During acquisition, the sensor operator should closely monitor the number of GPS satellites in view and the resulting dilution of precision (DOP). Lower numbers for the DOP indicate stronger satellite geometry and, therefore, increased positional accuracy. The planner should carefully balance the project’s accuracy requirements against the DOP during the flight window. Higher accuracies will generally correspond to lower maximum DOPs.
In the graphic on page 26, the maximum positional dilution of precision (PDOP) was set at 3.0. The area in red therefore indicates times that would be appropriate for LiDAR collection. Notice there are two times during the planned flight window where the PDOP is greater than 3.0. The first is fairly short in duration, and the plane might stay in the air and wait that one out. The second is fairly long and could provide an opportunity for a planned fuel stop and a break for the flight crew between collection sessions. Software allows the crews to look at each day’s GPS configuration to best determine flight times.
It is equally important to plan the locations for ground-based GPS occupations that will be later used in post processing the airborne GPS data captured by the LiDAR unit. Baseline lengths (distances from these ground units to the aircraft) should be reasonable to ensure the project’s accuracy requirements are met. For high-accuracy projects, the maximum baseline lengths could be as little as 10 to 15 miles. For projects with lower accuracy requirements, baseline lengths of 30 to 40 miles may suffice.
Airports often provide good options for GPS occupations. Control points here are generally located in controlled areas that provide opportunities for unattended occupations during flights. Often flight crews can set up these occupations prior to the day’s first mission and pick up the equipment after completing the collection for the day.
But airports have to be carefully evaluated. The runway length is important for safe takeoffs and landings. Lengths that are appropriate for small single-engine aircraft may not work for larger twin-engine platforms. Fuel considerations are also important. Planners need to consider the types of fuels available at the airport and the hours of operation at all potential locations. Night flights may be precluded if a smaller airport does not have fuel facilities that operate 24 hours a day.
The continuously operating reference system (CORS) often provides options for post processing the airborne data. Many states have dense CORS networks that collect data at intervals that are appropriate for processing airborne data. CORS could be used as a primary base option during processing or serve as important backup data in case something goes wrong with the primary collection.
Often, additional ground GPS occupations are necessary beyond the CORS or airport occupation options. Since GPS data collection begins before the flight mission and ends afterward, the logistics of getting personnel to these designated locations can be challenging. These challenges are magnified because communication between the air and ground can be difficult-if not impossible-during flight operations on small aircraft.
Unlike aerial photography, LiDAR is an active sensor and can acquire data 24 hours a day. As mentioned earlier, it is important to look at the GPS configuration to determine appropriate flight times. But other factors can be equally important.
Commercial air traffic can be a big consideration in the planning for acquisition in urban areas. My firm has completed projects in the immediate vicinity of Atlanta Hartsfield, Los Angeles International and the trio of major airports in the New York City area (Newark, LaGuardia and JFK) that would have been nearly impossible to acquire during daylight hours due to the high number of commercial flights. There are normally periods of significantly reduced air traffic during the nighttime hours that can be much easier for acquisition. Working with the air traffic controllers in these urban areas is crucial in the planning process.
Weather can also play an important role in flight-time planning. In some locations, night conditions are generally better suited for collection. For example, many parts of Florida have low clouds during most of the day that often dissipate after dark. LiDAR acquisition can take place below clouds but not above. Moreover, the winds can be lighter with overall less turbulent conditions during nighttime flying. All of these factors can lead to opportunities for night flying.
Planning is critical to the success of all projects. It is important for planners to have a clear understanding of client needs and expectations as well as a strong background in LiDAR. They should also possess a full understanding of how their sensor functions in different project environments and how the many options for acquisition affect the accuracy and cost of projects. A little extra effort upfront provides big rewards in the overall project.