Updated: Sep 12, 2018
Taking an Interest in the Area of Interest:
You’re going to get a great data set when you use LiDAR, but you need to understand what the data set is going to be used for — oftentimes the terrain and vegetation in question will dictate the necessity for a higher data point. As mentioned before, the higher the number of points within the data set, the better the definition of the data set is going to be — and for that matter, the accuracy of the data set plays into the picture as well.
Consider, for example, the flood plain maps that the Federal Emergency Management Agency (FEMA) creates using LiDAR. Typically, this kind of mapping project will need a point sample spacing of 1.4 meters, which means that you’ll be mapping a point roughly every 1.4 meters on average. This is a large area, and you’re typically going to need to achieve an accuracy of about 0.5 meters horizontally and 15 to 18.5 centimeters vertically. You’ll also achieve a 2-foot contour specification.
But in some cases, terrain and vegetation will be such that you’ll have to collect data with a higher point density in order to achieve the required accuracy specifications. For example, electric utility companies and the engineering firms that work for them usually require between 20 to 40 points per meter in order to properly map power lines.
These kinds of collections are typically done with helicopters, though it is becoming increasingly common to work with fixed-wing aircraft (planes), depending on the accuracy requirements. Transportation engineers often require engineering-grade information. Mobile mapping LiDAR can achieve this, provided that adequate ground control points within the project are also utilized for high accuracy data calibration.
What kind of eyes does a LiDAR have? The collection sensor uses a powerful laser that includes a transceiver and receiver, along with a geodetic-quality global positioning system (GPS) receiver, an inertial measurement unit (IMU), and a scanner. The laser typically operates at 532 to 1550 nm on the light spectrum, but this varies depending on whether you’re operating an airborne, terrestrial, or mobile mapping system.
The transceiver emits the LiDAR pulses, and return pulses are picked up by the receiver. The GPS reports the location of the platform of the sensor, such as where the aircraft is that’s car-rying an airborne system. The IMU measures the attitude of the sensor on its platform — what is known as the roll, pitch, and heading of the platform. Typically, a terrestrial scanner doesn’t have an IMU (find out more about the collection process for this type of scanner in the next chapter). A mirror attached to the scanner spreads the pulses across the surface the system is mapping.
See the forest for the trees
LiDAR can essentially “see through” vegetation to the same extent that humans can — for instance, when no foliage is present. If the point sample spacing isn’t high enough or if the vegetation is very thick, odds are lower that the LiDAR system will be able to emit pulses through the vegetation.
The best way to map through vegetation is to do it when there are leaf-off conditions. The fewer leaves on the trees, the better the representation of the ground under the trees will be. If you’re mapping an area of evergreen conifers, you’re not going to get a leaf-off condition, so you’ll need to greatly increase the concentration of points to have a better chance of getting to the ground. In cases of vegetation mapping, the typical point sample density will be 8 to 12 points per meter. In areas of light vegetation, the sample density can be less because the points will better represent the ground.
Flat as a pancake?
The flatter the surface in an area of interest, the easier it is to map and the smaller the number of points you’ll need to define that area. That’s fine if there’s not too much vegetation, but if there is, the LiDAR system will have trouble getting to the ground and more points will be needed.
Areas with higher relief — that is, more fluctuations in elevation — are more difficult to map, for a number of reasons. For one thing, high relief areas often tend to be high vegetation areas. Beyond that, the fluctuations in elevation mean the point spacing will change based on the distance to the ground from the LiDAR sensor. To adjust for this, you’ll have to fly these areas with more overlap between flight lines. That allows you to sample the areas twice from two different angles. This also increases the ability to get through the vegetation.
Mapping (or ignoring) those man-made features
LiDAR systems are very good at mapping man-made features, but as with other scenarios, you’ll need to tailor the LiDAR collection parameters to the application. If you want to map the man-made features, you’ll need to use a different collection process than you would if you want to remove them.
Young, James. LiDAR for Dummies. Hoboken: Wiley Publishing, Inc., 2011.