Skip to content

Tech Deep Dive: Lidar (Part 1)

  • by

Welcome to our technology deep dive series. Over the next three blog posts we are going to explore one of the most exciting technologies used in forest management: Lidar 

In part 1 we’ll cover some background on Lidar, what it is, and the available platforms, part 2 will cover applications and finally part 3 will walk you through how to get started working with the technology including examples and available software. 

Let’s get started!

A canopy height model created from Lidar data

Lidar stands for light detection and ranging. It is analogous in function to radar, but uses pulses of light instead of radio waves. Like radar, Lidar’s primary function is to determine the distance from a scanner to an object. It accomplishes this by measuring the return time of a pulse from the scanner and back. Based on the travel time and the location of the scanner, the location of the object which returned the pulse can be determined. 

Most Lidar platforms combine GPS and inertial measurement to determine the relative location of a scanned object. This allows for highly accurate and highly detailed measurements. Most airborne lidar platforms have a point density of a few dozen returns per square meter. Terrestrial lidar has much higher point density. 

Besides the location, the strength of the return signal (Known as the intensity) can also be collected for each pulse. Different surfaces (Such as asphalt versus foliage) will return different intensity levels and the intensity data can be used to determine what type of surface was captured. Intensity data can be rendered as a sort of pseudo aerial image and is also useful for point classification and feature extraction. 

The most useful feature of Lidar for forestry is its ability to receive multiple returns from a single pulse. When a lidar pulse passes through a non solid object (Like a tree canopy) parts of the pulse are returned at different times. Each return represents a separate hit as the pulse moves from the top of the canopy, through the foliage and then finally to the ground. 

This ability to “see through” foliage makes Lidar indispensable for terrain mapping as it allows an accurate scan of terrain even when covered by trees. This incidental information is often discarded if the purpose of a Lidar mission is terrain scanning. But for forestry applications this by-product is pure gold. 

Lidar can accurately map the shape and structure of a tree canopy, including the height and density of vegetation. Various methods exist to build models to predict forest inventory information from Lidar data. We will cover these methods in a later part of this series (Be sure to sign up for our email list so you don’t miss future posts.) 

Lidar Platforms

There are three main ways Lidar data is collected: Airborne (plane or helicopter), terrestrial (handheld, or vehicle) and spaceborne. 

Airborne Lidar is most commonly used in forestry. The use of aircraft allows for large areas to be captured economically, relative to manual sampling. Additionally the accurate terrain data (Which comes in the form of a digital elevation model or DEM ) has a wide range of uses for forest managers. 

Lidar collection is less sensitive to weather conditions than aerial photography, however the two are often collected at the same time using the same aircraft. 

At a basic level attributes like the height of a forest canopy and crown closure can be calculated. 

A further step is to combine Lidar data with plot sampling to allow forest inventory metrics to be calculated. There are different methods to accomplish this, but they generally involve taking sample plots and then performing a regression analysis to map forest inventory metrics to the metrics of the Lidar data. This model can then be applied to the entire forest. 

Terrestrial Lidar is an emerging technology with many potential applications to forestry. This involves either handheld, stationary or vehicle mounted Lidar scanners that are moved through the forest, or used on specific plots. 

Because receiving a GPS signal through a forest canopy can be difficult, terrestrial Lidar often employs algorithms like SLAM which stands for “Simultaneous location and mapping”. This technology matches each scan of a Lidar device to the points collected on a previous scan so that the correct relative location of each return is retained, even when the scanner is in motion (Such as being moved by a person or on a vehicle). This allows for accurate information relative to, for example, a plot center, but is not capable of geolocating the data accurately. 

While airborne Lidar scans the forest from above, terrestrial Lidar scans from the side. This allows for far more detail with regards to crown and trunk form. Terrestrial Lidar can provide (At the plot level) far more accurate forest inventory metrics than airborne Lidar, or even manual cruising, since the exact shape of trees can be captured. 

The newest Lidar platform is spaceborne or satellite based Lidar. These systems are largely still in their infancy. While some spaceborne Lidar platforms are available, they are not designed nor well suited to forestry applications. Currently these platforms are constrained by the power requirements, which limits the resolution of the data and size of the acquisition missions.

All that being said, spaceborne Lidar has several advantages over airborne Lidar, such as more frequent (and less complex) data acquisition and the potential for global coverage. 

No matter the platform, Lidar scanners themselves are likely to become more cost effective in the coming years (especially terrestrial scanners). Additionally, unmanned aerial vehicles (UAVs) are being used more and more for aerial Lidar missions. This type of platform has distinct cost advantages over traditional fixed wing or helicopter acquisitions. 

Types of Lidar scans 

Two broad types of Lidar exist: topographic and bathymetric. Topographic uses a laser in the near infrared range and is used to map the land surface. Bathymetric lidar uses green lasers and can penetrate the water surface to map near shore terrain. Some scanners combine both types of lasers allowing for both bathymetric and topographic scans as part of the same mission. 

Lidar data is typically represented by a series of points. Each point represents a hit on a surface. The surface could be terrain, vegetation or any other object detected by the scanner. 

Data represented in this way is termed discrete return Lidar. Another more complex form is known as full waveform Lidar. 

Whenever a pulse is sent and received from a Lidar scanner, the return does not come back as a series of points, instead it comes back as varying intensity. The strength of this return can be represented as a wave, or the distribution of intensity from a return. 

In discrete return Lidar any point on the waveform above a certain minimum intensity is considered a return. All the other information associated with the distribution is discarded. 

Full waveform Lidar data is complex and the data sets are massive. Since each pulse has a return wave associated with it and there are usually millions (if not billions) of pulses used to scan even a small area. Even simply visualizing the data is difficult. 

However, full waveform offers an additional level of detail which, provided it can be processed, offers much more data than discrete return systems. Full waveform can provide more detailed information about the understory of a forest scan. The additional detail can also be useful in the identification of tree species, something that can be difficult with discrete return Lidar. 

That’s it for part 1, in part 2 coming up we’ll cover some of the most common applications of Lidar in forestry. Make sure you’re subscribed to our newsletter so you don’t miss any future posts.

Leave a Reply

Your email address will not be published. Required fields are marked *