Best use of Terrain Lidar for Vegetation

Commercially available airborne Lidar systems (sometimes called Airborne Laser Scanners, ALS) may be used to acquire data from which useful forestry parameters can be derived. Such instruments record the range to the first and/or last return signal (above a threshold) and some systems record up to 3 additional ranges. They are designed to identify the position of the ground, which is an important first step in measuring the properties of the forest (eg. the height of the canopy).  Methods of classifying Lidar returns as ground or non-ground hits are well established as is the process of using the resulting point data to generate a digital elevation model (DEM).

The following images are examples of DEMs for four forested areas within an established study site near Tumbarumba, NSW. The colours overlaid on the elevation mesh indicate the total fractional cover of vegetation inferred from the number of non-ground Lidar hits.

 

 

Several features are immediately evident from these images. The Ponderosa Pine site is an area of mature plantation and is therefore quite uniform.  The dark areas (low cover) in this image were clear felled. The boundary road between the two areas is clearly seen as is the location of short road providing access into the plantation. The other three areas are native forest and woodland and therefore display much less uniformity. The Snow Gum woodland is the most sparse. The Blue Gum site is on a steep slope and the location of a road traversing the slope is well defined in the DEM.

 

More advanced products derived from ALS data make use of the penetration of the laser to different levels within the canopy. The number of returns at each height gives a measure of the amount of foliage present. A plot of foliage amount versus height is called a foliage profile and can deliver important information about the canopy.

 

 

Identification of multiple canopy layers is a direct outcome of the foliage profile and there are many valuable uses of this information. For example, the presence of a thick understorey will affect trafficability into the area and may also pose an increased fire risk. The presence of multiple canopy layers is important for wildlife (particularly bird) habitat and is well established as an index of biodiversity. Alternative methods of measuring foliage profiles are laborious and often inaccurate due to inadequate sampling.  Lidar provides the capacity to measure foliage profiles over large areas quickly, repeatably and accurately located.

 

Predominant or top height is a common measure of canopy height within a given area. It is an average height of a number of the tallest trees in an area. An analagous measure can be made with airborne Lidar data that achieves very similar results. The following graph shows the comparison for the four study sites (PP: Ponderosa Pine, SG: Snow Gum, AA: Alpine Ash, BG: Blue Gum) with Lidar values and uncertainty shown in blue and field measured values and uncertainty in red.

While it is possible to obtain useful forest measurement from ALS data, there are limitations. An airborne instrument can only sample the vertically projected foliage area, which is an under-estimate of the true foliage area. This can be taken into account if the foliage angle distribution is measured or estimated. The measurement of foliage angle is possible with a ground-based scanning system such as ECHIDNA™.

 

The range-finding nature of ALS systems is also a limiting factor in determination of forest structure. In order to measure vegetation cover and foliage profiles, it is necessary to collect data from many laser shots, thus the data volume is large. The selection of (usually) only first and last returns means that the middle parts of the canopy are poorly sampled. The finite size of the laser beam (several centimetres in diameter) is also a limitation when no information is available to indicate how much of the beam was reflected. These limitations can be overcome by using an instrument that records the calibrated intensity of all of the returns (a Lidar waveform) from the canopy– an airborne canopy Lidar.

 

Ground-based scanning Lidar - ECHIDNA™

 

A ground-based scanning Lidar system samples aspects of the canopy that are not possible with an airborne system. The vertical view of any airborne system results in the sampling of vertically projected foliage area and insensitivity to trunks. A multi-angular scanning system samples the foliage angle distribution and this information can be used in combination with data from airborne systems to provide actual (rather than vertically projected) foliage profiles. In this way, detailed structural measurements can be made from the ground in representative plots and scaled up to a larger area with an airborne system.

The instrument is a ground based and flexibly specified sounding system. Because it sounds in multiple directions, it has the evocative name of “ECHIDNAä” and is aimed to provide a unique stand-alone forest measurement technology as well as provide a base for investigations of laser and Lidar system performance. It is likely that the instrument and variants of it have their own very extensive market as field tools for foresters. The opportunity of using the ECHIDNAä in the validation of airborne (or spaceborne) data is another market area of some consequence.

A Rangefinder Prototype ECHIDNA™ has been.assembled from off the shelf components (a simple laser rangefinder and computer-controlled pan-tilt platform). Experiments with this instrument have provided base-level validation of algorithms and input into the design specifications for a more sophisticated instrument.