Portable Laser System for Forest Structure
Understanding of the relationship between forest structure and function has been limited by our ability to measure structure. Most existing techniques for measuring the shape of the forest either cannot be used over a large area because they are very labor intensive or are not available to most scientists because they are very expensive. Therefore, the forest ecology lab set out to develop a new system that could rapidly generate structure measurements over ecologically meaningful spatial scales at a reasonable cost.
The resulting system consists of a laser rangefinder connected to a laptop computer for data acquisition mounted on a lightweight carrying frame. The laser rangefinder measures the distance to surfaces directly overhead up to 200 times per second. When the system is walked along a transect through the forest at a constant speed, the height measurements can be used to generate a large amount of information about the shape of the forest canopy along the transect path.
After some data processing, all of the individual height measurements collected along the transect can be used to generate a "slice" through the forest similar to images generated by a medical CT scan. The colored areas indicate where the leaves, branches and other canopy surfaces are located, although it is not possible to differentiate among the types of surfaces. The shade of the color is proportional to the density of material. Darker colors indicate more surfaces in that area. The four images above are from representative 30m transects in four forest patches around SERC. The young and intermediate forests are shorter with most of the surface area near the top of the canopy. In the taller mature and old growth forests, the surfaces are more evenly distributed throughout the canopy.
The height measurements from the transects can also be pooled into a mean profile for the forest. Because this system collects a large number of measurements rapidly, it is possible to get an estimate of the variability around the mean profile. This is not possible with the more labor intensive methods. The error bars are standard errors in the figures above.
When numerous closely-spaced parallel transects are walked, it is possible to create a three-dimensional view of the forest. The above images were generated from 31 parallel 30 meter transects that were 1 meter apart. The maximum heights measured by the laser are plotted to illustrate the outer surface of the canopy. The young forest is short and has a very smooth outer canopy. In the older forests, the outer surface is increasingly bumpy. We call the bumpiness of the outer surface "rugosity". A forest's rugosity generally increases with age and is proving to be correlated with other interesting forest functions.