LiDAR forest inventories

In order to continue innovating in our sector, at Ideas MedioAmbiental we are starting various forest inventory projects using LIDAR technology in order to be able to apply what we have learned in the training offered by Agresta S.Coop, this has highlighted the need to innovate in forest planning and especially in data collection processes in the forestry field to optimize efforts and resources, and explains the advantages and differences between the classic inventory and the LIDAR inventory. Making a point for those who are unaware of this technology, say that LiDAR (Light Detection And Ranging) is an active remote detection system based on a laser scanner, thus allowing us to determine the distance from a laser emitter to an object or surface using a pulsed laser beam. As with radar technology, where radio waves are used instead of light, the distance to the object is determined by measuring the delay time between the emission of the pulse and its detection through the reflected signal.

LIDAR flight scheme. Source: USDA

Returning to the forestry issue, it is known that the classic inventory has proven to be valid and effective, but that it consumes a lot of human resources that currently, and given the value of the forest resources of some mountains, are not available and are relegated to forest masses of high production value and quality. Accuracy in data is one of the most important characteristics of technology, together with the large amount of data generated in a LiDAR flight, which makes it an efficient technology that facilitates planning for all possible scales that allow inventorying forest areas that until now would be impossible to address within reasonable time and cost.

In addition to the above, it has the advantage of working with geographically extensive, continuous and highly accurate three-dimensional information, which means having an enormous volume of information on the structure of the forest. In addition, it allows obtaining high-precision digital terrain (MDT) and surface (MDS) models that provide additional information for planning work: detection of abandoned roads and streets, passable and non-passable areas for different forestry machinery, mapping of the type of land preparation in areas of reforestation (terraces, holes...), and finally, there is the possibility of reusing LiDAR information later for different purposes, which, unlike the classic inventory, would mean having to return to the field. Low costs per unit area, LiDAR will considerably reduce the costs of carrying out mapping and forest inventories; especially considering that flights can be combined (optical and LiDAR sensors working simultaneously on the same aircraft) and that the cost of the flight can be divided between the multiple applications it has.

It is also noteworthy that classic forest inventories (by systematic sampling, for example) measure a small percentage of the forest surface, so these forest inventories only provide information on that small area of the forest (sampling fraction) and is commonly referred to as sampling error. On the other hand, a LiDAR inventory provides measurements of the laser sensor on the entire work surface, so the sampling fraction would be 100% and therefore the sampling error disappears as understood in an inventory by systematic sampling.

The error in a LiDAR inventory is associated with the good fit of the regression that relates LiDAR measurements to forest inventory variables, namely, a LiDAR inventory requires two sampling phases: a first phase in which the auxiliary variable is measured over the entire surface and a second phase in which, in a relatively small number of plots, both the objective variable and the auxiliary variable are measured and correlated in a regression model. Agresta S.Coop offers us the main differences between a LiDAR inventory and a classic one, the main advantage of the LiDAR inventory compared to the “classic” inventory is that the errors in estimating the main mass variables in the stand, which is the management unit, are substantially improved. With classic inventories, good estimates are given in large surface units (barracks or strata) but poor in smaller ones (canton or stand). This advantage of LiDAR makes it possible to improve the planning of actions at the rodal scale and to apply flexible planning methods with more solvency, such as planning by stands, which respond much better to the demands of multifunctionality that society is making to our mountains.

TO DEAL WITH

“CLASSIC”

Volume of information

20,000 and 40,000 data on vegetation heights per ha

5-10 data on diameters and heights per ha

Sample fraction

100%

Around 1%

Mistakes

Lowers at stand level

Good on large surface levels (barracks, mountains, stratum)

Results

Difficulty in distinguishing species and diameter classes

Distinguish species and diameter classes

Costes

Low per unit area (especially working in large strata and with LiDAR information already captured)

Heights per unit of surface

In the following video you can see a plot where the number of feet and the amount of wood have been calculated

https://vimeo.com/11207734