News

Through presentations and a hands-on tutorial, participants in this workshop will become familiar with LiDAR technology, LiDAR data processing, and the steps involved in producing an enhanced forest inventory. In the hands-on tutorial, participants will learn how to use FUSION, a freely available software package, to both visualize and process LiDAR point clouds. Participates will calculate a digital elevation model (DEM) and a canopy height model (CHM) from a LiDAR point cloud, as well as calculate a range of metrics from LiDAR data that describe vegetation characteristics. Finally, participants will build linear models to predict forest attributes from these calculated LiDAR metrics. No previous experience with LiDAR or computer programming is required to participate in this workshop.

Location Date Registration Link
Vancouver Oct. 2, 2017 Waiting List
Quesnel Oct. 4, 2017 Full
Edmonton Oct. 19, 2017 Waiting List
Kapuskasing Oct. 23, 2017 https://www.eventbrite.ca/e/awarecwfc-lidar-workshop-kapuskasing-tickets-37742722499
Huntsville Oct. 25, 2017 https://www.eventbrite.ca/e/awarecwfc-lidar-workshop-huntsville-tickets-37742955195
Montreal Oct. 27, 2017 https://www.eventbrite.ca/e/billets-awarecwfc-lidar-workshop-montreal-37742969237
Fredericton Oct. 31, 2017 https://www.eventbrite.ca/e/awarecwfc-lidar-workshop-fredericton-tickets-37743003339
Corner Brook Nov. 3, 2017 https://www.eventbrite.ca/e/awarecwfc-lidar-workshop-corner-brook-tickets-38080233002

 

 

For more information, please contact us at forestry.aware@ubc.ca or call 604-822-1944.

As part of the AWARE project and in collaboration with JD Irving, a wall-to-wall map of individual tree species at the Black Brook Forest was completed in April of 2017. Led by Benoît St-Onge, a professor at UQAM and AWARE investigator, Rachel Perron (M.Sc. student, Q22), with assistance from Jean-François Prieur (Ph.D. student, Q21) developed special methods and software to analyse a huge airborne lidar dataset for predicting the species of every single tree visible from the air. This is no small feat as the Black Brook forest occupies 2090 km2 and is populated by more than 145 million trees. Despite there being a good number of scientific studies on the identification of trees from liDAR on small patches of forests, there is no evidence that such a large number of trees had ever characterized in an entirely automated manner.

The process involved the gathering of reference tree crown samples in the field from existing sample plots and plantation maps, and through expert photo-interpretation to train a machine learning classifier. In parallel, the airborne liDAR tiles were converted to canopy height models and single tree crowns were automatically delineated using the SEGMA application developed at Benoît St-Onge's lab, resulting in tens of millions of polygons. The lidar point clouds were then extracted for each crown polygon and stored in individual files. Classification features ("metrics") were computed for each crown based on the characteristics of the individual point cloud and stored in a PostGIS database. Features included three-dimensional metrics (e.g., the average slope of the crown profile, the crown "porosity" to laser pulses, etc.) and intensity metrics (e.g., average and standard deviation of intensity per crown, etc.). The reference crown samples were used to train random forest classifications based on all these features. Several classification levels were employed to classify trees as deciduous/coniferous, at the genus, and species levels. Specific classification models were developed to separate spruces from balsam firs. Based on a training set composed of more than 1 500 crowns, the accuracies were 49% at species level, 58% at genus level, and 93% for deciduous/coniferous separation.

Once validated, the classification models were applied to all the individual crowns. This resulted in more than 2000 map tiles, each 1 km x 1 km and containing on average 75 000 processed crowns, i.e. polygons for which the tree position, size (height, crown area, etc.), and predicted species at different levels are recorded (see figure below). The sheer volume of the input and output data represented a challenge in itself. Simple operations such as clipping liDAR points with the crown polygons just could not be done in a reasonable amount of time with most existing GIS applications. To get around this lack of functionality, Rachel Perron developed specific software approaches for optimizing each process and digesting the enormous amount of data. This required significant effort and time, but resulted in a unique, high capacity set of tools.

In addition to the error estimation performed on the reference crowns, JD Irving staff are currently checking the output species information in the field. The inspection reveals that the classification performed as expected in some situations, but may be less reliable in others. This type of "real world" verification is rarely reported in the scientific literature and can only be done if a wall-to-wall map was generated. What can be learned from this is, among other issues, that the selection of sample crowns for such a large territory may represent a greater challenge than expected. For example, for any given species, trees of different size class might have to be sampled in a systematic manner. Such a posteriori validation provides a clearer view of new and important research questions. In addition to sampling strategies, future research will consist of including contextual information such as drainage and soil type, as well as additional remote sensing data in the form of multispectral airborne images to improve the accuracy of species identification.

Figure - Individual tree crown polygons coloured by predicted species (overlaid on a color infrared airborne image).

Each morning, the group would split into two teams of three individuals, with each team’s goal of measuring trees on two sample plots that day. Guided by GPS coordinates and maps provided from West Fraser Integrated Forestry Company, each team would drive on dirt, gravel, or grass roads to get as close to the plot that they could. Once there, they would hike up to two kilometers in dense brush on unmarked terrain in order to find the plot. The plots were circular with a radius of 11.2 m. On the plots, trees were marked with blue paint and a numbered metal tag. The group took measurements on the on each tree that included the diameter, height, height to living crown, species, and class (ranging from dominant to suppressed). The same measurements were taken on the plots in either 2004 or 2006. By re-measuring the plots, researchers will be able to determine how each plot changed in the 12 or 14 years since it was last measured.

The team faced many challenges in taking their measurements. In some instances, they were unable to find the locations of the plots, which had the center marked by a 1m-tall orange stake. The most frequent cause of this was a large fire in 2011 around the town of Slave Lake that burned 40% of the city and many of the sample plots. The group also encountered hazardous conditions traveling to and from the plots each day, with the area’s terrain varying between wet, marshy areas and dry, dense brush. Researchers came across black bears and had to be prepared for encounters with other wildlife, as they were also in grizzly country. Despite these challenges, the team of researchers accomplished their goal of measuring a total of 40 plots of trees. At an average of 80+ trees per plot, the group estimates that they measured over 3,200 individual trees.

The measurements that the team collected will be used to verify models and estimates that are made from various remote sensing techniques, such as satellite imagery and LiDAR.

 

slavelakegroupphoto

The team (left to right) – Piotr Tompalski, Yu Chen, Yuhao Lu, Chris Mulverhill, Ignacio San Miguel, Brandon Bung (Photo: Piotr Tompalski).

 

slavelakeyu

Yu Chen sizes up a tree before taking measurements (Photo: Piotr Tompalski).

 

slavelaketree

On the plots, each tree was marked with paint and a numbered metal tag (Photo: Piotr Tompalski).

 

slavelakebean

Yuhao Lu uses a GPS unit to search for the plot. Many of the northern plots had been burned in 2011 (Photo: Ignacio San Miguel).

The PRF team: Dr Paul Treitz (Queen's), Joe Rakofsky (McGill), Karin van Ewijk (Queen's), Stacey Leson (UBC), Rachel Perron (UQAM) and Jean-François Prieur (Sherbrooke U./UQAM) (left to right). (Photo: Rachel Perron)

The PRF team: Dr Paul Treitz (Queen's), Joe Rakofsky (McGill), Karin van Ewijk (Queen's), Stacey Leson (UBC), Rachel Perron (UQAM) and Jean-François Prieur (Sherbrooke U./UQAM) (left to right). (Photo: Rachel Perron)

 

The PRF (45o 57’ N, 77o 34’ W) is situated along the Ottawa River, northwest of Ottawa, on the southern edge of the Precambrian Shield. The forest encompasses approximately 10,000 ha and borders Algonquin Park. Its topography has been influenced by glaciation and consists of post-glacial deposits and outwashing, resulting in gently rolling hills with sandy loams or shallow sandy soils with bedrock outcrops. The overstory of this mixed mature forest is characterized by eastern white, red, and jack pine on dry, nutrient-poor sites, and trembling aspen and white birch on upland sites. Tolerant hardwood species such as sugar and red maple are also abundant, especially on nutrient-rich uplands. Red oak, albeit in lower abundances, is found on upper, south- and west-facing slopes with shallow soils while shade-tolerant conifers, such as eastern hemlock, are more commonly found in valleys and on north- and east-facing slopes.

Due to the ongoing research activities at PRF, there is a large network of existing sampling plots which were utilized in this field campaign. In all, a network of 80 plots that encompass a range of species compositions and basal area conditions have been selected for sampling (to meet the needs of the AWARE research questions being investigated). In 2016, the AWARE field crew was able to sample 57 plots, leaving the remainder to be sampled in 2017. For each plot, we measured all trees with a diameter at breast height (dbh) equal to, or greater than 8 cm that fell within a 17.84 m radius from the plot centre (i.e., a sample plot size of 0.1 ha or 1000 m2). Tree measurements included species, status (living or dead), origin (natural or planted), dbh, crown class (e.g., dominant, co-dominant or suppressed), and health or decay class depending on whether the tree was alive or dead. At the plot level we recorded stand development stage, noted indications of disturbances, documented terrain conditions (sloping, wet, rocky, etc.) and took panoramic photographs at the centre of the plot which will be used to create photogrammetric point clouds. Using an SX Blue GPS system, we collected accurate and precise locations of each plot centre. Through a joint data sharing effort with NRCan (Lisa Venier) we will also have access to measurements for trees with a dbh less than 8 cm (i.e., dbh and density of trees within a number of height classes) collected in transects within the 57 sampling plots. Their data collection started in early August and is ongoing.

The field data collected over the month of July coincided with the acquisition of multispectral lidar data by Teledyne Optech’s multispectral lidar system (Titan).  These field data will facilitate the development of methods for modelling tree diameter distributions and tree species identification/classification of the overstory (and potentially the understory).

The team standing around the white pine with the largest dbh in PRF (Photo: Karin van Ewijk)

The team standing around the white pine with the largest dbh in PRF (Photo: Karin van Ewijk)


Joe Rakofsky and Stacey Leson at the truck after finishing a plot. (Photo: Karin van Ewijk)

Joe Rakofsky and Stacey Leson at the truck after finishing a plot. (Photo: Karin van Ewijk)


This survey was carried out to meet the research objectives of the AWARE project led by Nicholas Coops of UBC and funded by NSERC and industrial partners (http://aware.forestry.ubc.ca/). Survey specifications were designed by AWARE researchers Benoît St-Onge (UQAM) and Paul Treitz (Queen's), assisted by Peter Arbour (Operations Manager, PRF, Canadian Wood Fibre Centre) and Teledyne Optech. During the entire month of July, field measurements were collected by a team of AWARE researchers including postdoctoral fellow Karin van Ewijk (Queen's); graduate students Jean-François Prieur (Sherbrooke U./UQAM) and Rachel Perron (UQAM); field assistants Joseph Rakofsky (McGill) and Stacy Leson (UBC); and Paul Treitz. The lidar and field data will serve to quantify tree diameter distributions and identify tree species, among other research objectives.

Teledyne Optech Titan 3

Teledyne Optech Titan three channel false-color intensity image of a section of the Petawawa Research Forest (NIR in red, SWIR in green, and green channel in blue).

Subscribe to Our Newsletter – The Point Cloud






The first annual general meeting for AWARE will be held at the Forestry Centre, Grenfell Campus, Memorial University in Corner Brook, NL from May 24th-26th, 2016.  Over the three days, we will present the results of our first year of research, hear talks from local and international speakers, and run two technical workshops.

Please download the program for more information.

AWARE AGM Y1 Program v 1