Happy November and congratulations to Rachel Perron (Q22-Species ID), who submitted her MSc report in September! She’s now just a few revisions away from graduation. I’d also like to mention that Murray Woods from the Ontario Ministry of Natural Resources and Forestry will be joining the CWFC under a one year Interchange Canada Agreement. Murray Woods, who was already providing his significant expertise to AWARE, will become even more involved with our research with the change. We look forward to his increased involvement.
Silvilaser is a very important conference for LiDAR and forestry, and AWARE was very much a part of Silvilaser this year. A large contingent of our researchers presented their work at Silvilaser 2017, including Kaysandra Waldron (Q2), Karin van Ewjik (Q4), Paul Treitz (Q4), Jean-Romain Roussel (Q8), Tristan Goodbody (Q11), Piotr Tompalski (Q13), Benoit St-One (species ID) and Jean-Francois Prieur (Q21). Kudos to all who presented at this conference!
LiDAR is not the only area that AWARE is researching. Aurelie Schmidt (Q17a- Ecosystem Services), who is also getting close to completion, presented her research at Conférence Vision Géomatique 2017 in October.
Bastien Vandendaele is a PhD candidate at the Université de Sherbrooke (Canada) and the Université de Liège – Gembloux Agro-Bio Tech (Belgium) under the supervision of Pr. Richard Fournier and Pr. Philippe Lejeune. His research focuses on UAV-based LiDAR. For his research in Canada, he is working on Q24 – Theme 4 within the AWARE project. For his Belgian research (beginning in September 2018), he will be working with the CIB-Olam forestry company within the DynafFor project (Dynamique des forêts d’Afrique centrale).
Bastien completed his M.Sc. in bio-engineering with a specialization in Forests and Natural Areas Management at the Université de Liège in 2015. During his M.Sc. program, he worked as an intern in the Département de la Nature et des Forêts in Belgium where he was introduced to GIS and contributed to the implementation of the new natural reserve area of Genappe. He was also involved in the IFSA (International Forestry Students’ Association) as a member of the Belgian representative committee. This experience allowed him to enrich his formal education by gaining practical experience and developing a wider perspective on forest management practices.
As a forest engineer, Bastien’s professional interests are focused on the enhancement of forest resource management. Having realized the strong potential of LiDAR technology for supporting forest inventory, Bastien chose to do his final M.Sc. internship at the Applied Geomatics Department at Université de Sherbrooke and at UQÀR to develop new skills in remote sensing and learn about boreal forest management. He spent 6 months under the supervision of Prs. Richard Fournier and Robert Schneider working on the use of terrestrial LiDAR for the estimation of structural forest attributes. From this experience, he developed a great interest in research related to forest environment and moved to Canada in 2016 to begin his Ph.D. For his research on Q24 in AWARE, Bastien works in collaboration with FPInnovations and the Northern Hardwoods Research Institute Inc. to investigate the potential of UAV-based LiDAR data for enhancing tree-level structural attributes estimation (i.e. stem diameter, height and merchantable volume). For his research within DynAfFor project, he aims to assess the potential of UAV-based LiDAR data to enhance allometric AGB models of tropical trees.
In his leisure time, you might find Bastien exploring the landscapes of Canada, climbing or, faithful to traditional habits in Belgium, enjoying a good beer with friends in a local microbrewery.
ALS-based estimates of structure in a mixedwood boreal forest
Chris Mulverhill is an MSc student working under the supervision of Nicholas Coops (UBC). His research project (Question 7 of theme 2) investigates the use of Airborne LiDAR (ALS) to predict stem size distributions (SSDs) across a boreal mixedwood forest near Slave Lake, AB.
The SSD, which shows the relative frequency of tree diameters on a given area (e.g. plot or stand), is a critical component of a forest inventory. SSDs are versatile and are useful for a range of timber production and ecological monitoring purposes, such as describing attributes such as structure, age, and volume, or model inputs to provide information for planning and management considerations. Landscape-level knowledge of SSDs provides a multidimensional insight below the forest canopy that might be missed by only using height, volume, or basal area alone to describe a stand. ALS provides an opportunity to meet this information need by being able to estimate stand-level SSDs over large areas, and understanding patterns of their variance across species and age gradients, when combined with other remotely sensed information.
SSDs are frequently summarized with a variety of statistical models or probability density functions (PDFs). These PDFs have two or more parameters that change the shape of the frequency distribution to characterize different sets of data. Parameters are fit to plot data by using an optimization technique such as maximum likelihood. Models for predicting each parameter value can be extrapolated across the landscape to provide an estimated SSD for any stand in the area. To do this, Chris uses plot measurements and ALS data collected in 2007 collected in the Slave Lake study area. To date, his research has focused on determining the most accurate techniques for estimating plot-level SSD for this study area, which features a wide range of tree species, ages, and stand structures. In addition to testing different techniques for optimizing parameters to fit plot data, Chris’s research also aims to determine the PDF that best suits the entire study area (Figure 1). His research has seen success in this process and will proceed by connecting the estimated SSDs to area-specific harvesting specifications to provide insight into the quantity and mix of harvested timber.
Figure 1 – Comparing fitted frequency distributions of ground-measured SSDs (top) to those estimated by ALS (bottom) for a variety of PDFs (colors) and fitting techniques (line type) on sample plots
Our next AGM will be held at FPInnovations Headquarters in Montreal on June 5-7th, 2018. We will use two days for AWARE researchers and industrial/government supporters only. The third day will highlight the research progress since the start of AWARE and will be open to researchers and industry outside of AWARE. More details will be released at the end of the year.