The Point Cloud, AWARE’s Electronic Newsletter
Vol. 4, Issue 1. Date: 9/6/2018
aware.forestry.ubc.ca Previous Issue
Not displaying correctly?
View this email in your browser
Welcome to the September edition of the Point Cloud. After four years of research, AWARE has now started the last of the 25 research questions in our mandate! We would like to welcome Xavier Gallagher-Duval, who will be working as an MSc student under Richard Fournier on Q16. Xavier is no stranger to some of our collaborators in Newfoundland, as he has already done two work terms on AWARE related topics for his undergraduate studies.
Also, congratulations to Shane Furze, who successfully defended his PhD thesis and will be heading to the Great Lakes Forestry Centre for a post-doctoral position this autumn.
This summer, AWARE collected data in three of our core sites. In New Brunswick, we teamed up with the Government of New Brunswick to enhance their aerial surveys of the Black Brook Forest. To provide better data for DAP applications, we have increased overlap to 80%. Additional field work is planned at Black Brook this September.In Alberta, a team of six researchers went to Slave Lake for two weeks to gather additional field data. This data will be used for Q7, Q12 and additional work on Q13 and Q5.
Last of all, in Newfoundland, Aurelie Schmidt spent the summer consulting with Resource Innovations. She provided her expertise from developing eco-system services frameworks (Q17a) to Resource Innovations, who are contractors for CBPP. Also in Corner Brook, Sam Herniman spent two weeks collecting habitat data for Q18. He has written up a bit more on his field work in this issue.
Abdelmounaime Safia is originally from Algeria where he studied geodesy and remote sensing at National Center for Space Techniques. There, he obtained his master’s (M.Sc) and his first doctoral degree. He moved to Canada to continue his research where he obtained a second Ph.D. in remote sensing at Sherbrooke University. For this PhD, he worked on improving the extraction of texture information in multiband images and developed a new texture model that characterizes texture of multiband images as set of intra-band spatial interactions (i.e. the classic grayscale texture) plus a set of inter-band spatial interactions. . This model is suitable for multispectral and hyperspectral images and can be tested using a standard multi-band texture database proposed by Abdelmounaime .
In 2016, Abdelmounaime joined AWARE to work on the application of Terrestrial LiDAR System (TLS) in forestry (Q19) at the individual tree (Theme 3) and plot levels. Based on the RDI (relative density index), he developed a new method to estimate 3D distributions of forest material. This been integrated into L-Vox for estimating 3D forest attributes at the individual tree level and plot level. In can also be used for ALS data. He is finishing three publications based on his work in AWARE.
Outside of his AWARE research, Abdelmounaime is also interested in applying his remote sensing knowledge around the world. He has done independent research and consulting on topics ranging from quantifying drivers of change in cultural parks in the Sahara to estimation of small-scale crop production at the village level in Mali.
In the coming months, Abdelmounaime aims to work on the use of remote sensing sensors (LiDAR, multispectral and hyperspectral) in UAV platforms for ultra-high resolution mapping. With the radical revolution on the why people nowadays are communicating, Abdelmounaime is interested on worldwide collaboration projects that needs alternative workplace that extends research horizons.
 A. Safia and D.-C. He, “Multiband compact texture unit descriptor for intra-band and inter-band texture analysi s,” ISPRS J. Photogramm. Remote Sens., 2014.
 S. Abdelmounaime and H. Dong-Chen, “New Brodatz-Based Image Databases for Grayscale Color and Multiband Texture A nalysis,” ISRN Mach. Vis., vol. 2013, pp. 1–14, 2013.
Estimating Stand Age from Airborne Laser Scanning Data to Improve Ecosite Based Models of Black Spruce Wood Quality in the Boreal Forest of Ontario
Rebecca Wylie is an MESc student working with her supervisor Jeff Dech at Nipissing University. Her research (Question 14 of Theme 2) aims to enhance forest resource inventories by creating
models that provide reliable estimates of wood quality and could enable value chain optimization approaches that consider the market potential of trees prior to harvest. Ecological land classification units (e.g. ecosite) and forest structural metrics derived from Airborne Laser Scanning (ALS) data have been shown to be useful predictors of wood quality. However, much of the variation in wood quality among trees is driven by differences in age, and this variation has been unaccounted for in models because age is poorly represented in most inventory systems.The objectives of this study were (i) to develop a model to predict mean stem age of black spruce-dominated stands across a representative boreal forest landscape, and (ii) refine models of black spruce wood quality by including age as a predictor variable. The study was conducted in the Hearst Forest in northern Ontario, and included age data from 116, 400m2 plots and wood quality data from 80, 400m2 plots representing a broad range of forest conditions. Plots were linked to a raster (20 x 20m) of ALS derived structural variables covering the entire forest. A stand age model was produced using a non-parametric approach that combined k Nearest Neighbour (k-NN) with random forests (RF) as the distance metric. This model performs well, with a root mean squared distance of 15 years and explained 62% of the variation (Figure 1). The wood density model used RF to produce a model that performed well predicting 13% of the variation within the sample population and had an RMSE of 59.1 kg/m3. Although the introduction of age into wood quality models did not improve performance, it did bring this type of large-scale wood quality prediction closer to becoming operational by accounting for changes across the entire stem associated to age.The non-parametric RF-k-NN age model developed in this study imputed accurate age measurements for black spruce dominated stands, that can be utilized as variables in FRI systems. These variables are an important tool to support sustainable forest management decisions. This type of indirect predictive modelling, creates measurements that are more efficient and consistent than traditional FRI methods, given that the technique applies the same rules to the entire land base. Descriptive wood quality maps (Figure 2) can be used as planning tools to optimize the value chain of forest timber harvesting by selecting specific wood qualities prior to harvest. The speed at which these wood quality variables can be calculated allows for enhanced forest inventories to be more reactive to economic demands than is possible with the long implementation cycles of traditional FRI’s.Figure 1. Imputed vs. observed RF-k-NN age prediction model, from mean plot ages (n=116) of black spruce dominated stands and ALS structural data in the Hearst forest located in the boreal region of Northeastern Ontario.
Figure 2. Predicted mean cell level wood density (kg/m3) map from random forest model based on ALS structural variables and wood density measurements from 109, 12mm increment cores in the Hearst forest of Northeastern Ontario.
NL Field Season Update (by Sam Herniman)
I just returned from Newfoundland where we have been conducting fieldwork for the creation of habitat suitability models for birds in Newfoundland (that’s question 18, if you’re wondering). Generally, we want to get an idea of what types of forest, different bird species are found in by comparing the location of each bird with the habitat information we can find with ALS data.The fieldwork involved Jenna McDermott, a masters student at Memorial University, and I (Sam Herniman), driving around the wilds of Newfoundland on ATVs for two weeks and conducting point counts. The people at Corner Brook Pulp and Paper helped us with most of this by supplying the ATVs and driving us between each survey area in the 900 square km watershed (slightly smaller than the area of Hong Kong) we were surveying. Thanks so much for your help, CBPP!
The process of a point count is very simple, but requires an expert ear (i.e. Jenna) to do accurately. We would stand in each forestry plot for 11 minutes and record the number of each bird species Jenna heard or saw. The hard bit, is waking up in the morning because these surveys can only be carried out between sunrise (about 5:00) and 10:00. Overall, it was a very successful trip in which we were able to conduct the number of point counts I had hoped for (31) and get a good idea of the bird diversity of the area.
Now that I am back, I can start creating models that show the relationship between the bird abundance data we collected and the habitat metrics we have calculated from the ALS acquisition. I’ll let you know how it goes!
Due to the late start of many of our research questions and limited opportunities for collaboration for some industrial partners, AWARE had a shortfall in in-kind contributions that resulted in a holdback of funding from NSERC for Y4. With the holdback of funding and the uncertainty of gaining enough in-kind support in the next two years to restore funding, AWARE has scaled back some of its research operations. As continuing to fund all our students and postdocs is our key objective for the remaining project, we have postponed the final Ontario and Alberta field seasons scheduled for Y5, and are putting on hold the proposed Alberta LiDAR survey until these issues have been resolved. We are also planning to reschedule the timing and location of the Y4 AGM that was planned for Alberta.
We continue to work with our industrial partners to strengthen collaborations and ensure our research outcomes are being utilised by our industrial support as much as possible . With sufficient in-kind support, we may be able to restore with-held funds and re-allocate funding to these previous planned activities.
AWARE AGM and Research Showcase
AWARE held its third annual general meeting in Montreal on June 5-7th. The first two days followed the same format as the past AGMs, with presentations to AWARE researchers, partners and government supporters. On the third day, AWARE opened our doors to the rest of the forest industry with a Research Showcase Day. During this day, we highlighted our research results of the last three years, had a poster session to allow our students to talk about their research to others one-on-one, and had breakout sessions to discuss emerging research areas. The session proved to be popular, with a good mix of industrial, government and academic professionals. AWARE will hold a similar event at the end of our project in 2020.