Spring had turned to summer so quickly with AWARE researchers busy with field work and completing projects. Karen van Ewijk completed her AWARE projects and has now moved on to a new employment at Lim Geomatics. With her expertise on Lidar data processing and forest attribute modelling, Karen was hired as a Data Scientist. Congratulations Karin!
Catherine Frizzle is a PhD candidate under Richard Fournier (Applied Geomatics) and Mélanie Trudel (Hydrology) at Sherbrooke University. She completed her undergraduate studies in biology and her master in Environmental Studies both from Sherbrooke University. She is currently working on Question 17b building upon the framework developed from Q17a ecosystems services (ES), Aurélie Schmidt’s work, by developing ecological function indicators from LiDAR data.
Ecosystems Services (ES) are the benefits that humans derive directly or indirectly from ecosystems. Forests provide many ES, some of which may seem to conflict with timber harvesting, such as erosion control and water flow regulation. Q17b focuses on deriving ecological function indicator from LiDAR to support sustainable forest management. Topographic and hydrologic attributes are available from LiDAR data: DEM, hydrologic network, topographic wetness index, slope. Forest metrics also help to simulate the role of the forest in hydrological processes at the watershed scale. The Soil and Water Assessment Tools (SWAT) hydrologic model will be used to represent those processes and identify if the function indicators of erosion control and water flow regulation from LiDAR behave similarly to the SWAT outputs in the sub-basins of the Harry River watershed in Newfoundland.
Besides her PhD project, Catherine is highly involved in watershed integrated resource management. Working for a nonprofit watershed organization in Québec for the past 15 years, she lead many projects with stakeholders not always having the same management gold in the watershed, so she is fully aware of the potential outcomes of her PhD project in support of environmental reporting. She also teaches applied geomatics for integrated watershed management to graduate students and coaches supervised student projects. Her personal time is spent with her family enjoying life at its best.
Q9: Deriving hydrologic indicators from DTMs to support land classification schemes for forest management
Ayla Brombach is a Master’s student working with Dr. Jeff Dech at Nipissing University. Her research project, Question 9 of Theme 2, aims to derive hydrologic indicators from imagery and LiDAR-based digital terrain models (DTMs) to predict moisture regime classes for balsam fir and black spruce-dominated plots across the island of Newfoundland. Predicted moisture regimes will then be included in random forests simulations to estimate key average wood quality attributes such as density.
This study builds on recent work completed in the boreal forest of northern Ontario, which demonstrated the value of ecological land-classification (ELC) systems as indicators of wood quality attributes, and suggested that broad-scale patterns of moisture availability are linked to differences in wood attribute characteristics such as wood density and fibre length. Given that ELC systems are specific to a particular region, a more general approach would be to identify the mechanisms driving these patterns of wood attribute characteristics and then develop a classification scheme that captures the variation in this mechanistic driving force. Thus, the current study focuses on the prediction of soil moisture regime. The majority of Ayla’s work involved the use of image-based DTM’s to develop a depth to water model, which was used to create a moisture regime classification for the 2 x 2 km area surrounding each of 194 field plots.
The water depth model was derived from a two-step process, which included stochastic depression analysis (SDA) to determine the elevation of the surface water, and a kriging routine to interpolate the elevation of the water table over the map area. While developing this model Ayla has explored several variants to elucidate the best method for both the depression identification and interpolation. She first examined the optimum number of iterations and probability threshold for the SDA. After a review of the literature, she tested 10 to 500 iterations and thresholds of 0.55 and above. Ayla found that 50 iterations with a threshold of 0.60 produced best results from her dataset. For the interpolation, inverse distance weighting (IDW), thin-plate splines and several kriging methods were tested. Co-kriging with the original DTM elevations had the best results. Ongoing work will involve finalizing the moisture regime classification and testing its accuracy, subsequent wood attribute modelling, and comparison of the LiDar-based DTM depth to water predictions to the DNR imagery-based DTM.
Figure 1: Plot 19501709 (a) and Plot 19500108 (b) showing both the original DTM (16 km2) and interpolated water table. The interpolated water table follows DTM and intersects showing areas of surface water.
The AWARE e-Lecture series generally titled An Update on Canadian Forestry Applications of LIDAR and Digital Photogrammetry is here. The AWARE Experience will be running Wednesdays from September 25 to October 30, 2019, from: 1:30 p.m. – 2:30 p.m. EST / 2:30 p.m. – 3:30 p.m. AST / 10:30 a.m. – 11:30 a.m. PST. Registration to the e-lectures are for free and opens September 6, 2019 from the Canadian Institute of Forestry (CIF-IFC) website.
Save the date for the AWARE’s year 5 annual general meeting, which is set to take place at the University of Toronto’s John H. Daniel’s Faculty of Architecture, Landscape and Design building on Feb. 19-20, 2020. More details will be announced in the coming days.