Advances in Visual Applications: Visualizing & quantifying landscape change in SW Alberta using Mountain Legacy Project photography
By Mary Sanseverino with the Visual Applications Team
INTRODUCTION
In 2016 Mountain Legacy Project (MLP) scientists joined Landscapes in Motion – a collaborative research project whose goal is to improve the understanding of how, where, and when historical wildfires have occurred throughout Alberta’s southern Rocky Mountains. LiM’s work affords a better understanding of the forces at work on these landscapes and looks for answers to questions like these: Which forests had their beginnings in large, severe fires? Which have been shaped by less intense burns? How have they grown back?
Three teams form the core of LiM’s research – a Fire Regime team, a Spatial Modelling team, and an Oblique Photography team. Combined the teams go beyond what any could accomplish separately. Not only has MLP provided scientists and expertise to the Oblique Photography group, but its extensive collection of historic imagery – 1888 to the 1950s – is an excellent fit with the study area (Figures 1A and B). And LiM collaboration has encouraged MLP researchers to build out expertise in visualizing and quantifying landscape change within the images to even deeper levels.
Landscape in Motion Study Area
Figure 1A: The Landscapes in Motion project area covers nearly 20,000 square kilometres in the southern Rockies. From landscapesinmotion.ca
Mountain Legacy Project Data Points in LiM Study Area
Figure 1B: Between 400 – 500 stations with 2000 – 2500 historic images in the study area. Visit explore.mountainlegacy for image detail.
The LiM project provided both resources and collaboration that helped MLP researchers develop an end-to-end process for visualizing, classifying, and quantifying landscape change as seen in the images. Close collaboration, especially with the Fire Regime team, brought MLP focus onto both a region within the overall study area and a specific set of matching images.
The region in focus encompasses the upper Elbow and Sheep River watersheds. 17 historic images from 8 common station locations in this area were selected for in-depth analysis. In all, these 17 images, taken in 1895, 1897, 1916, and 1940, look out over approximately 350 sq km of land. Figure 2 gives an overview of the Stations, images, and Field of View (FoV) each covers. The Study Areas denoted on the Figure 2 map show part of the Fire Regime team’s research locations.
Figure 2: Selected MLP stations and images with approximate image FoV.
Map designed and created by James Tricker, PhD Candidate, UVic School of Environmental Studies, and MLP Research Associate.
This interactive web-based report details the end-to-end process and presents image visualizations and land cover quantification from each of the images.
METHODOLOGY reports on the web-based visualization and quantification tools created and developed in-house by MLP. Although currently used in the MLP/LiM research settings, these tools are under Creative Commons licence and are freely available for use by researchers, students, managers, and the general public.
LOCATIONS gives interactive access to images, data, and land cover masks and visualizations from each image.
RESOURCES gives access to papers, references, and bulk downloads of images, land cover masks, and quantification data.
__________________________________________________
Visual Applications Team
Eric Higgs
Mountain Legacy Project: Principal Investigator
Dr. Eric Higgs is a Professor in the School of Environmental Studies at the University of Victoria, and Past Chair of the Society for Ecological Restoration. He directs the Mountain Legacy Project, which over the last two decades has repeated more than 8,500 historical survey photographs, many of them in the southern Rockies. His work has helped to establish repeat photography methods which are used to track landscape change over a century or longer.
Julie Fortin
MLP Research Associate
Julie Fortin is a researcher with a passion for science, mountains and landscape dynamics. She is fascinated not only by how the Earth’s systems interact and change over time, but also by how we represent those systems and how we quantify that change. Much of her research involves remote sensing, change detection algorithms, and ecological modelling.
Mike Whitney
Software Developer & MLP Research Associate
Mike Whitney is responsible for the development of Mountain Legacy’s Image Analysis Toolkit (IAT). His commitment to creating efficient web-based tools that can be accessed by anyone with an Internet connection and run on cost-effective platforms has helped to let MLP’s vast collection of mountain imagery “speak” about change to researchers from all over.
Mary Sanseverino
MLP Research Associate
Mary Sanseverino is a Teaching Professor Emerita in the Department of Computer Science at UVic. A long time mountaineer and photographer, Mary’s research interests have involved work with computational photography, making her a good fit with the Mountain Legacy Project. She has been associated with the project since 2010, going out into the field in 2012, 2014, 2015 to 2018, and even in 2019 with a new left hip! She is the corresponding author for this report: msanseve@uvic.ca
With assistance from these MLP team members
Maria Catanzaro
Mountain Legacy Project Research Assistant
Maria worked at the detailed and time consuming task of preparing vegetation, height, and density masks.
James Tricker
PhD Candidate in UVic's School of Environmental Studies and MLP Research Associate
As well as making maps designed to communicate MLP/LiM collaboration, James provides invaluable GIS expertise.
Kristyn Lang
Mountain Legacy Project Research Assistant
Kristyn works on adding more vegetation land cover masks, creating Virtual Photos, and producing georeferenced viewsheds.
Maya Frederickson
MSc student at UVic's School of Environmental Studies and MLP Resesarch Associate
Maya’s MSc research focuses on the same geographic area as the LiM project. She has tested IAT and WSL-Monoplotting on several of the images used here.
__________________________________________________
Project Partners
