Authors: Mir Mustafizur Rahman*, University of Calgary, Maria Strack, University of Waterloo, Greg McDermid, University of Calgary
Topics: Land Use and Land Cover Change, Remote Sensing, Environment
Keywords: remote sensing, peatland, disturbance, land cover, image classification
Session Type: Paper
Start / End Time: 1:10 PM / 2:50 PM
Room: Forum Room, Omni, West
Presentation File: No File Uploaded
Peatlands store a large amount of soil carbon and exchange greenhouse gases (GHG) with the atmosphere. Roads can be extremely disruptive to low-lying peatland ecosystems, often triggering persistent changes in environmental factors such as land cover, microtopography, and hydrology, which may have implications on GHG release rates. We investigate the change in the land cover due to the construction of mineral roads in two peatland sites in Northern Alberta, Canada. To do so, we obtain multispectral satellite data and 3D structural information ~3 years before and ~ 3 years after the construction of the roads. We then separately classify the before images and the after images into representative land cover classes. The classified images are then compared to locate and quantify land cover changes. Our results revealed that (i) land covers are altered due to construction of road, (ii) impacts are maximum at the edges of the roads and then gradually decrease as distance from the road increase and no effects are observed beyond 40 m from the edge of the road, and (iii) when the road is placed along the general flow direction, the disturbance is similar on both sides of the road. However, when the road is placed across the general flow direction, disturbance on the upstream side is significantly higher than that on the downstream side. Our findings are expected to assist in controlling and guiding industrial activity in environmentally sensitive boreal peatlands.