Name
Utilizing vegetation indices to capture post-disturbance ecosystem recovery in boreal landscapes
Date & Time
Monday, May 8, 2023, 4:45 PM - 5:00 PM
Description
Pilot projects assessing the viability of boreal landscape reclamation are ongoing in the Athabasca oil sands region and are planned for future upscaling. Thus, feasible metrics for monitoring ecosystem development and function post-construction are required. Biophysical surface conditions are a critical component of ecohydrological feedbacks which govern ecosystem function. Capturing temporal and spatial changes at the surface improves our ability to assess ecosystem trajectory and better quantify surface-atmosphere exchanges of energy, water and carbon. We tested two commonly used vegetation indices (NDVI and EVI) using a Google Earth Engine script created by the authors and readily available Landsat 8 satellite imagery to capture the first seven years of development for a recently constructed fen-upland complex. Results indicate NDVI over-estimated and frequently misclassified wet and/or dark soils as dark vegetation, particularly in the period immediately following construction. Inclusion of the blue band in EVI post-processing was better suited to capture vegetation emergence and changes particularly in the constructed peatland. Additionally, the script was applied to nearby natural landscapes to provide context to the results from the reclaimed site. Furthermore, we found a strong, positive correlation between EVI, and eddy-covariance measured gross ecosystem productivity and albedo. Our results indicate EVI is a well-suited vegetation index for recovering, boreal landscapes in the region. Incorporation of remotely sensed vegetation indices following disturbance can be a helpful monitoring tool to capture spatial and temporal ecosystem evolution and function particularly when paired with field data.
Location Name
Cedar
Full Address
Banff Park Lodge Resort Hotel & Conference Centre
201 Lynx St
Banff AB T1L 1K5
Canada
Abstract
Pilot projects assessing the viability of boreal landscape reclamation are ongoing in the Athabasca oil sands region and are planned for future upscaling. Thus, feasible metrics for monitoring ecosystem development and function post-construction are required. Biophysical surface conditions are a critical component of ecohydrological feedbacks which govern ecosystem function. Capturing temporal and spatial changes at the surface improves our ability to assess ecosystem trajectory and better quantify surface-atmosphere exchanges of energy, water and carbon. We tested two commonly used vegetation indices (NDVI and EVI) using a Google Earth Engine script created by the authors and readily available Landsat 8 satellite imagery to capture the first seven years of development for a recently constructed fen-upland complex. Results indicate NDVI over-estimated and frequently misclassified wet and/or dark soils as dark vegetation, particularly in the period immediately following construction. Inclusion of the blue band in EVI post-processing was better suited to capture vegetation emergence and changes particularly in the constructed peatland. Additionally, the script was applied to nearby natural landscapes to provide context to the results from the reclaimed site. Furthermore, we found a strong, positive correlation between EVI, and eddy-covariance measured gross ecosystem productivity and albedo. Our results indicate EVI is a well-suited vegetation index for recovering, boreal landscapes in the region. Incorporation of remotely sensed vegetation indices following disturbance can be a helpful monitoring tool to capture spatial and temporal ecosystem evolution and function particularly when paired with field data.
Session Type
Breakout Session