This land-use and land-cover (LULC) classification dataset is derived from 2017 Sentinel-2 multispectral imagery (10 m spatial resolution) for the Cold Lake Oil Sands Area portion that falls within the Lower Athabasca Regional Plan (LARP) area. As part of Alberta’s Land-use Framework, LARP was developed in 2012 to set the stage for robust growth, vibrant communities, and a healthy environment within the region. One of its implementation objectives is to balance the economic development of oil sands and impacts on ecosystems and the environment. This objective is achieved through enhanced science-based monitoring to improve the characterization of landuse over time and understand cumulative effects on the environment.
This land classification raster dataset contains 11 classes: 0 - Unclassified, 1 - Shrubland, 2 - Grassland, 3 - Coniferous Forest, 4 - Deciduous Forest, 5 - Open Water, 6 - Wetland (Bog/Fen/Marsh/Swamp), 7 - Exposed - Barren Land, 8 - Agricultural Land, 9 - Developed Footprints, and 10 - Burned Areas - Little Biomass. These categories can be used as baseline data for planning, managing, and monitoring surface infrastructure needs and impacts.
The Overall accuracy is the total number of correctly classified pixels divided by the total number of pixels in the error matrix N.
The Producer accuracy is an indication of the probability of a reference pixel being correctly classified. It is calculated by dividing the number of correctly classified pixels in a class by the total number of pixels in that class.
The User accuracy is a ratio of the total number of correctly classified pixels in a class divided by the total number of pixels that were classified in that class. It indicates the probability that a pixel classified on the map represents that class on the ground.
The Kappa coefficient indicates if the accuracy level is significantly better than a random result, providing a better comparison of different classifications.
LULC Classification result of 2017
Overall Accuracy = (27433144/32478270) 84.4661%
Kappa Coefficient = 0.7917
Class|Producer Accuracy (%)|User Accuracy (%)|Producer Accuracy (Pixel Ratio)|User Accuracy (Pixel Ratio)
Shrubland|61.22|73.01|814617/1330695|814617/1115722
Grassland|83.45|34.20| 24839/29766| 24839/72638
Coniferous|80.97|64.67|4265333/5267639|4265333/6595290
Deciduous|86.84|86.20|10838133/12479920|10838133/12573497
Open Water|99.71|99.42|6043887/6061361|6043887/6079237
Wetland|73.78|93.78|5126639/6948196|5126639/5466589
Exposed Land|80.64|38.72|140535/174285|140535/362994
Developed|96.11|84.39|179161/186408|179161/212303
The accuracy assessment of the LULC classification result was conducted using the Digital Integrated Dispositions (DIDs), Alberta Merged Wetland Inventory (AMWI), Alberta Ground Cover Classification (AGCC), and Boreal Wetland Probability (BWP) datasets.
The Agricultural Land class was excluded from accuracy assessment since this class was attributed based on agricultural data sourced from Agriculture and Agri-Food Canada. In addition, the Burned Areas class was excluded from the accuracy assessment since this class was attributed by matching historical wildfire data sourced from the Government of Alberta that fall within the Exposed - Barren Land class.
A major component of this dataset is derived from Ducks Unlimited Canada's drained and boreal Enhanced Wetland Classification inventories.
Http://www.capf.ca/pdfs/AWCS%20Draft%20requesting%20Public%20Feedback.pdf
Process steps were performed in ENVI 5.5 software to produce the classification map:
1. A Sentinel-2 cloud-free composite dataset (10 m spatial resolution) was produced based on 2017 mid-summer imagery using the Google Earth Engine as an input to the land-use/land-cover (LULC) classification process.
2. Ground-reference datasets were created from DIDs, AMWI, AGCC, and BWP data.
3. A Maximum Likelihood classification algorithm was applied to the Sentinel-2 cloud-free composite dataset to produce LULC classifications for the Shrubland, Grassland, Coniferous Forest, Deciduous Forest, Open Water, Wetland (Bog/Fen/Marsh/Swamp), and Exposed - Barren Land classes.
4. The Developed Footprint class was produced using the Constraint Energy Minimization and Spectral Angle Mapper partial unmixing method followed by a K-means clustering. Post-classification techniques (i.e., majority filter, high-pass filter, clump, and sieve) were applied to all of the classes to refine the result and to minimize false detections. False detections (e.g., sandbars and fire scars misclassified as developed footprint) were further reduced by matching the DIDs areas that fall within this class.
5. The Agricultural Land class was attributed based on the agricultural data obtained from the Agriculture and Agri-Food Canada. In addition, Burned Areas class was attributed by matching historical wildfire data sourced from the Government of Alberta with the Exposed - Barren Land class.
6. Developed Areas class was refined using Normalized Difference Vegetation Index (NDVI) <0.3 derived from the 2017 SPOT-6 imagery (6 m spatial resolution).
Part of the process steps were adapted from our previous work published in the following journal paper.
Chowdhury, S., Chao, D.K., Shipman, T.C. and Wulder, M.A. (2017): Utilization of Landsat data to quantify land-use and land-cover changes related to oil and gas activities in west-central Alberta from 2005 to 2013: GIScience and Remote Sensing, p. 700-720, at http://dx.doi.org/10.1080/15481603.2017.1317453.