The time-series land change detection data of the oil sands surface mineable area were derived from Landsat multispectral data. They contain 2 classes: 1 – land disturbance and 2 – vegetation recovery. These categories can be used as baseline data for planning, managing, and monitoring surface infrastructure needs and impacts.
The oil sands surface mineable area, Township 89 to 103, Range 3 to 14, west of the 4th Meridian, 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 ecosystem and environment. This is to be achieved through enhanced science-based monitoring for improved characterization of the environment and collection of the information necessary to understand cumulative effects.
The time-series are stacked to create a 6-band TIFF image with each band representing a change period.
The overall accuracy is defined as the total correct 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 as the total number of correct classified pixels in a class divided by total number of pixels in that class.
The User accuracy is defined as the total number of correct classified pixels in a class divided by 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 is used to find out if the accuracy level is significantly better than a random result, providing a better comparison of different classifications.
Land change detection results from 1985-1990, 1990-1995, 1995-2000, 2000-2005, 2005-2010, and 2010-2015.
Overall Accuracy = (1111297/1377514) 80.6741%
Kappa Coefficient = 0.7528
Year-to-Year |Producer Accuracy (%)|User Accuracy (%)|Producer Accuracy (Pixel Ratio)|User Accuracy (Pixel Ratio)|
1985-1990 | 61.38 | 72.96 | 25968/42309 | 25968/35591 |
1990-1995 | 73.98 | 58.71 | 59310/80175 | 59310/101026 |
1995-2000 | 78.75 | 91.73 | 259298/329261 | 259298/282660 |
2000-2005 | 84.07 | 74.79 | 255769/304229 | 255769/341996 |
2005-2010 | 89.57 | 75.95 | 332442/371166 | 332442/437731 |
2010-2015 | 71.30 | 100.00 | 178510/250374 | 178510/178510 |
Process steps performed in ENVI 5.1 to produce this land disturbance classification dataset:
1. Pre-release versions of annual Landsat Best Available Pixel Composite (LBAPC) datasets (1984 to 2012) with 30 m spatial resolution were obtained from the Natural Resources Canada Pacific Forestry Centre for testing purposes. Of these, the 1985, 1990, 1995, 2000, 2005, and 2010 datasets and in-house produced 2015 LBAPC datasets were used as inputs to produce this time-series land change detection stack.
2. Normalized Difference Built-up Index (NDBI) imageries were produced from 1985 to 2015 LBAPC datasets with 5-year gap.
3. Change detection results were produced by subtracting initial state NDBI (e.g., 1985) from the final state NDBI (e.g., 1990) for every 5-year gap from 1985 to 2015. The change detection result with NDBI difference from 0.2 to 1 was assigned as land disturbance and -1 to -0.2 was assigned as vegetation recovery. NDBI difference > -0.2 and < 0.2 represents very little to no changes and might contain noise from sensor artifacts and atmosphere.