This imagery dataset is a TIFF format file containing a land-use and land-cover classification of the Upper Peace Region of Alberta, derived from 2017 Sentinel-2 multi-spectral imagery.
This land classification raster dataset contains 12 classes : 0 - Unclassified, 1 - Shrubland, 2 - Grassland, 3 - Coniferous Forest, 4 - Deciduous Forest, 5 - Open Water, 6 – Wetland (Bog/Fen/Marsh/Swamp), 7 - Exposed Land/Bare Soil, 8 - Cropland/Pasture, 9 - Developed Footprints, 10 - Mixed Developed Footprints, and 11 - Fire Scar.
The Overall accuracy is the total number of correctly classified pixels divided by the total number of pixels in the error matrix.
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:
Overall Accuracy = (214838551/260704666) 82.4069%
Kappa Coefficient = 0.7423
Class | Producer Accuracy (%) | User Accuracy (%) | Producer Accuracy (Pixel Ratio) | User Accuracy (Pixel Ratio)
Shrubland | 98.11 | 28.13 | 2999522/3057412 | 2999522/10662197
Grassland | 98.30 | 24.07 | 1368531/1392219 | 1368531/5684658
Coniferous | 76.13 | 93.84 | 78854469/103576555 | 78854469/84026610
Deciduous | 83.63 | 94.68 | 92212277/110256398 | 92212277/97395146
Open Water | 89.53 | 99.87 | 549824/614099 | 549824/550521
Wetland | 92.45 | 59.36 | 33605523/36350385 | 33605523/56610490
Exposed Land | 96.87 | 90.03 | 4754199/4907902 | 4754199/5280838
Developed | 89.91 | 100.00 | 494206/549696 | 494206/494206
The accuracy assessment of the LULC classification result was conducted using the Digital Integrated Dispositions (DIDs) and Alberta Merged Wetland Inventory (AMWI) data from the Government of Alberta, and the Alberta Biodiversity Monitoring Institute (ABMI) Wall-to-wall Land Cover datasets for 2000.
The Cropland/Pasture class was excluded from accuracy assessment since this class was attributed based on agricultural land data, sourced from Agriculture and Agri-Food Canada. In addition, the Fire Scar 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 Land/Bare Soil class. The Mixed Developed Footprints class was combined with the Exposed Land class for accuracy assessment.
Process steps were performed using ENVI 5.4 and GRASS GIS 7.2.2 software to produce the classification map:
1. A Sentinel-2 cloud-free composite dataset was produced using the Google Earth Engine as an input to the Land-Use/Land-Cover (LULC) classification process.
2. Ground-reference datasets were created from the Government of Alberta’s DIDs and AMWI data and the 2000 ABMI Wall-to-wall Land Cover data.
3. A Random Forest classification algorithm was applied to the Sentinel-2 cloud-free composite data to produce LULC classifications for the Shrubland, Grassland, Coniferous Forest, Deciduous Forest, Open Water, Wetland (Bog/Fen/Marsh/Swamp), and Exposed Land/Bare Soil 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 Cropland/Pasture class was attributed based on Normalized Difference Vegetation Index values > 0.7, contained in agricultural land data obtained from Agriculture and Agri-Food Canada. In addition, the Fire Scar class was attributed by matching historical wildfire data sourced from the Government of Alberta with the Exposed Land/Bare Soil class.
6. Classification maps were refined by keeping the Coniferous Forest, Deciduous Forest, and Wetland classes of the 2015 LULC classification included in our AER/AGS DIG 2018-0004 release consistent for the areas that remained unchanged, based on image-to-image changes extracted using the Normalized Difference Built-up Index from 2015 Landsat-8 data to 2017 Sentinel-2 data. In addition, the Developed Footprint class in the 2015 LULC classification that remained unchanged from 2015 to 2017 was re-classified as Mixed Developed Footprint in this LULC classification to differentiate it from the detailed Developed Footprint class generated from Sentinel-2 data.
These process steps were adapted from 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 & Remote Sensing, 54(5), p.700-720, at http://dx.doi.org/10.1080/15481603.2017.1317453.