Skip to main content
Have a personal or library account? Click to login
Regional-Scale Analysis of Vegetation Dynamics Using Satellite Data and Machine Learning Algorithms: A Multi-Factorial Approach Cover

Regional-Scale Analysis of Vegetation Dynamics Using Satellite Data and Machine Learning Algorithms: A Multi-Factorial Approach

Open Access
|Oct 2023

Figures & Tables

Figure 1:

Overall flowchart of vegetation classification process in the GEE platform.

Figure 2:

Greater Sydney region, Australia.

Figure 3:

The schematic diagram of RF for season-based vegetation mapping.

Figure 4:

Confusion matrix used in the proposed RF model's training process: (a) with only spectral indices; and (b) all input variables. (i), (ii), and (iii) present the normalized confusion matrix for summer, autumn, and winter seasons, respectively.

Figure 5:

Visualization results of season-based vegetation mapping achieved by the proposed model for summer season: (a) original multi-temporal Sentinel-2 image, (b) results of RF+spectral indices, (c) results of RF+spectral indices+topographic factors, and (d) results of RF+spectral indices+topographic factors+texture information.

Figure 6:

Visualization results of season-based vegetation mapping achieved by the proposed model for autumn season: (a) original multi-temporal Sentinel-2 image, (b) results of RF+spectral indices, (c) results of RF+spectral indices+topographic factors, and (d) results of RF+spectral indices+topographic factors+texture information.

Figure 7:

Visualization results of season-based vegetation mapping achieved by the proposed model for winter season: (a) original multi-temporal Sentinel-2 image, (b) results of RF+spectral indices, (c) results of RF+spectral indices+topographic factors, and (d) results of RF+spectral indices+topographic factors+texture information.

Figure 8:

Input variable importance in season-based vegetation mapping achieved by the RF model for the Greater Sydney region: (a) for summer, (b) for autumn, and (c) for winter.

Area of pixels in each class for each season in square km_

ClassArea (km2)
SummerNon-vegetation1,364.3155
Grass1,691.4873
Trees9,225.3798
Crops86.8179

AutumnNon-vegetation1,335.9587
Grass1,537.0343
Trees9,440.5423
Crops54.4652

WinterNon-vegetation1,370.4691
Grass1,485.4252
Trees9,440.9971
Crops71.1091

Spatial and spectral resolutions of Sentinel-2 satellite data_

BandCentral wavelength (nm)Spatial resolution (m)
Coastal aerosol44360
Blue49010
Green56010
Red66510
Vegetation red edge70520
Vegetation red edge74020
Vegetation red edge78320
NIR84210
Vegetation red edge86520
Water vapor94560
SWIR-Cirrus1,38060
SWIR1,61020
SWIR2,19020

Quantitative results were achieved by the suggested RF method for season-based vegetation mapping_

Precision (%)Recall (%)F1 score (%)OA (%)Kappa (%)
SummerRF+spectral indicesNon-vegetation82.4591.2286.6190.6586.11
Grass91.5483.9687.59
Trees98.9397.6487.59
Crops67.0175.0170.78
RF+ spectral indices+topographic factorsNon-vegetation82.9991.3986.9991.2987.06
Grass92.8084.4688.43
Trees99.2497.9498.58
Crops67.8778.1572.65
RF+ spectral indices+topographic factors+texture informationNon-vegetation84.2792.0888.0092.5688.96
Grass93.5087.3190.30
Trees99.3798.2898.82
Crops74.9880.8977.82

AutumnRF+spectral indicesNon-vegetation81.4588.7284.9390.0885.27
Grass90.1083.5086.68
Trees98.9397.8398.38
Crops66.6173.6469.95
RF+ spectral indices+topographic factorsNon-vegetation80.7092.3086.1190.6086.06
Grass90.3383.9887.04
Trees98.9697.8598.40
Crops71.6973.4672.57
RF+ spectral indices+topographic factors+texture informationNon-vegetation81.8792.4986.8691.6487.60
Grass91.9785.7088.72
Trees99.4498.2198.83
Crops73.5076.5675.00

WinterRF+spectral indicesNon-vegetation81.8891.6186.4791.3587.17
Grass90.9984.7387.75
Trees99.2598.2998.77
Crops73.8477.1075.44
RF+ spectral indices+topographic factorsNon-vegetation82.6291.9387.0392.0888.27
Grass91.0588.0489.52
Trees99.3798.3798.87
Crops78.4675.8777.14
RF+ spectral indices+topographic factors+texture informationNon-vegetation83.3692.5787.7292.8989.46
Grass92.7886.9689.78
Trees99.5098.6399.06
Crops80.1682.9781.54

The number of input variables for the RF method used to create season-based vegetation maps_

CategoryDescriptionInput variables number
TopographicElevation, slope, aspect3
Spectral bandsBlue, green, red, vegetation red edge, near-infrared, and SWIR10
Spectral indicesNDVI, NDBI, MNDWI, NDTI4
Textural informationVariance, contrast, dissimilarity, homogeneity, correlation5×10
Total variable 67
Language: English
Submitted on: Jul 19, 2023
Published on: Oct 21, 2023
In partnership with: Paradigm Publishing Services
Publication frequency: 1 issue per year

© 2023 Abolfazl Abdollahi, Biswajeet Pradhan, Abdullah Alamri, published by Macquarie University, Australia
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.