Skip to main content

Table 3 Examples of application of clustering methods towards SDGs using EO data

From: Monitoring sustainable development by means of earth observation data and machine learning: a review

SDGs Field Main finding References
SDG 2 (Zero Hunger) Agriculture The proposed methodology based on K-Means and crop images, had a good performance estimating the rice yield [51]
SDG 7 (Affordable and Clean Energy) Renewable energy sources The choice of the clustering technique plays a crucial function in the forecasting of the gross wind power output [74]
SDG 9 (Industry, Innovation and Infrastructure) Mining The results showed that FCM was superior to K-Means and Self-Organising Map for mineral favourability mapping [90]
SDG 11 (Sustainable Cities and Communities) Land change The proposed approach based on K-Means, demonstrated better detection accuracies and visual performance for land cover and land change detection, compared to several methods [91]
Seismic The method analysed was reliable and effective in the identification of sequences of earthquakes [92]
Construction The proposed method used to segment individual buildings had a good performance with datasets acquired from densely built-up areas [76]
Land cover The proposed clustering method outperformed the original approach for remote sensing segmentation in land cover classificatio [93]
SDG 13 (Climate Action) Wildfires The presented algorithm for global burned area mapping was capable to adapt to different ecosystems and spatial resolution data [54]
Geomorphology The proposed DBSCAN methodology for geomorphological analysis allowed the detection of movements of a rock glacier [94]
Climate The techniques used such as K-Means and DBSCAN demonstrated their suitability for predicting climate types [53]
SDG 14 (Life Below Water) Sandbars The proposed algorithm demonstrated a high potential to be used for the extraction of sandbars positions [95]
SDG 15 (Life on Land) Soil degradation Assessment of spatial variability and mapping of soil properties provide an important link in identifying soil degradation spots [96]
Agriculture Optimised kernel-based FCM gave more accurate agriculture crop maps when compared with the classical FCM and K-Means [97]
SDG 17 (Partnerships for the Goals) Sustainability level The results obtained using Hierarchical Cluster Analysis showed that Sweden has the highest level of sustainability among the European countries; while, Greece, Bulgaria and Romania were the countries with the lowest performance [98]