3
datasets for mining activities. The datasets sourced from
S&P Global encompassed geographical coordinates and
detailed spatial boundaries of mining concessions under
study, enabling precise delineation of operational mining
areas. The extracted base polygons served as reference for
accurate geographical alignment of the satellite imagery
within GEE platform. This approach ensures minimization
of potential georeferencing errors characteristic of multi-
source geospatial data integration. The base polygons were
converted into GEE feature assets, enabling direct compu-
tational integration of custom mine site geometries. This
conversion was critical for localizing geospatial analysis,
facilitating precise spatial delineation and efficient process-
ing of the selected mine satellite imagery within the GEE
computational platform.
We compiled cobalt ore production data from the
United States Geological Survey (USGS) (USGS, 2024)
annual reports, focusing on mines within the Democratic
Republic of Congo’s (DRC) Cobalt- Copper Belt region.
The sample mines for the purpose of this study represented
5 of the region’s mining operations (Figure 1), selected
through a structured screening process focusing on data
completeness and relevance to the study.
The comprehensive annual production volumes from
2000 to 2023 were compiled. This production data was
relevant for establishing correlation between ore extrac-
tion intensity and land use impact. Quantification of veg-
etation health using the Normalized Difference Vegetation
Index (NDVI) was conducted. NDVI is a widely used
index derived from multispectral data and is calculated as a
normalized ratio between the red and near-infrared bands,
estimated according to Equation 1.
NDVI R
R600h
800 680
800 =+
-^R
^R h (1)
where R800 and R680 represents reflectance at 800 nm and
680 nm respectively.
Chlorophyll’s high near-infrared reflectance makes
NDVI a reliable indicator for assessing plant “greenness”
(Dong et al., 2019 Gago et al., 2015 Rasul et al., 2018
Vinod Janse et al., 2018). We calculated the mean and
standard deviation across the 23-year time series for each
mine location and layer. vegetation cover changes are used
as a proxy for environmental impact induced by mining
activities over time. The derived NDVI metrics provided a
quantitative framework for assessing spatial and temporal
vegetation trends within the Cobalt-Copper Belt region.
We employed data visualization techniques to plot NDVI
metrics (mean and standard deviation) and mine produc-
tion data, enabling intuitive interpretation of temporal
vegetation dynamics. Correlation coefficient analysis was
subsequently performed to quantitatively assess the strength
of relationship between ore production and vegetation
changes. This integrated approach allowed for statistical
identification of potential correlations between production
intensity and vegetation health across the Cobalt-Copper
Belt mining sites. By focusing on the selected mines and
modeling approach, we analyzed key environmental indi-
cators, such as vegetation loss and land cover changes, to
quantify biodiversity impacts.
Figure 1. The approximate locations of the Copper-Cobalt Belt in the Democratic Republic of Congo (DRC) and the selected
mines
datasets for mining activities. The datasets sourced from
S&P Global encompassed geographical coordinates and
detailed spatial boundaries of mining concessions under
study, enabling precise delineation of operational mining
areas. The extracted base polygons served as reference for
accurate geographical alignment of the satellite imagery
within GEE platform. This approach ensures minimization
of potential georeferencing errors characteristic of multi-
source geospatial data integration. The base polygons were
converted into GEE feature assets, enabling direct compu-
tational integration of custom mine site geometries. This
conversion was critical for localizing geospatial analysis,
facilitating precise spatial delineation and efficient process-
ing of the selected mine satellite imagery within the GEE
computational platform.
We compiled cobalt ore production data from the
United States Geological Survey (USGS) (USGS, 2024)
annual reports, focusing on mines within the Democratic
Republic of Congo’s (DRC) Cobalt- Copper Belt region.
The sample mines for the purpose of this study represented
5 of the region’s mining operations (Figure 1), selected
through a structured screening process focusing on data
completeness and relevance to the study.
The comprehensive annual production volumes from
2000 to 2023 were compiled. This production data was
relevant for establishing correlation between ore extrac-
tion intensity and land use impact. Quantification of veg-
etation health using the Normalized Difference Vegetation
Index (NDVI) was conducted. NDVI is a widely used
index derived from multispectral data and is calculated as a
normalized ratio between the red and near-infrared bands,
estimated according to Equation 1.
NDVI R
R600h
800 680
800 =+
-^R
^R h (1)
where R800 and R680 represents reflectance at 800 nm and
680 nm respectively.
Chlorophyll’s high near-infrared reflectance makes
NDVI a reliable indicator for assessing plant “greenness”
(Dong et al., 2019 Gago et al., 2015 Rasul et al., 2018
Vinod Janse et al., 2018). We calculated the mean and
standard deviation across the 23-year time series for each
mine location and layer. vegetation cover changes are used
as a proxy for environmental impact induced by mining
activities over time. The derived NDVI metrics provided a
quantitative framework for assessing spatial and temporal
vegetation trends within the Cobalt-Copper Belt region.
We employed data visualization techniques to plot NDVI
metrics (mean and standard deviation) and mine produc-
tion data, enabling intuitive interpretation of temporal
vegetation dynamics. Correlation coefficient analysis was
subsequently performed to quantitatively assess the strength
of relationship between ore production and vegetation
changes. This integrated approach allowed for statistical
identification of potential correlations between production
intensity and vegetation health across the Cobalt-Copper
Belt mining sites. By focusing on the selected mines and
modeling approach, we analyzed key environmental indi-
cators, such as vegetation loss and land cover changes, to
quantify biodiversity impacts.
Figure 1. The approximate locations of the Copper-Cobalt Belt in the Democratic Republic of Congo (DRC) and the selected
mines