Mapping soil organic carbon stock through remote sensing tools for monitoring iron minelands under rehabilitation in the Amazon

The mining activity is fundamental for human development, but it requires changes in the environment, negatively affecting soil organic carbon (SOC) stock. The only way to restore the original SOC stock is via rehabilitation practices. Considering adopting more efficient techniques and technologies for the rehabilitation process, this study aimed to assess SOC stock in rehabilitating iron minelands and to use remote sensing tools to predict it. The study was conducted in two waste piles (WP-W and WP-S4) each presenting a chronosequence (initial, intermediate, and advanced stages) in addition to the forest (reference), all located in the Carajás Mineral Province, southeastern Amazon, Brazil. Results showed a gradual increase in the SOC stock for the WP-S4 from the initial to the advanced stage (average of 26.5 Mg ha−1). There were no differences among the rehabilitation stages for the WP-W. The difference in SOC stocks between the two waste piles is mainly related to the soil fertility status, indicating a higher SOC accumulation potential for the WP-S4. Principal component analysis revealed that the attributes presenting major influence on the SOC stock are K, SOC, cation exchange capacity, B, and total N (positive); and soil bulk density, pH, and sand (negative). The regression model results showed that the remote sensing indices Normalized Difference Vegetation Index (NDVI) and Bare Soil Index (BSI) are useful to derive digital maps of SOC stock and assist in the monitoring of the rehabilitation in the waste piles of the iron mineland.