Assessment of Landscape Evolution through Pedo-Geomorphological Mapping and Predictive Classification Using Random Forest: A Case Study of the Statherian Natividade Basin, Central Brazil

Understanding the relationship between geological and geomorphological processes is essential for reconstructing landscape evolution. This study investigates the influence of geology and geomorphology on landscape development in central Brazil, specifically within the Natividade Group distribution area. To achieve this, we integrated remote sensing data (Sentinel-2 and SRTM imagery) with geospatial analyses to generate two key maps: (i) a pedogeomorphological map, which classifies landforms and soil-landscape relationships, and (ii) a predictive geological-geomorphological map, which models spatial patterns of geological and geomorphological influence using machine-learning techniques. The identified pedoforms were grouped into three major slope classes, each reflecting distinct relationships among lithology, relief, and soil development. This classification structure enhances the interpretation of landscape evolution by linking physical terrain characteristics to underlying geological controls. The pedogeomorpho-logical map identified nine distinct pedoforms, reflecting variations in drainage density and patterns, relief and lithology. The spatial distribution of these parameters indicates strong correlations between soil development, geomorphological processes, and the underlying geological framework. Pedoforms associated with resistant lithology, such as quartzite-rich metasedimentary rocks, are linked to shallow, poorly developed soils, especially related to the Natividade Group, whereas areas with phyllite, schist, and Paleoproterozoic basement rocks of the Almas and Aurumina Terranes exhibit deeper, chemically weathered soils. These findings highlight the role of soil formation as a key indicator of landscape evolution in tropical climate regions. The predictive mapping approach employed a Random Forest classifier trained with 15 environmental predictors derived from remote sensing datasets. This analysis resulted in six landscape classes, revealing the ongoing interaction between geology, geomorphology, and surface pro-cesses. While the predictive model effectively delineated geological-geomorphological units, its moderate accuracy suggests that additional variables, such as geophysical data, could enhance classification. The results confirm that the interplay between geology and geomorphology significantly influences landscape evolution, though other factors, such as climate and vegetation, also play crucial roles. Tectonic events, including the Rhyacian amalgamation (~2.2 Ga), Statherian taphrogenesis (~1.6 Ga), Neoproterozoic orogeny (~600 Ma), and the formation of the Sanfranciscana Basin (~100 Ma), have left imprints on the region’s current landscape. The combination of remote sensing techniques with ge-ological and geomorphological studies provides a comprehensive framework for un-derstanding landscape evolution and supports land-use planning, environmental con-servation, and risk assessment in geologically complex regions.