DATA-DRIVEN CHARACTERISATION OF LANDFORMS AND SUBSURFACE STRUCTURAL MAPPING USING REMOTE SENSING AND RANDOM FOREST IN THE DEMOCRATIC REPUBLIC OF CONGO

This study aimed to evaluate the effectiveness of the Random Forest (RF) algorithm and remotely sensed datasets in modelling landforms and subsurface geological structures in the Kolwezi Region of the Democratic Republic of Congo (DRC). Ten (10) controlling parameters (i.e., topographic, hydrologic, slope, geological) were selected for this study. Landforms and subsurface structures were modelled using RF and Aster dataset. The prediction performance of RF model was validated using the area under curve (AUC) and the receiver operating characteristic (ROC), respectively. The RF yield an AUC of 0.7819 and an accuracy of 78 %, confirming its effectiveness. The results reveal strong correlations between geomorphological features and subsurface structures, showing that RF is a reliable tool for understanding surface–subsurface interactions.