MACHINE LEARNING ALGORITHM IN THE PREDICTION OF GEOMORPHIC INDICES FOR APPRAISAL THE INFLUENCE OF LANDSCAPE STRUCTURE ON FLUVIAL SYSTEMS, SOUTHEASTERN - BRAZIL

Abstract

The Abaeté hydrographic basin was influenced by sediments (Proterozoic and Cretaceous) and by volcanism (Upper Cretaceous). The objective was to identify whether the drainage configuration is due to structural factors or by neotectonics. We use geomorphic indices (Basin Asymmetry Factor, Stream Length-gradient Index - SL, and Channel Steepness Index - k_sn). We elaborate longitudinal profiles in the rivers of sub-basins. We apply the Random Forest (RF) Machine Learning algorithm in the prediction of the SL and k_sn indices, with the selection of relevant covariates. The RF was efficient in predicting, with better performance in the k_sn (R² 0.38), and indicating the areas of influence of the indices. The highest values of the indices are in zones of lithological contact with different resistances, favoring a sharp change in channel slope (knickpoints). In these zones, there is also a predominance of sub-basins tilted, and the longitudinal profiles of the rivers show uplift or subsidence. Therefore, structural factors conditioned the drainage of the basin.