A Quantitative Provenance Analysis (QPA) Approach to Quantify Controls on Sediment Generation and Sediment Flux in the Upper Reaches of the Magdalena River (Colombia): 1. Natural and Anthropic Controls on the Sand Fraction

Abstract

AbstractThis study investigates the upper Magdalena river basin (Colombia), which is characterized by tectonic activity, tropical climate, heterogeneous lithological assemblages, and anthropic influence. It aims to comprehend factors controlling sediment generation, flux, and composition by incorporating petrographic data from 27 recent fluvial samples along with geochemistry, geomorphological parameters, historical precipitation, land cover, landsliding and suspended sediment load data. The analysis shows mismatches between sand composition and drainage lithology distribution. A new set of endmember mixing models is presented, quantifying the impacts of erodibility, sand generation potential (related to lithology), and sediment connectivity between hillslopes and river outlets (related to geomorphology). Results show enhanced sediment generation from sedimentary rocks (up to 40%) due to high erodibility, and decreased sediment flux from low‐ to medium‐grade metamorphic rocks (up to 60%) due to controlling morphometric parameters. Overrepresentation of plutonic and high‐grade metamorphic rocks and underrepresentation of volcanic rocks (up to 100%) are controlled by mineralogical and textural parameters. Enhanced landsliding activity in areas with volcanic activity occasionally overshadows basin‐wide compositional trends driven by morphological and lithological characteristics, as it controls rare overrepresentation of volcanic and low‐grade metamorphic detritus in the sand fraction. Sediment retention by hydroelectric dams significantly decreases suspended sediment flux by 30%, while deforestation plays a minor role in sediment flux. This study underscores the importance of coupling sediment generation, geomorphology and the distribution of stochastic events related to seismic activity in any sediment production‐focused study and for developing numerical models in Quantitative Provenance Analysis (QPA).