Intra-storm variations in the contributions of geological formations to suspended sediment: a comparison between Bayesian and FingerPro sediment fingerprinting methods
摘要
Quantifying the contributions of erosion-prone geological formations represents a robust approach for identifying sediment sources via sediment fingerprinting techniques. Nevertheless, intra-event variations in these contributions remain poorly understood, particularly in watersheds of northern Iran. In this study, we adopt an innovative approach that integrates geochemical tracers and weathering indices to assess the dynamic influence of geological formations on sediment production during individual rainfall events. To achieve this, 23 suspended sediment samples were collected at the watershed outlet during rainfall events, along with 28 surface soil samples from different geological units identified as potential sediment sources. Geochemical elements in these samples were measured, and statistical analyses were performed using the FingerPro package in R software. The selected optimal tracers included terbium (Tb), dysprosium (Dy), yttrium (Y), cerium (Ce), and the weathering index SOC. The results indicated that, using a multivariate mixing model in FingerPro, the geological units TRJs (Shemshak Formation), PIc (Polymictic Conglomerate), and Q (Quaternary Formation) contributed 73%, 18%, and 9% to sampled sediment, respectively. Using a Bayesian model, the relative contributions were 76% for TRJs, 14% for PIc, and 10% for Q. Additionally, the specific sediment contribution (relative contribution per hectare) was determined to be 0.015%, 0.015%, and 0.095% for the TRJs, PIc, and Q units, respectively. These findings provide a better understanding of sediment dynamics and highlight the importance of management strategies based on geological characteristics to reduce erosion and protect water quality in vulnerable watersheds.