Abstract
Water quality monitoring is essential for managing watersheds; however, values below detection limits are often mishandled, which can hinder interpretation and lead to these measurements being ignored or discarded. This study aimed to analyze metal concentrations and their variability in freshwater rivers, focusing on datasets with a high proportion of censored values, where many measurements fell below the detection limits. This study was conducted in Hu’u, Sumbawa, Indonesia, from 2019 to 2023. Thirteen dissolved metals were analyzed: As, Ba, B, Cd, Cr(VI), Co, Cu, Fe, Pb, Mn, Hg, Se, and Zn. Significant differences between watersheds were observed for six metals, including As, Ba, B, Fe, Mn, and Zn (p < 0.05), whereas no significant seasonal effects were detected for any parameter. The findings indicate that the watersheds are likely impacted by nearby agricultural and domestic activities, as evidenced by the higher metal concentrations near settlements and cornfields. Conversely, more remote watersheds with limited agricultural exposure exhibited lower pollutant concentrations. This study practically enhances the representation of censored data concentrations with multiple detection limits, not only by visualizing censored values but also by illustrating data variability.