A global framework to model spatial ecosystems exposure to home and personal care chemicals in Asia

Abstract This paper analyzes spatially ecosystem exposure to home and personal care (HPC) chemicals, accounting for market data and environmental processes in hydrological water networks, including multi-media fate and transport. We present a global modeling framework built on ScenAT (spatial scenarios of emission), SimpleTreat (sludge treatment plants), and Pangea (spatial multi-scale multimedia fate and transport of chemicals), that we apply across Asia to four chemicals selected to cover a variety of applications, volumes of production and emission, and physico-chemical and environmental fate properties: the anionic surfactant linear alkylbenzene sulphonate (LAS), the antimicrobial triclosan (TCS), the personal care preservative methyl paraben (MeP), and the emollient decamethylcyclopentasiloxane (D5). We present maps of predicted environmental concentrations (PECs) and compare them with monitored values. LAS emission levels and PECs are two to three orders of magnitude greater than for other substances, yet the literature about monitored levels of LAS in Asia is very limited. We observe a good agreement for TCS in freshwater (Pearson r = 0.82, for 253 monitored values covering 12 streams), a moderate agreement in general, and a significant model underestimation for MeP in sediments. While most differences could be explained by uncertainty in both chemical/hydrological parameters (DT50 water , DT50 sediments , K oc , f oc , TSS) and monitoring sites (e.g. spatial/temporal design), the underestimation of MeP concentrations in sediments may involve potential natural sources. We illustrate the relevance of local evaluations for short-lived substances in fresh water (LAS, MeP), and their inadequacy for substances with longer half-lives (TCS, D5). This framework constitutes a milestone towards higher tier exposure modeling approaches for identifying areas of higher chemical concentration, and linking large-scale fate modeling with (sub) catchment-scale ecological scenarios; a major limitation in model accuracy comes from the discrepancy between streams routed on a gridded, 0.5° × 0.5° global hydrological network and actual locations of streams and monitoring sites.