Backward particle tracking analysis for probabilistic contaminant source assessment

<p>Groundwater resources in the South of Portugal are under growing pressure due to an unbalanced water budget, mostly due to low natural recharge in recent years which, in some cases is leading to historically lowest observed groundwater levels. Additionally, groundwater quality issues have also been studied, mostly related to the use of fertilizers in agriculture, nonetheless, emerging contaminants such as pharmaceuticals are a growing risk and recent field campaigns have identified the occurrence of such substances in groundwater of South of Portugal. Moreover, the effects of climate change are expected to worsen the current situation. Modelling groundwater flow and contaminant transport with deterministic models is a practical tool to predict how groundwater might respond to climate change&#8217;s effects. However, to include uncertainty regarding highly complex geological heterogeneity units as well as input and outputs can be quite demanding and challenging with deterministic models. The stochastic approach is a promising technique that incorporates such uncertainties into the models. Markov chain transitions, a facies-based geostatistical approach represent a widely implemented stochastic approach for the generation of different scenarios with spatial distributions of hydraulic conductivity in heterogeneous aquifers. Particle tracking presents an alternative means of simulating the advective movement of a contaminant, offering the possibility of interpreting particle behavior. Backward particle tracking is computationally fast, numerically stable, and can reduce the uncertainties of source location estimates, optimizing the design of monitoring networks. In this work, a methodology consisting of a stochastic version of the backward particle tracking is applied to a multi-layered coastal aquifer in Faro (Algarve) to provide insight regarding the source of the pharmaceuticals detected in groundwater. To attain this task a three-dimensional flow and transport FEFLOW model for the study area (Campina de Faro aquifer system) was used.</p><p>KEYWORDS: Groundwater, Emerging contaminants, Pharmaceuticals, Portugal, Campina de Faro</p><p><br>ACKNOWLEDGMENTS<br>This work was supported by eGROUNDWATER &#8211; Citizen science and ICT-based enhanced information systems for groundwater assessment, modeling, and sustainable participatory management (GA n. 1921) and MIT Portugal Partnership 2030 (MPP2030-FCT) under the Doctoral Grant PRT/BD/153504/2021 Climate Science & Climate Change.</p>