An advection-reaction model of emerging contaminants in ground and surface waterbodies with known topology: a data-based approach

Method name: Data-driven advection-reaction modeling based on chemical monitoring data and network topology

A simple data-based advection-reaction (reactive transport) model applicable to both rivers and aquifers monitoring networks is proposed. It is built on (a) available monitoring data, and (b) graph-theoretical concepts, specifically making use of the Laplacian matrix to capture the network topology and the advection process. The method yields useful information regarding the dynamic spatial behavior of the variables monitored, expressed in terms of quantitative parameters like characteristic length, entropy, first-order decay constants, synchronization between sites, and the external inputs/outputs to the system. The model was tested in an unconfined shallow aquifer located in the lower Besòs River (Spain), in which 37 pharmaceutical compounds were monitored at 7 sites, alongside two campaigns (February and May 2021). Characteristic lengths were, on average, of the same order (24.5 m) as the mean distance between consecutive monitoring sites (33.6 m), thus reflecting an adequate monitoring network design. From an estimated mean advection velocity (0.24 m·h-1), first-order decay constants were calculated for each compound and campaign, with mean values of 0.025 h-1 (February) and 0.005 h-1 (May). Whereas entropy was generally slightly larger values in February than in May (mean values of 1.02 and 0.9 entropy units respectively), synchronization showed the opposite trend (mean values of 62.4% and 68.8% respectively). The input/output profiles were generally site-dependent, regardless of the compound, and campaign considered. • A new advection-reaction modeling approach directly based on experimental data obtained from monitoring campaigns together with the network topology is proposed. • The method yields new quantitative information regarding the dynamic behavior of the variables monitored, useful for both research and management purposes.

This research has been supported through the grant CEX2018-000794-S funded by MCIN/AEI/ 10.13039/501100011033.

With funding from the Spanish government through the "Severo Ochoa Centre of Excellence" accreditation (CEX2018-000794-S).

Peer reviewed