Multiyear and seasonal wide-scale indicators for French surface waters contamination by WFD substances

This study offers an unprecedented valuation of the French surface waters WFD chemical monitoring dataset, covering 101 substances (metals, industrial and persistent organic pollutants (POPs), plant protection product (PPP) and biocides active substances, combustion residues) measured monthly on 4000 sites of the 6 main continental river basins, during 12 years (2009-2020). The concentration data were first made comparable through an original process removing the bias induced by the space-and-time heterogeneity of the monitoring labs performance, to gather a reference workable set of monthly contamination indicators. These were then used to display the substances' seasonal and interannual timeseries, revealing, e.g. the succession of PPP active substances contamination peaking periods in the 6 basins, or the long-term trends of the concentrations of the various chemicals, sometimes evidencing insufficiencies in the monitoring performance. These environmental observations were put in regard of the knowledge of the substances ban, restriction or reduction measures, to assess how streams' chemical quality responds to them. Additionally, the observed contamination features and their variations over the years are discussed in terms of changes in their usages, product substitution, emission sources, and linked to environmental processes like runoff, river dilution and physicochemical conditions. Some original findings and interpretation are provided on glyphosate and AMPA wide-scale data inter-relation, and some light is cast on the efficacy of the recent national policies restricting pesticides use in populated areas. For PPPs, the developed water contamination indicators were compared to tonnage data. We assessed their degree of linear relationship, which we propose to quantitatively express through a substance specific basin-to-river contamination coefficient. The interannual variations of this coefficient appear to be related to the changes in the water contamination seasonal patterns. We were also able to describe and validate the dependency of this coefficient on the molecular properties of the substances, conferring some capabilities for predicting the relative environmental risk induced by non-yet monitored compounds. We finally discuss the relevance of the developed indicators to complement the national chemical pollutants management system currently in place.