A seasonally explicit counterfactual framework reveals incomplete turbidity recovery in a tropical river seven years after a major mine-tailings disaster

Abstract River channels have been extensively impacted by tailing inflows worldwide. Evaluating the environmental impacts of tailing dam failures using methodologically rigorous, time-series–based approaches is an essential step in supporting recommendations for the design of long-term monitoring and management actions. In southeastern Brazil, the 2015 Fundão tailings dam failure discharged approximately 43 million m 3 of iron ore tailings into the Doce River. Several studies have shown that turbidity is the most conservative metric for assessing water quality recovery in the Doce River. Here, we analyzed a 26-year turbidity record from 13 monitoring stations along the upper and middle reaches of the Doce River to quantify and characterize the response to the Fundão tailings dam failure seven years after the event. To do this, we used an integrated framework for trend analysis of environmental monitoring data, combining seasonal decomposition, interrupted time-series models with counterfactual “no-impact” scenarios, upstream–downstream contrasts, and standardized effect sizes. From a more conservative perspective, we showed that turbidity recovery was not complete by 2022; it remained partial, largely confined to the dry season and the furthest downstream portion of the evaluated reach of the Doce River. We argue that post-failure evaluation in the Doce River, and likely in other tailings-impacted rivers, should maintain high-frequency, long-term measurements, with particular emphasis on the rainy months and stations closest to the mine tailings spill site.