Continuously monitoring and managing manganese (Mn) concentrations in drinking water supply reservoirs are paramount for water suppliers since high soluble Mn loads lead to discoloration of potable water. Traditional Mn management approaches involve water samplings and laboratory analyses typically on a weekly basis; if critical thresholds are exceeded, then appropriate treatment procedures are exploited. Despite the Mn level currently being manually sampled throughout the year, in subtropical monomictic lakes such as Hinze dam, critical Mn concentrations in the epilimnion, where the water is drawn, are typically recorded only during winter during lake circulation. Vertical profiling system (VPS) installed can continuously collect physical parameters such as water temperature, pH or dissolved oxygen, which determine the transport process of Mn in the lake. Therefore, a long-term historical database gives opportunities for the development of a data driven prediction model to autonomously forecast future Mn concentration values.

In the present study, VPS and samplings data were collected and analysed, and prediction models applying nonlinear regression techniques and data-driven equations were developed and assessed; they were able to forecast future Mn concentrations from 1 to 7 days ahead with correlation coefficients higher than 0.83 on an independent test dataset. Importantly, the peak concentrations in the epilimnion during the lake destratification were accurately predicted. The model also displays the probabilities of the Mn to exceed certain key-thresholds, thus assisting operators in Mn treatment decision-making. Such a tool is very beneficial for the water supplier, since costly and time consuming water samplings for monitoring Mn concentrations can be avoided, thus relying only on the real time VPS-based model outputs.