Water quality collapse in the Guaíba system

Context

In April–May 2024, southern Brazil’s worst flood on record, driven by synchronized waves from the Jacuí, Taquari, Caí, and Sinos rivers, pushed the Guaíba system to an all-time peak of 5.37 m in Porto Alegre. Infrastructure was overwhelmed for roughly 30 days; 2.3 million people were affected, 185 died, and five of six drinking-water treatment plants shut down, leaving 1.23 million residents without potable water.

This research project (internally named CLIMASANO, led by the sanitation group at IPH/UFRGS and funded by FAPERGS) set out to assess what the flood did to raw water quality in the Guaíba system. The team sampled 92 sites across the metropolitan area during the flood peak itself, collecting rare empirical data while the disaster was still unfolding. I joined as a one-year postdoc researcher, responsible for building the spatial database that organizes all analytical results.

Approach

The study covered an unusually broad analytical panel: physicochemical parameters (pH, conductivity, turbidity, color, suspended solids, phosphorus, ammoniacal nitrogen, dissolved organic carbon), microbiological indicators (total coliforms, E. coli), ecotoxicity assays (Daphnia magna), antimicrobial resistance genes (carbapenemase genes blaKPC and blaNDM), and pathogenic viruses (Hepatitis A and E, Norovirus GI and GII, SARS-CoV-2). Dozens of physicochemical and microbiological parameters were analyzed across the 92 individual samples, with advanced molecular and virological screening performed on 20 composite samples.

I designed a GeoPackage-based spatial database organizing these records for direct querying in QGIS, with source layers, a hexagonal binning grid for spatial aggregation, and filtered views by theme and campaign. A companion PDF report documents raw values and maps. The database is embargoed on Zenodo while new campaigns are incorporated; I’m still expanding it, with plans to migrate the grid to H3 indexing.

Outcomes

The main output is a paper in Scientific Reports (August 2026), “How water quality under extreme floods compromises ecosystem services, sanitation and One Health.” During peak flooding, turbidity exceeded 470 NTU, total phosphorus reached 21× the legal limit, and E. coli surpassed U.S. EPA recreational thresholds by more than an order of magnitude. Multidrug-resistant bacteria and carbapenemase genes were detected across flooded urban zones, and pathogenic viruses, including Hepatitis A, Norovirus, and SARS-CoV-2, were found in the floodwaters. The study provides one of the few datasets collected during a major flood peak, rather than after.

We also produced a dataset on the operational status of sanitation infrastructure during the disaster, compiled through Lei de Acesso à Informação requests to municipal governments, documenting how treatment, sewage, and supply systems performed across affected cities.

Resources:

• Albornoz, Possantti, et al. (2026): How water quality under extreme floods compromises ecosystem services, sanitation and One Health. Scientific Reports.
• CLIMASANO Database on Zenodo: Water quality GeoPackage database (embargoed)
• Sanitation Infrastructure Dataset on Zenodo: Operational status of sanitation systems during the 2024 disaster