An Open Science Early Information System for Forecasting Hypoxia in the Tidal Elbe

Manuscript (draft, prepared for the journal Environmental Management) with supplementary material on the software quality assessment. Low oxygen conditions (hypoxia) in water bodies threaten ecosystem health and ecosystem services. We present an open-source system that provides daily forecasts of dissolved oxygen and hypoxia up to seven days ahead in the tidal Elbe, specifically in the highly urbanized Hamburg Port area. The system integrates open data from local water quality and meteorological stations and weather forecasts, applies an auto-regressive model trained on historical observations, and delivers the results through an automated, documented web service. In a hindcast, hypoxia is forecast at the two downstream stations with an F1-Score of 0.87 one day and about 0.6 seven days ahead; skill is lower upstream. The software described is archived at 10.5281/zenodo.22917791. The deposit contains the typeset manuscript and supplement (PDF) and the source files of the manuscript without the third-party publisher templates.