Département Eaux de surface

Bloom-forming algae present a unique challenge to water managers as they can significantly impair provision of important ecosystem services and cause health risks to humans and animals. Consequently, effective short-term algae forecasts are important as they provide early warnings and enable implementation of mitigation strategies. In this context, machine learning (ML) emerges as a promising forecasting tool. However, the performance of ML models is heavily dependent on the availability of appropriate training data. Consequently, it is essential to determine the volume of data necessary to develop reliable ML forecasts. Understanding this will guide future monitoring strategies, optimize resource allocation, and set realistic expectations for management outcomes. In this study, we used 30 years of fortnightly measurements of 13 different parameters from a lake in the English Lake District (UK) to examine the impact of training data duration on the performance of ML models for forecasting chlorophyll-a two weeks in advance. Once training data availability exceeded four years, a Random Forest model was found to consistently outperform naive benchmarks (mean absolute percentage error 16.4 % lower than the best-performing benchmark). With more than 5 years of training data, model performance generally continued to improve, but with diminishing returns. Furthermore, it was found that equivalent and, in some cases, better performance could be achieved by only using a subset of the most important input features. Additionally, it was found that reducing the sampling frequency had negative impacts on performance, both due to the reduced number of training observations available, and increased forecast horizon. Our findings demonstrate that for lakes ecologically similar to the study site, a consistent and regular sampling programme focused on monitoring a limited number of key parameters can provide sufficient observations for generating short-term algae forecasts after approximately five years of data collection. Importantly, this result provides justification for the initiation of new monitoring programmes for sites where algal blooms are a concern, and suggests that there are likely many pre-existing monitoring datasets which would be suitable for training algae forecast models.

Industrial activities of a silk dyeing factory in Thalwil, on the shore of Lake Zurich, Switzerland, caused extreme Sn contamination of lake sediments. In this study, we determine the contamination source, spread, and age using a multiproxy approach. We used X-ray fluorescence spectroscopy (XRF) core scanning and further geochemical analyses to assess the contamination spreading and thickness in the sedimentary column. We found elevated Sn levels throughout sediments of Lake Zurich, ranging from 177 gkg^-1 in front of the former silk factory to 0.05 gkg^-1 at the southeast end (background: ca. 0.006 gkg^-1). The rapid concentration drop away from the shore suggests quick precipitation of a sparingly soluble inorganic Sn compound, which is confirmed by Scanning Electron Microscope Imaging in tandem with Energy-dispersive XRF spectroscopy (SEM-EDX) data. The Sn XRF profile of a varved core indicates a contamination onset in the early 1890s, a maximum around 1900, and a gradual decrease to low levels in the 1940s. High Sn concentrations in turbidite layers from the deep basin indicate that mass movements physically remobilised Sn. However, in stable conditions, in-situ porewater measurements (conc. < 0.5 μgL^-1) using dialyse plates show little Sn remobilisation into the lake water (0.05 mga^-1m^-2). The low remobilisation, reducing conditions, and high sulphide contents in the contaminated layers suggest that Sn is firmly bound to the sediments. Combined with the low toxicity of Sn, we conclude that the Sn contamination poses no threat to lake biota or drinking water production.

Swiss geosciences are at the heart of major national and global challenges that our society is facing, including climate change and weather extremes, long-term environmental trends, natural hazards (e.g. landslides, floods, seismic and volcanic events), the energy transition towards a green and blue economy, sustainable management of natural resources (e.g. water, soils, ecosystems, critical metals, building materials), and the increasing interest towards extra-terrestrial life and resources. In order to remain at the forefront of geosciences research and provide the fertile environment necessary for groundbreaking findings addressing these societal challenges, the Swiss Geosciences Community proposes five large infrastructures, organised into four pillars built on a common datadriven research foundation. [...]