Département Eaux de surface

Many components of the Earth system feature self-reinforcing feedback processes that can potentially scale up a small initial change to a fundamental state change of the underlying system in a sometimes abrupt or irreversible manner beyond a critical threshold. Such tipping points can be found across a wide range of spatial and temporal scales and are expressed in very different observable variables. For example, early-warning signals of approaching critical transitions may manifest in localised spatial pattern formation of vegetation within years as observed for the Amazon rainforest. In contrast, the susceptibility of ice sheets to tipping dynamics can unfold at basin to sub-continental scales, over centuries to even millennia. Accordingly, to improve the understanding of the underlying processes, to capture present-day system states and to monitor early-warning signals, tipping point science relies on diverse data products. To that end, Earth observation has proven indispensable as it provides a broad range of data products with varying spatio-temporal scales and resolutions. Here we review the observable characteristics of selected potential climate tipping systems associated with the multiple stages of a tipping process: This includes i) gaining system and process understanding, ii) detecting early-warning signals for resilience loss when approaching potential tipping points and iii) monitoring progressing tipping dynamics across scales in space and time. By assessing how well the observational requirements are met by the Essential Climate Variables (ECVs) defined by the Global Climate Observing System (GCOS), we identify gaps in the portfolio and what is needed to better characterise potential candidate tipping elements. Gaps have been identified for the Amazon forest system (vegetation water content), permafrost (ground subsidence), Atlantic Meridional Overturning Circulation, AMOC (section mass, heat and fresh water transports and freshwater input from ice sheet edges) and ice sheets (e.g. surface melt). For many of the ECVs, issues in specifications have been identified. Of main concern are spatial resolution and missing variables, calling for an update of the ECVS or a separate, dedicated catalogue of tipping variables.

Waterborne pathogens pose a serious threat to public health, emphasizing the need for reliable and efficient water treatment technologies. Wastewater treatment plants employ a range of processes to reduce microbial contamination, with membrane filtration emerging as a promising solution due to its ability to physically remove pathogens without the production of harmful chemical by-products. This study investigates the effectiveness of a gravity-driven membrane (GDM) filtration system for pathogen removal from wastewater and evaluates the associated public health risks with and without treatment. A quantitative microbial risk assessment model was employed to estimate infection probabilities for various waterborne pathogens. The results demonstrated a significant decrease in pathogen concentrations following treatment, with up to a 10⁴-fold reduction in norovirus infection risk. Three critical factors influencing membrane performance were identified: membrane integrity, pore size characteristics, and membrane fouling. Maintaining membrane integrity was found to be essential for ensuring consistent pathogen removal. While nominal pore size is commonly used to predict rejection efficiency, the overall pore size distribution was found to have a greater influence on virus retention. Additionally, although membrane fouling is often considered detrimental, it was shown to enhance virus removal by up to two orders of magnitude. These findings underscore the potential of GDM systems for effective virus removal and highlight the importance of proper membrane design, maintenance, and monitoring in ensuring long-term operational efficiency and maximizing public health protection in wastewater treatment applications.

Forest ecosystems are vital for biodiversity, climate regulation, and ecosystem services. Their resilience depends not only on species diversity but also on intraspecific variation—the genetic and phenotypic differences within species—which underpins adaptive capacity to environmental change. However, large-scale, continuous monitoring of intraspecific variation remains challenging. Here, we present a remote sensing approach using Sentinel-2 time series of five vegetation indices as proxies for pigment content, canopy structure, and water content to detect intraspecific variation in seven tree species across a broad environmental gradient in Switzerland. Using pure-species plot data from the Swiss National Forest Inventory, we decomposed variation into spatial, temporal, and spatiotemporal components. We found that spatial variation dominated in evergreen species (48–86%), while temporal variation was more pronounced in deciduous species (56–82%), reflecting their stronger seasonality. These findings demonstrate that species-specific Sentinel-2 time series can effectively track intraspecific variation, providing a scalable method for forest monitoring. This approach opens new pathways for studying forest adaptation, informing management strategies, and guiding species selection for conservation under changing climate conditions.