The ocean is losing oxygen due to climate change, threatening ecosystems, especially in naturally low-oxygen areas called Oxygen Minimum Zones (OMZs). Using the IPSL-CM6A-LR Large Ensemble, this study identifies when climate-driven changes in OMZ volumes and regional deoxygenation emerge from natural variability. We highlight hemispheric asymmetries due to ocean ventilation and provide model-based estimates for the timing of detectable OMZ evolution.

Hyperspectral imagery combined with machine learning enables accurate large-scale mapping of plant traits but struggles with uncertainty when facing unfamiliar environmental conditions. This study introduces a distance-based method that measures dissimilarities between new and training data to reliably quantify uncertainty. Results show it effectively identifies uncertain predictions, greatly improving the reliability of global vegetation monitoring compared to traditional methods.

Enhanced weathering can significantly slow climate change by capturing CO2 and converting it into stable bicarbonate. In study, this method removed up to 1.5 tons of carbon per hectare in a single growing season. We had however very little evidence of bicarbonate formation. Our findings show that enhanced weathering facilitated instead carbon accrual via not only carbonate precipitation but also enhanced biogeochemical activities promoting additional carbon storage.

We simulate virtual idealized drainage at 10 pristine peatland sites. Over time, the emission factors of CO₂ flux decrease and the reduction of CH₄ emissions is amplified. The sensitivities of flux changes to drainage are primarily controlled by initial CO₂ and CH₄ fluxes, initial soil carbon content, peat vegetation community, air temperature and initial water table depth. Using GWP100 (100-year Global Warming Potential), our simulation suggested only very small net GHG (greenhouse gas) emission changes when peatland is drained for 50 years.

We have developed a simple mathematical model to investigate how water levels, soil temperature, and plant carbon inputs control methane release from wetlands through production and consumption in the soil. The model captures seasonal and inter-annual variations in emissions and identifies conditions, particularly water depth, that maximise emissions, depending on wetland characteristics. This work supports predicting greenhouse gas fluxes and guiding wetland management and restoration.

We analyzed three decades of carbon concentration and transport in the subpolar gyre of the North Atlantic Ocean, a key region in ocean carbon cycle. Natural carbon remained stable, but human-derived carbon rose by over 30 %. Upper layers exhibit strong seasonal to interannual variability driven by physical and biological processes. Carbon transport was tightly linked to circulation patterns. In the longterm, transport of human-derived carbon is proportional to the increase in its concentration.

We analysed wood density variation in mainland France using 110,000 records from 156 tree species. Models calibrated for 44 species revealed species identity as the dominant driver, followed by tree size and temperature. While neglecting wood density variation had little impact on forest biomass estimates nationally, it caused significant bias at finer scales, highlighting the need to account for wood density variation in forest biomass and carbon assessments.

The September 2022 flood in central Italy left a short-lived yet significant imprint offshore, with patchy sediment deposition, pollutant hotspots (Polycyclic Aromatic Hydrocarbons (PAHs), Poly- and Perfluorinated alkyl substances (PFASs)) and shifts in benthic microbial communities. These findings reveal how extreme events, though transient, can reshape coastal systems, stressing the need for event-based monitoring and improved understanding of flood-driven sediment, contaminant, and ecosystem dynamics.

Peatlands not only store large amounts of carbon but also harmful metals. In a Swedish wetland, we sampled and analysed peat and water samples and found that iron, lead and carbon are released together, especially after dry periods. Lead levels were often toxic and mostly came from past gasoline pollution. The lake trap some of this released lead down stream of the wetland. We showed that climate change may cause peat to break down, releasing carbon and harmful metals into nearby water systems.

We studied how off-season waterlogging affects CO₂ fluxes and dissolved carbon dynamics in cultivated boreal mineral soils. The study was conducted with intact soil profiles in a greenhouse. Waterlogging reduced immediate CO₂ efflux, but CO₂ accumulated in porewater and was released to the atmosphere upon soil drying. Cumulative emissions remained unaltered. Our results suggest that temporary waterlogging does not suppress total CO₂ production in mineral soils as much as commonly assumed.

Forests release trace gases that influence air and climate. While plants are the main source, soil and leaf litter can also release significant amounts, especially in tropical forests like the Amazon. We measured these fluxes in different forest types and found soil and litter to be active sources and sinks. This can improves climate models by including realistic forest processes, vital for understanding and protecting the Amazon.

On intertidal flats, it is hard to distinguish sediment mixing by animals from reworking by waves and currents. We combined tracers to identify reworking of grains of different sizes on the short- and long term. Coarse (sand) grains were less mobile than fine (mud) grains, and partly kept their layering after deposition. The luminescence properties of sand grains can be used for sediment dating and can show sediment mixing, but this method needs to be tested more for young, intertidal sediments.

Plant waxes are an important tool for inferring past changes in vegetation and the water cycle. However, the mechanisms governing plant wax production and stable isotope values are not well understood in the Arctic. We found that terrestrial Arctic plant waxes are not significantly influenced by environmental parameters including latitude, temperature, precipitation amount, humidity, and elevation. These findings agree with our understanding of plant waxes in other regions of the world.

This study reconstructs past marine eukaryotic communities in Storfjordrenna using sedimentary ancient deoxyribonucleic acid (DNA) metabarcoding. Biodiversity remained stable over the past 13,300 years, with changes during major climate transitions. Cercozoans and marine stramenopiles dominated, indicating ecological flexibility. These findings highlight the value of sedimentary ancient DNA for tracking long-term marine ecosystem changes.

The Amazon forest is important for carbon storage, but climate change might push parts of it towards a tipping point into a degraded state. By studying satellite trends and tree diversity across different spatial scales, we found a larger tipping risk at smaller spatial scales than for the whole region. We also found that higher tree diversity makes the forest more stable and thus less likely to tip, although the effect is relatively weak, highlighting the importance of protecting biodiversity.

This study investigates non-biological soil CO₂ ventilation in a semiarid Mediterranean shrubland and standardizes the criteria for its detection. Ten ventilation events were identified during the SCARCE (Synchronized Characterization of Aerosol, Radon and Carbon dioxide Exchanges in drylands) campaign, with several atmospheric parameters acting as drivers. Surface pressure emerged as the primary driver at the site, while friction velocity and boundary layer turbulence were also found to be relevant, highlighting the value of Doppler lidar for soil–atmosphere exchange studies.

In Northern Europe, current temperatures limit the time available for soil preparation and crop growth. Warming may extend the growing season and improve growing conditions, but higher temperatures also increase evapotranspiration and raises the risk of water stress. We evaluated the role of various climatic conditions on crop yield fluctuations in Sweden over 1965–2020 and found that unless Sweden receives more rain in the growing season, crop yields will likely decrease with warming climates.

This study uses machine learning to predict global sea surface chlorophyll a, which is important for monitoring marine ecosystems and the carbon cycle. Using forecasts of sea surface temperature, salinity, height, and mixed layer depth, we generate global predictions up to six months ahead in just minutes. Our approach matches state-of-the-art numerical methods while being faster and more resource-efficient.