The representation of light propagation in seawater is critical for modelling marine biogeochemistry. We analyse results from a radiative transfer model that accounts for the absorption and scattering of light in the ocean with their respective uncertainties. We compare these results with in situ and remote-sensed data. Our analysis highlights the benefits of accounting for model uncertainties while using advanced representations of light in modelling frameworks.

Nitrate is a crucial nutrient in oceans, whose excess can trigger uncontrolled algae growth that damages marine ecosystems. We used machine learning to generate skilled, gap-free, bi-decadal surface nitrate data from sparse observations, revealing areas on the North-West European Shelf that are more vulnerable to excess algae growth if nutrient pollution occurs. We also looked at bi-decadal trends in coastal nitrate and the impact of winter nitrate on spring phytoplankton blooms.

To investigate if greenhouse gases (GHG) emissions from ponds vary among clear-water ponds with macrophytes and turbid-water ponds with phytoplankton, we measured CO₂, CH₄, and N₂O concentrations and emissions in two clear- and two turbid-water urban ponds in the city of Brussels from June 2021 to December 2023. Differences in  CH₄ ebullitive emissions were observed between clear- and turbid-water ponds but none for CO₂ and N₂O emissions.

Treeline ecotones are exposed to pronounced differences in irradiation and nutrient availability in complex mountain relief. We compared growth phenology and growth rate of Picea abies on north- and south-facing slopes at sites differing in nutrient availability at the lower part of the treeline ecotone. Our results suggest that temperature alone governs growth phenology, but nutrient availability modulates the growth rate in the peak season when temperature no longer limits cambial activity.

Climate warming speeds up lake eutrophication, creating “dead zones” where aquatic life suffocates due to oxygen depletion. The sediments of Amsoldingersee, a Swiss lake, revealed how climate shifts impacted the lake around 10 000–18 000 years ago. (1) Algal composition differed between both cold and warm periods. (2) Nutrient additions from dust controlled algal growth more than temperature. (3) Cold periods with ice cover led to oxygen depletion. (4) Algal communities recovered after anoxic phases.

We integrated palaeoecological and geochemical data to discern the impact of catastrophic events on the development of peatlands within pine monocultures. An approach that integrates these methods is not commonly employed but offers a more comprehensive understanding of past ecosystem transformations. We used multi-proxy research of the peat core and neodymium isotope record. We support the results of our analyses with the recognition of statistically significant critical transitions.

We combine two types of carbon dioxide (CO₂) data from Dutch peatlands in a machine learning model: from fixed measurement towers and from a light research aircraft. We find that emissions increase with deeper water table depths (WTDs) by 4.6 tons of CO₂ per hectare per year for each 10 cm deeper WTD on average. The effect is stronger in winter than in summer and varies between locations. This variability should be taken into account when developing mitigation measures.

Thawing permafrost at Iškoras in northern Norway is transforming peat plateaus into thermokarst ponds and wetlands. These small ponds show striking oversaturation of dissolved greenhouse gases, such as carbon dioxide (CO₂) and methane (CH₄), partly owing to organic matter processing. Streams nearby emit CO₂, driven by turbulence. As permafrost disappears, carbon dynamics will change, potentially increasing emissions of CH₄. This study highlights the need to integrate these changes into climate models.

Soils are a large reservoir of carbon (C) on land, and there is uncertainty regarding how this reservoir will be affected by climate change. Currently, active research on (1) how changing precipitation patterns, a key aspect of climate change, will affect soil C and (2) how vulnerable subsoils are to climate change is still being undertaken. In this study, we examined subsoils after 20 years of experimentally manipulated precipitation shifts to see whether increasing precipitation would affect C amounts and chemistry.

Archaeal molecular fossils (tetraethers) have been used around the globe to track changes in climate. Little is known about the archaeal response to environmental change in soda lakes, especially lakes influenced by hydrothermal inputs. For the first time in Lake Magadi, we show tetraethers tracking abrupt changes in methane and non-methane producers due to hydrothermal inputs to the lake. This study provides insight into the role of hydrothermal water and methane production in soda lakes.

The 2015–2016 El Niño caused severe droughts in tropical forests, but its impact on the Cerrado, Brazil's largest savanna, was unclear. Our study tracked the productivity of two key Cerrado vegetation types over 5 years. Before the El Niño, productivity was higher in the transitional forest–savanna, but it dropped sharply during the event. Meanwhile, the savanna showed minor changes. These findings suggest that transitional ecosystems are particularly vulnerable to drought and climate change.

This study evaluates machine learning approaches for upscaling evapotranspiration from the site to the global scale. Sequential models capture temporal dynamics better, especially with precipitation data, but all models show biases in data-scarce regions. Improved upscaling requires richer training data, informed covariate selection, and physical constraints to enhance robustness and reduce extrapolation errors.

Redox potential, indicative of the active pathways of organic matter decomposition, was monitored for 2 years in a boreal peatland with three drainage regimes. Contrary to expectations, the water table level and redox potential were not found to be correlated in a monotonic fashion; thus, the relationship between the water table level and redox conditions is not modellable using non-dynamic models.

The rapid expansion of rubber cultivation constitutes a significant land-use change in Southeast Asia. Despite fertilization being a common practice in rubber plantations, its impact on soil methane (CH₄) dynamics has remained poorly understood. Our study demonstrates that fertilization not only reduces soil CH₄ consumption but also increases CH₄ production, transforming rubber plantations from a net CH₄ sink to a source. Implementing rational fertilization could enhance atmospheric CH₄ removal.