This limited observational base gives rise to high uncertainty regarding the future terrestrial carbon sink. Many questions need answer to determine if this negative feedback to climate change will be sustained under future environmental changes, or whether increases in autotrophic respiration or carbon turnover might counteract it, for example through accelerated tree mortality or more frequent and more severe disturbance events (e.g. drought, fire, insect outbreaks) Shifts in the dynamics of plant mortality, establishment, and growth are expected to significantly influence forest composition.
Uncertainties and/or data gaps in large-scale empirical products of vegetation dynamics, carbon fluxes and stocks may be overcome by extensive collections of field data and new satellite retrievals of forest biomass and other vegetation properties. Such novel datasets may be used to evaluate, develop and parametrize global vegetation models and hence to constrain present and future simulations of vegetation dynamics. Where no observations exist, exploratory modelling can investigate realistic responses and identify priorities for field and experimental campaigns. We welcome contributions that make use of observational approaches, vegetation models, or model-data integration techniques to advance understanding of the effects of environmental change on vegetation dynamics, tree mortality as well as carbon stocks and fluxes at local, regional or global scales and/or over long periods.
Orals: Wed, 6 May, 14:00–18:00 | Room N1
The carbon (C) sink in aboveground woody biomass (B) of mature forests remains one of the most uncertain components of the global C budget, with contrasting estimates from forest inventories, remote sensing, and ecosystem models, and an uncertain contribution from environmental change. The self-thinning relationship—describing the decline in tree density as mean tree size increases—encapsulates the carrying capacity for B and varies across forest types and environments. Assessing its temporal stability enables the separation of C uptake driven by changes in forest area or recovery from past disturbance from changes in the carrying capacity of mature forest biomass, potentially induced by environmental change.
Here, we compiled 105,763 inventories in natural forests spanning all major forest biomes worldwide to quantify temporal changes in the self-thinning relationship at the global scale. We detected a gradual and pervasive upward shift of the tree density–size relationship across all biomes, suggesting a thickening of mature forests and a partial relaxation of self-thinning constraints. The most pronounced thickening occurred in forests located in warm, dry climates and in regions with low nitrogen deposition and high soil phosphorus availability, whereas forests characterized by high soil C:N ratios and elevated organic carbon content showed the weakest responses. The observed shift in the self-thinning relationship implies a global C sink of 1.9 Pg C yr-1 [95% confidence interval: 1.77-2.07 Pg C yr-1], highlighting the changing carrying capacity of aboveground biomass stocks in mature forests as a key mechanism underlying the persistent terrestrial C sink.
How to cite: Stocker, B. D. and Marqués, L. and the Global Forest Inventory Data Analysis Team: Increasing carbon storage capacity across global forest biomes, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19021, https://doi.org/10.5194/egusphere-egu26-19021, 2026.
Photosynthetic theory predicts that rising atmospheric CO₂ should enhance photosynthesis in tropical trees, potentially increasing stem growth and strengthening the tropical forest carbon sink. However, consistent positive CO₂ effects on stem growth are rarely detected in observational studies. Here, we investigate whether tree light exposure, climatic variability, and statistical limitations can explain this apparent discrepancy.
We used a previously parameterised and tested forest model to simulate tropical tree populations under fixed and rising historical CO₂ and climate representative for SE Asian lowland tropical forest. To represent realistic variation in light availability, trees were simulated in gaps of different sizes, explicitly resolving height-dependent light gradients, constraints on maximum canopy size, and dynamic changes in light conditions as trees grow. Simulations were conducted under source-limited conditions.
Across simulations, CO₂ effects on growth were weak compared to the effects of climate and light availability. The simulated CO₂ response was comparable in magnitude to effects reported in temperate forest FACE experiments, but substantially stronger than those typically inferred from tree-ring studies. CO₂ effects were amplified in cooler years but showed little sensitivity to precipitation variability.
Using the simulated data, we then evaluated whether recommended statistical approaches for the detection of CO₂ effects in tree rings could recover the CO₂ signal obtained from the model simulations. We found that, despite its relatively strong magnitude, the CO₂ effect was difficult to detect reliably. Two out of four tried methods detected a CO₂ effect, but its presence and strength were strongly dependent on the statistical model assumptions.
These results highlight the challenges of attributing CO₂ effects on tree growth in real-world observational data, which are subject to substantial noise and may exhibit weaker responses. Progress in detecting CO₂ effects may benefit from closer integration of simulation experiments and statistical inference to guide study design and interpretation.
How to cite: Zwartsenberg, S. A., de Vries, J., Sterck, F. J., Anten, N. P. R., and Zuidema, P. A.: When models meet data: Limits to detecting CO₂ effects in tropical forests, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11409, https://doi.org/10.5194/egusphere-egu26-11409, 2026.
Tropical forests are rapidly changing due to land-use conversion, with major consequences for biodiversity and the terrestrial carbon balance. Secondary (regrowth) forests now cover a larger global area than primary forests, making their carbon accumulation central to the future strength of the terrestrial carbon sink. In sub-Saharan Africa, slash-and-burn agriculture remains the dominant disturbance and is expected to intensify as the human population triples by 2100. As a result, secondary forests are increasingly reshaping regional vegetation structure, yet biomass recovery rates, drivers, and uncertainties remain poorly constrained, limiting carbon stock estimates and vegetation model performance.
We compiled a new pan-African field dataset of aboveground carbon (AGC) stocks across Afrotropical secondary forest succession using 969 inventory plots from 31 sites in 12 countries. AGC recovery trajectories were modeled within a hierarchical Bayesian framework that propagates observational and process uncertainty. Median times to recover 90% of old-growth AGC (t90) ranged from 36 to 91 years, with uncertainty intervals at some sites extending beyond a century. The early-successional AGC accumulation (first 20 years) varied widely, from 12 to 100 Mg C ha-1. Mean annual precipitation emerged as a dominant control on AGC recovery, outperforming other climatic (temperature, photosynthetically active radiation, seasonality, and maximum climatic water deficit), soil (chemical and physical), and landscape predictors (forest cover and distance metrics). Absolute AGC accumulation during the first two decades was more predictable (R² = 0.56) than recovery relative to old-growth reference conditions (R² = 0.21), a pattern especially relevant for carbon sequestration assessments.
Overall, forest carbon recovery across the Afrotropics is relatively slow, heterogeneous, and often highly uncertain. By synthesizing extensive field data, this study provides empirical benchmarks for validating vegetation models and improving projections of the terrestrial carbon sink. However, the large uncertainties underscore the need to expand long-term forest inventories and reinforce the importance of conserving the remaining old-growth forests alongside the carbon sequestration potential of natural regeneration.
How to cite: Van de Velde, V., Boeckx, P., Makelele, I., Ewango, C., Bienvenu, A., N'Guessan, A. E., N'Dja, J. K., Hérault, B., and Bauters, M.: Environmental controls on carbon stock recovery rates in Afrotropical secondary forest, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10545, https://doi.org/10.5194/egusphere-egu26-10545, 2026.
The proliferation of satellite missions targeting observations of terrestrial land surfaces has substantially increased efforts to map and monitor carbon stored in vegetation using remote-sensing datasets. One such effort, ESA’s Climate Change Initiative (CCI) Biomass project, is about to release an 18-year record of global aboveground biomass (AGB) maps covering the periods 2005–2012 and 2015–2024 at a spatial resolution of 1 hectare. This latest release (version 7.0) significantly extends previous data records and reduces temporal inconsistencies caused by the diverse set of satellite observations required to construct such a long time series. With this release, the CCI Biomass dataset enables the tracking of carbon dynamics in terrestrial vegetation over the past two decades and is now comparable with other global, satellite-based, spatially explicit datasets.
In this presentation, we highlight three key findings derived from the CCI Biomass dataset.
- We identify spatially consistent biomass accumulation in tropical secondary forests in Brazil, in agreement with sample-based estimates derived from in situ measurements where available. Spatially resolved growth trends, combined with forest age information from the MapBiomas dataset, indicate higher growth rates in the western Amazon, with peaks of up to 10 Mg ha⁻¹ yr⁻¹ at approximately 10–15 years of forest age. In contrast, growth rates in the eastern Amazon do not exceed 5 Mg ha⁻¹ yr⁻¹.
- We detect contrasting biomass trends in primary forests of the Brazilian Amazon. Below a biomass threshold of 250 Mg ha⁻¹, forests exhibit an average accumulation of approximately 1–2 Mg ha⁻¹ yr⁻¹, whereas above this threshold high-biomass forests show a decline of around −1 Mg ha⁻¹ yr⁻¹. These findings are consistent with recent evidence suggesting a weakening of the Amazon carbon sink. However, they remain unconfirmed and will need further investigation, particularly with respect to statistical significance, given the substantial pixel-level uncertainty in the CCI Biomass estimates.
- A comparison among several global AGB datasets derived from satellite data (e.g., Xu et al., Boitard et al., Li et al., Santoro et al.), including CCI Biomass, reveals broad agreement in the spatial distribution of biomass. However, absolute AGB estimates can differ by up to 100% among datasets, irrespective of geographic location. Moreover, temporal biomass trajectories often diverge, showing differences in the magnitude of fluctuations and, in some cases, opposing growth trends. Overall, our analysis underscores the need for a systematic intercomparison of remote-sensing-based AGB datasets using a common framework to assess their accuracy and uncertainty.
How to cite: Santoro, M., Cartus, O., Favrichon, S., Heinrich, V., Urbazaev, M., Araza, A., and Herold, M.: Tracking terrestrial biomass from space: patterns, trends, and uncertainties, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13930, https://doi.org/10.5194/egusphere-egu26-13930, 2026.
The Amazon rainforest, home to nearly 40% of Earth’s tropical forest biomass, plays a central role in the global carbon sink. Recent observations indicate a decline in net carbon uptake, attributed to intensifying hydro-climatic extremes, deforestation, and forest degradation. Seasonal droughts are lengthening, wet-season onsets are delaying, and strong El Niño events are becoming more frequent, yet the mechanisms by which these changes influence ecosystem processes remain poorly understood.
Leaf litterfall is a major pathway of carbon and nutrient transfer from vegetation to soils, integrating climate seasonality, atmospheric demand, and plant physiological strategies. Here, we analyze eight years of litterfall observations from a lowland Amazon rainforest together with meteorological data to identify seasonal and interannual climate controls on canopy turnover.
We find that litterfall dynamics differ systematically between years with typical and anomalous litterfall patterns. In years with typical seasonality, litterfall patterns are well explained by atmospheric moisture and pressure variables, with maximum relative humidity and barometric pressure emerging as dominant predictors. In contrast, anomalous years show distinct responses depending on hydro-climatic context: wetter wet seasons followed by intense dry seasons are associated with elevated litterfall, best explained by cumulative water deficit, whereas drier wet seasons are linked to suppressed litterfall driven by high evaporative demand, captured by potential evapotranspiration. Across all conditions, litterfall reflects the combined influence of antecedent hydrological states and immediate atmospheric demand, challenging phenology frameworks based primarily on precipitation or temperature alone.
Moisture-demand and pressure variables such as vapor pressure deficit, potential evapotranspiration, and barometric pressure may thus provide powerful yet underutilized insights into vegetation turnover, although better models will emerge from the integration of other biological processes such as stem flow and root dynamics. As climate change alters seasonality and increases hydro-climatic extremes, disruptions in canopy turnover are likely to influence vegetation dynamics, nutrient cycling, and the future resilience of the Amazon forests.
How to cite: Fleischer, K., Bleichrodt, J., Hofhansl, F., Damasceno, A., Pereira, I., Fuchslueger, L., Valverde-Barrantes, O. J., P. Martins, N., Aleixo, I. F., F. Lugli, L., Miron, A. C., and Garcia, S. and the AmazonFACE team: Climate Controls on Canopy Turnover: The Role of Atmospheric Demand and Hydrological Legacies in an Amazonian Lowland Forest, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12959, https://doi.org/10.5194/egusphere-egu26-12959, 2026.
Phenological change is among the clearest biological fingerprints of climate change, shaping carbon cycling, ecosystem productivity, and trophic interactions. In this work, we synthesize more than 2,400 site-years of high-frequency phenocam imagery—from the PhenoCam Network, Australia’s Terrestrial Ecosystem Research Network, and Europe’s Integrated Carbon Observation System—to evaluate continental-scale patterns in phenology and vegetation greenness. We show that phenological and greenness metrics differ markedly across primary vegetation types and climate zones, with baseline phenology varying strongly across vegetation–climate combinations, revealing substantial ecological structuring of seasonal dynamics. In contrast, temporal trends are generally modest, heterogeneous, and seldom statistically distinguishable from zero; only a small number of vegetation–climate subgroups display detectable directional change. Generalized additive models fitted to multi-year mean site values indicate that climate and vegetation type together explain up to 58% of cross-site deviance, and that climate–vegetation interactions improve model performance by ~10% on average—most strongly for length-of-season metrics. Because Earth system models depend on realistic seasonal dynamics to constrain carbon–climate feedbacks, our results identify where model representation most needs improvement and which components of vegetation seasonality are most sensitive to climate forcing. Taken together, the findings suggest that spatial variation in plant phenology is strongly governed by vegetation–climate coupling, whereas coherent phenological shifts over time have yet to emerge at continental scales. We highlight the value of harmonized phenocam data for detecting early signals of ecological change and the need for continued international coordination toward a global phenocam dataset.
How to cite: Hagen, C. J., Jay, K. R., Loescher, H. W., Murphy, B. A., Post, A. K., Richardson, A. D., SanClements, M. D., and Staudhammer, C. L. and the GERI Phenology Team: Climate and vegetation jointly govern continental-scale patterns of plant phenology and vegetation greenness, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7621, https://doi.org/10.5194/egusphere-egu26-7621, 2026.
Lianas exist in virtually every tropical forest, play a vital role for their functioning but also exert a strong pressure on their hosting trees. The exact influence of lianas on tree structure is poorly understood, despite its potential implications. Here, we investigated the relationship between liana infestation and tree height across the pantropics. Our analysis based on 50,473 tree records revealed substantial individual tree height reduction with liana infestation. Controlling for species, size and site, trees were on average 4.2% and 7.8% shorter when moderately and heavily infested by lianas compared to their liana-free counterparts. Substantial tree height reduction for heavily infested trees was found in a majority of the abundant species in our dataset, as well as in 46 of the 66 sites that we compiled in this study. Among the sites where a substantial liana-induced tree height reduction was identified, its magnitude varied between 3.6% and 39.3% for a 50 cm DBH tree under heavy liana infestation. Liana impact on tree height varied with disturbance, elevation and climate. In lowland, old-growth forests, liana impacts on tree height strengthened under drier conditions and warmer temperature regimes, particularly in sites characterised by higher minimum temperatures, warmer warm-season conditions, and reduced thermal seasonality. As global warming is predicted to exacerbate these climatic drivers, individual tree height reduction with heavy liana infestation could aggravate pantropically by 48% (middle of the road emission scenario) to 92% (business as usual) by the end of this century. Given the central role of tropical forest structure in governing carbon sequestration and climate feedback, these findings, coupled to the globally observed increase in tropical liana abundance, suggests that lianas could disproportionately influence Earth system functioning under future climate change.
How to cite: Verbeeck, H., Krishna Moorthy, S. M., and Meunier, F. and the Data contributors and collaborators: Lianas reshape tree height allometry across tropical forests, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14844, https://doi.org/10.5194/egusphere-egu26-14844, 2026.
Accurate monitoring of forest CO2 sequestration is essential for the European Union’s mission to achieve carbon neutrality. However, current estimates of the European forest carbon sink vary greatly depending on the methodology used. In particular, forest dynamics, such as harvest, disturbances and regrowth, are poorly captured by current approaches. Inverse modelling based on observations of atmospheric CO2 is a key approach to constraining global and regional carbon budgets and quantifying the forest CO2 uptake. CO2 inversions capture hourly-to-seasonal variability in surface CO2 fluxes well, such as during droughts. Nevertheless, due to the rapid mixing of atmospheric CO2, inverse modelling systems require additional information from local observations to capture the effects of forest dynamics over longer timescales and across space. National forest inventories (NFIs) provide a promising and underused observational data stream of aboveground biomass, biomass increment, and forest demography across space and time.
Here, we integrate forest demography data from NFIs with a biosphere model to constrain European-wide land carbon fluxes from hourly to decadal scales. Our approach combines the Simple Biosphere Model 4 (SiB4) with European forest inventory data and the forest resources model EFISCEN-Space. Including forest demography in our simulations leads to an average increase in the total European land sink strength of roughly 50 TgC yr-1 over 2000-2020, compared to a baseline simulation without demography. This additional uptake is mainly attributed to managed forests in Central Europe. Climate extremes such as the 2018 and 2022 droughts introduce additional variability in European forest carbon fluxes, which we show both spatially and temporally in our optimised net ecosystem exchange (NEE) fluxes. Finally, we evaluate the simulated CO2 mole fractions from our modelling system across the network of European atmospheric monitoring sites. This analysis marks an important step towards including NFIs as an additional constraint in our inverse modelling system for the carbon cycle, CarbonTracker Europe. With this development, we aim to bridge atmospheric and ground-based data in estimating the European forest carbon sink.
How to cite: van de Sande, M. A. J., van der Woude, A. M., Hooghiem, J. J. D., Filipek, S., Schelhaas, M.-J., Zuidema, P. A., Nabuurs, G.-J., and Peters, W.: Constraining Hourly to Decadal Forest Carbon Fluxes with European National Forest Inventories, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9817, https://doi.org/10.5194/egusphere-egu26-9817, 2026.
Tree mast seeding i.e., synchronous, highly variable seed production among years, has wide consequences for ecosystem functioning. In a large-seeding year, a pulse of resources is made available to wildlife over the large spatial scale by virtue of the synchronous reproduction by millions of trees. But how trees synchronize masting over the vast geographical areas? A major mechanism governing the annual allocation of resources to seed production is weather variation. If individuals respond to the same “weather cue” across extensive regions, masting synchrony can emerge. This requires, however, the timing of the cue window being well conserved across species range, but mechanisms facilitating such stability remain unknown. Here, we investigated factors driving masting synchrony in European beech, which extends to the thousands of kilometres throughout its geographic range. We used a moving window analysis to determine how correlations between annual seed production and mean temperatures in 61 populations of European beech sampled across the species' range fluctuate at a fine temporal scale. We found that correlation coefficient values between seed production and temperature rapidly increased after the summer solstice, compared to before it. Moreover, using temporally-restricted permutation tests we showed that this abrupt increase in masting-weather correlations at the solstice is not driven purely by chance. Beech achieves high spatial synchrony in seed production by anchoring the weather cue window to the summer solstice - the longest day of the year that occurs simultaneously across the whole Northern Hemisphere. Beech abruptly opens its temperature-sensing window on the solstice, hence widely separated populations all start responding to weather signals in the same week. This enables cohesive timekeeping across distant populations inhabiting diverse climatic regions and creates a high precision timing of the Moran effect, leading to the subcontinental-scale synchrony of beech mast seeding.
How to cite: Szymkowiak, J., Journé, V., Foest, J., Hacket-Pain, A., Kelly, D., and Bogdziewicz, M.: Summer solstice orchestrates the subcontinental-scale synchrony of European beech (Fagus sylvatica) mast seeding, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10688, https://doi.org/10.5194/egusphere-egu26-10688, 2026.
Understanding how changes in vegetation cover affect nutrient dynamics is essential for predicting future ecosystem responses to environmental change. In this study, we integrated ground-based observations, remote sensing data, and dynamic process-based modelling to investigate vegetation dynamics under changing environmental conditions and disturbances driven by both natural processes and human activities. Our primary objective was to assess how land-use change, disturbances, and management practices influence nutrient cycling in plant ecosystems.
To address this objective, we compiled detailed land-use and land-cover data for 270 currently forested sites across five European countries (Croatia, Hungary, Slovakia, Czech Republic, and Poland). Data sources spanned from the second half of the 18th century to the present and included historical military maps, aerial photographs, satellite imagery, and forest management plans. The analysis revealed that 84% of sites experienced a vegetation change, while one third of them underwent multiple types of change. Management interventions were the most common driver occurring on more than a half of sites, followed by shifts in tree species composition at over one third of sites and deforestation observed at one quarter of sites. Natural disturbances were identified only at one fifth of sites.
Subsequently, we simulated vegetation dynamics using the process-based model Biome-BGCMuSo. Each site was modelled under two simulation set-ups: one considering only the current ecosystem state, and another incorporating the documented vegetation changes over the past 200 years. This design enabled us to isolate the effects of historical vegetation dynamics on ecosystem stocks and fluxes. In total, 40 variables related to carbon, nitrogen, and water cycling were analysed. The magnitudes of differences between the two set-ups varied among ecosystem components, sites, and species, and were strongly linked to the type and frequency of vegetation changes. The most pronounced negative effects, reaching up to 50% difference, were observed in soil, litter, and coarse woody debris carbon and nitrogen stocks, as well as in net ecosystem exchange and heterotrophic respiration following deforestation.
Our results highlight the critical importance of accounting for historical vegetation changes in ecosystem modelling. By demonstrating how legacy effects shape present-day nutrient dynamics, this study underscores that ecosystem functioning reflects not only current conditions but also the cumulative influence of past land use, management, and disturbance history.
The study was funded by the EU NextGenerationEU through the Recovery and Resilience Plan for Slovakia under the project No. 09I03-03-V04- 00130.
How to cite: Merganičová, K., Lieskovský, J., Ortmann-Ajkai, A., Kaim, D., Ostafin, K., Stych, P., Marjanovic, H., Hidy, D., Barcza, Z., and Merganič, J.: How changes in vegetation cover affect nutrient dynamics of forest ecosystems, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16286, https://doi.org/10.5194/egusphere-egu26-16286, 2026.
Information on forest responses to global change drivers remains fragmented across scientific disciplines, spatial scales, and methodological approaches, limiting our ability to understand and predict forest vulnerability and resilience. Reconciling ground-based physiological observations with landscape- to continental-scale evidence from modelling and remote sensing is essential to advance our understanding and inform effective policy for forest management.
Here, we synthesise current knowledge on how European forest condition has changed over recent decades by integrating evidence across spatial scales and methodological frameworks. We focus on three key metrics: tree growth, water-use efficiency (WUE), and tree mortality, which together link physiological processes to ecosystem-level responses and forest health. These metrics were assessed in relation to major global change drivers, including climate change (drought and climate extremes) and atmospheric pollution (rising CO₂ concentrations and changing nitrogen and sulphur deposition).
As part of Working Group 2 of the COST Action CA21138 CLEANFOREST, we conducted a systematic review of trends in European forest growth, WUE, and mortality reported since 1990. We extracted the direction of reported trends (positive, negative, or neutral) from more than 500 peer-reviewed studies, spanning dendrochronology, ecosystem flux measurements, forest inventories, modelling, and remote sensing, alongside detailed information on forest type, climate zone, tree species/genus, and site characteristics. The resulting database comprises over 1,300 observations across Europe.
Our synthesis reveals pronounced spatial, temporal, and ecological asymmetries in the European forest evidence base. Studies are heavily concentrated in Central Europe, with substantial gaps in Mediterranean, Eastern European, and high-latitude regions. Reported trends indicate predominantly negative growth and increased mortality in southern Europe, in contrast to more neutral or positive signals in central and northern regions. Across scales, tree-level observations often suggest physiological compensation (e.g. increasing WUE), whereas landscape-scale assessments more frequently reveal stagnating growth and intensified mortality, highlighting a mismatch between local adjustments and ecosystem-level responses.
By reconciling evidence across scales, this systematic review draws a comprehensive picture of the complex response of European forests to the changing climate, identifies key knowledge gaps, biogeographic biases, and opportunities to better integrate long-term monitoring networks with emerging approaches. Such integration is essential for understanding mechanisms underpinning forest responses to climate extremes and atmospheric deposition, and for improving projections of future forest functioning under global change conditions.
How to cite: Creek, D. and the COST Action CA21138 CLEANFOREST Working Group 2: Forest responses to global change: reconciling evidence from tree to continental scales in Europe, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10478, https://doi.org/10.5194/egusphere-egu26-10478, 2026.
Predicting the future carbon balance of forests, and their carbon sequestration capacity, requires a precise understanding of how gross primary productivity (GPP) is partitioned among autotrophic respiration and the different compartments of the net primary productivity (NPP). Studies show an important variability of the NPP:GPP ratio across forest, and a partial decoupling between GPP and wood production in forest ecosystems. This suggests an interannual variation of carbon allocation among tree functions and organs, which is generally not accounted for in most dynamic vegetation models. Using GPP estimated from eddy-covariance measurements and independent above-ground NPP measured over 6 to 20 years on six forest sites (5 sites in France belonging to the ICOS network and one site in Brazil), we explored the interannual variations in GPP partitioning to aboveground growth and its distribution among years and aboveground organs (wood, leaves, fruits, flowers).
The partitioning of GPP to aboveground biomass varied considerably across sites, with Mediterranean evergreen forests showing the lowest values (15% and 18%), temperate forests intermediate values (21-37%), and the tropical eucalypt site showing the highest fraction (48%). At four of the sites, biomass production exhibited a larger inter-annual variability than did GPP, suggesting a greater sensitivity to environmental controls of the carbon sinks than the carbon source. All sites but one exhibited a significant correlation between annual aboveground NPP and annual GPP, but with small R² values between 0.2 and 0.6, thus showing a rather weak coupling between the two productivities. The coupling was generally even weaker for wood production alone (generally considered as the main carbon sequestration in forests) than for total aboveground NPP (which also includes short-lived organs such as leaves that will rapidly decompose). Finally, we observed that the inter-annual correlations between GPP and biomass production varied depending on the onset of the GPP integration time-window, indicating different temporal lags between assimilation and growth according to species and organs.
This works highlights the necessity to take into account inter-annual variations of carbon allocation in forest carbon balances, and to better understand of the climatic drivers of sink activity including potential lags between assimilation, storage and growth.
How to cite: Poughon, J., Cailleret, M., Delpierre, N., Berveiller, D., Chipeaux, C., Courtois, P., Cuntz, M., Domec, J.-C., Guillemot, J., Joetzjer, E., Kempf, J., Lafont, S., Le Maire, G., Marloie, O., Morfin, A., Nouvellon, Y., Ourcival, J.-M., Simioni, G., and Limousin, J.-M.: Inter-annual variation in carbon allocation explains partial decoupling between assimilation and growth in six forest sites, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4036, https://doi.org/10.5194/egusphere-egu26-4036, 2026.
Europe's forests store nearly 40 PgC and provide a critical carbon sink of ~0.2 PgC yr-1, yet climate-driven disturbances increasingly threaten this capacity. Although disturbance rates from windthrow and bark beetle outbreaks have risen in recent decades, it remains unclear whether these events increasingly affect the oldest and largest trees, which store a disproportionate share of carbon. Here, we combine three decades of satellite-derived disturbance maps with spatially explicit data on forest age, biomass, and species composition to reveal patterns of structural selectivity across Europe. We show that natural disturbances have shifted toward older, carbon-rich stands, with disturbed forest area > 60 years old nearly tripling since 2010 (from 0.38 to 1.06 Mha). This structural shift is most pronounced in spruce-dominated regions of Central Europe (effect size = 1), where compound heat and drought events have amplified susceptibility to bark beetles. Biomass losses from natural disturbances in spruce forests increased eightfold between the early (2011-2016) and recent (2017-2023) periods. Trend-based projections indicate that, if current patterns of structural selectivity persist, natural disturbances could expose biomass carbon stocks equivalent to approximately 20 % of Europe’s contemporary forest carbon sink by 2040 (~0.05 PgC yr -1 or ~0.7 PgC cumulative). Our findings reveal a previously unquantified vulnerability: climate-driven disturbances increasingly affect forest structures with high per-hectare carbon stocks, amplifying disturbance-related carbon exposure and weakening the long-term effectiveness of Europe’s forest carbon sink. Adaptive management strategies that promote structural and compositional diversification in high-risk regions will be critical to stabilise forest carbon storage under continued climate change.
How to cite: Besnard, S., Viana-Soto, A., Hartmann, H., Patacca, M., Heinrich, V. H. A., Kowalski, K., Santoro, M., De Keersmaecker, W., Van De Kerchove, R., Herold, M., and Senf, C.: Natural disturbances increasingly affect Europe’s most mature and carbon-rich forests , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-3036, https://doi.org/10.5194/egusphere-egu26-3036, 2026.
Changes in forest disturbances can have strong impacts on forest structure and carbon dynamics. Yet, we lack consistent long-term data on forest disturbance regime shifts and their implications for carbon stocks. Observational evidence is mandatory for developing adaptation strategies in vulnerable regions such as the European Alps. Here, we present an observation-based, alpine-scale characterization of forest disturbances and their temporal evolution, together with modelled impacts on above-ground biomass (AGB) under different disturbance regimes. We applied a Landsat-based time series change detection approach to classify stand-replacing disturbances across the Alps for the period 1984-2024. Disturbance regimes were characterized using metrics such as event frequency, severity, return intervals and temporal trends. Disturbance regime parameters (probability scale, clustering degree and intensity slope) were derived by decade and used in a model inversion framework to assess AGB responses under different disturbance regimes. Our results indicate a marked intensification of disturbances in recent years. While disturbance peak years were synchronized between the western and the eastern Alpine ranges, the western Alps did not exhibit an increasing disturbance trend. In contrast, disturbance severity in the eastern Alps has significantly changed in the last decade compared to previous decades with both a mean rise in disturbances and a higher frequency of years characterized by an extreme number of disturbance events. Spatially, this increase was widespread across the eastern Alps and not confined to distinct hotspot areas. Outputs from dynamic carbon simulations showed that under the current disturbance regime (2014-2024) AGB can be reduced by 25% relative to the past disturbance regime (1984-1994), with convergence times between regimes spanning between 10 to 100 years. Overall, our findings provide robust observational evidence of an ongoing forest disturbance regime shift in the eastern Alps and demonstrate its substantial impacts on forest carbon dynamics. This work provides spatial and temporal information for understanding changes in carbon dynamics in alpine forests as well as an empirical foundation for improving disturbance-aware carbon modelling. The outcomes can inform adaptation and management strategies.
How to cite: Candotti, A., Wang, S., Carvalhais, N., and Tomelleri, E.: Recent intensification of forest disturbances in the Alps and implications for carbon dynamics, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12467, https://doi.org/10.5194/egusphere-egu26-12467, 2026.
Forest stand age is a key variable in modeling carbon storage and uptake in forest ecosystems. However, consistently estimating stand age across Europe remains challenging due to labor-intensive field measurements and heterogeneous estimation methods. In addition, the availability and coverage of stand age field data vary widely across countries, with some lacking data entirely. Here, we present a spatially explicit forest stand age dataset for 2010 across Europe at multiple spatial resolutions. For regions with field data, stand age was estimated by integrating national forest inventory data with satellite-based disturbance information. For regions without field data, we used machine-learning models combining climate variables, satellite-derived tree height, and other Earth observation metrics to generate continent-wide stand age estimates. The resulting datasets provide the spatial distribution of forest stand age across Europe at 0.5° resolution for continental-scale analyses and ~ 1 km² resolution for regional applications, including associated uncertainty estimates, highlighting regions where old-growth forests are more likely to persist. By leveraging one of the largest compilations of forest field observations together with Earth observation data, our approach substantially reduces uncertainty relative to previous spatially explicit stand-age products, enabling their use in ecosystem modeling, biodiversity conservation, and climate adaptation planning across multiple scales.
Funding acknowledgement:
CLIMB-FOREST Horizon Europe project (No. 101059888), European Union.
How to cite: Rodes-Blanco, M., Astigarraga, J., Pugh, T., and Ruiz-Benito, P. and the ClimbForest (stand age): Pan-European forest stand age estimates using field-based and earth observation data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12999, https://doi.org/10.5194/egusphere-egu26-12999, 2026.
Forests play a central role in climate change mitigation, yet large uncertainties persist in quantifying their future carbon sequestration potential, particularly under increasing disturbance pressure. Here we develop a spatially explicit, species-resolved assessment of present and future forest carbon dynamics in France by integrating high-resolution Earth observation products with national forest inventory data. Using 10–30 m maps of above-ground biomass, tree species composition, and disturbance history derived from satellite observations, we reconstruct species- and region-specific forest growth curves through a space-for-time approach, independently constrained using field measurements from the French National Forest Inventory. These growth curves are assimilated into a data-driven bookkeeping model that tracks annual biomass gains and losses from growth, harvest, and natural disturbances across metropolitan France from 2020 to 2050.
Our results indicate that, in the absence of disturbances, French forests could increase above-ground carbon stocks by up to 32% by 2050, driven primarily by the growth of young conifer stands. However, when realistic disturbance and harvest regimes extrapolated from recent trends are accounted for, net carbon accumulation is reduced to 23%, highlighting the strong moderating influence of unplanned disturbances and management. Carbon sequestration potential and resilience vary markedly across species and regions: mixed and deciduous stands generally exhibit greater robustness, whereas intensively managed conifers—particularly fir–spruce, Douglas fir, and maritime pine—are disproportionately vulnerable to combined disturbance and harvest pressures.
By explicitly representing forest demography, species composition, and disturbance regimes at high spatial resolution, our framework delivers a realistic projection of the French forest carbon sink to mid-century. These results provide actionable insights for forest-based mitigation strategies, revealing where carbon storage can be enhanced, where resilience is limited, and how current management trajectories may constrain future climate benefits.
How to cite: Pellissier-Tanon, A., Xu, Y., Ritter, F., Besic, N., and Ciais, P.: Integrating remote sensing-based maps of biomass and disturbances for future forest carbon sequestration scenarios in France, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18502, https://doi.org/10.5194/egusphere-egu26-18502, 2026.
Forest resilience under climate change depends on adequate seed production, yet it remains largely unclear how tree fecundity is changing and how this may constrain current and future range expansion. Using a long-term dataset spanning broad climatic gradients, we show that climate warming is already disrupting reproductive synchrony and fecundity across Europe.
In European beech (Fagus sylvatica), we analysed seed production records from 341 sites (mean record length 31.7 years) to test how climate change is altering masting dynamics. We show that increasing frequency of the main reproductive cue strongly erodes year-to-year variability and synchrony in seed production, particularly at colder, high-latitude and high-elevation sites. In these regions, masting variability has declined by up to ~54%, with SSP2.45 projections for 2070 indicating reductions of up to ~83%. With masting underpinning the production of pollinated, unpredated seeds, this challenges the assumption that cold-range margins are refugia from climate impacts and indicates that disruption of reproductive dynamics is likely to become the norm for this species.
Extending beyond a single species, we used a temporal attribution framework to analyse three decades of fecundity change in five dominant taxa (40,530 annual observations, 348 sites) for oaks (Quercus robur, Q. petraea), European beech, Scots pine (Pinus sylvestris), and silver fir (Abies alba). Across all species, viable seed production declined by 32–65%, with summer warming emerging as the dominant driver. Growing season drought and spring temperature had comparatively minor effects. Weather effects varied with climate, indicating diverging short-term (within-site) and long-term (across-site) sensitivities, and suggesting potential for local adaptation or acclimation.
Together, these results show that reproduction may emerge as a key bottleneck for forest resilience under climate change, as warming drives populations beyond their optimum reproductive niches. Integrating reproductive processes into forest projections and management is needed to avoid overlooking critical transitions in forest dynamics.
How to cite: Foest, J., Szymkowiak, J., and Bogdziewicz, M.: Future Forests Start with Seeds: Warming-Driven Disruption of Forest Fecundity, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5076, https://doi.org/10.5194/egusphere-egu26-5076, 2026.
Forest ecosystems are the main terrestrial carbon sinks and therefore play a critical role in global climate mitigation strategies. However, their long-term maximum capacity to store carbon under present and future climate conditions remains uncertain. To address this knowledge gap, we investigated the limits of above-ground biomass accumulation in existing forests and non-forested areas, and predicted how climate change scenarios could impact this carbon storage capacity. We evaluate the physical constraints on carbon accumulation, using a machine learning framework that integrates climate, pedologic, and hydrological parameters with natural forest disturbances, and CO₂ fertilization effects on forest growth. The results show that relying on forests to offset business-as-usual carbon emissions is possible only to a very limited extent. Projections of future forest carbon in existing forests remain highly uncertain and critically dependent on the methods and assumptions used for the assessments. Climate model-based estimates–which are often used in international policy and IPCC reports–suggest substantial future increases in carbon storage capacity, mainly driven by a strong CO₂ fertilisation effect. On the contrary, experimental and satellite-based evidence suggests much weaker increases or even a stabilisation of the terrestrial sink. Furthermore, our results suggest that ecosystem vulnerability to climate extremes will increase the frequency and severity of natural disturbances in the tropical and mid-latitude regions, ultimately reducing the effective long-term capacity of forests to store carbon. Potential increases in the carbon storage of boreal ecosystems could partially offset these losses, but only over longer timescales that cannot compensate for near-term declines elsewhere. These results highlight the need to reframe the potential of forest-based solutions for climate mitigation: not as an offset for anthropogenic carbon emissions, but as an essential buffer that helps prevent the land sector from becoming a net carbon source. For this purpose, forest conservation and sustainable management aimed at increasing ecosystem resilience to climate should be prioritised as transitional measures that support emission reductions, rather than substitutes for them.
How to cite: Roebroek, C., Caporaso, L., Cescatti, A., Duveiller, G., Davin, E., and Seneviratne, S.: Physical limits of carbon storage on land under present and future climate, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20254, https://doi.org/10.5194/egusphere-egu26-20254, 2026.
Phenological development in semi-arid landscapes is susceptible to climatic variability, making it a powerful early indicator of climate stress. The Bundelkhand region of Central India—characterised by erratic monsoon rainfall, frequent droughts, high atmospheric dryness, and fragile agroecosystems—offers a critical natural laboratory for assessing climate-driven phenological changes. This study develops an integrated framework that combines multi-sensor remote sensing, phenology–climate machine learning models, and bias-corrected CMIP6 projections to quantify how vegetation dynamics and crop productivity in Bundelkhand will evolve under future climate scenarios. A multi-year phenology record was constructed using Sentinel-2 NDVI/EVI (10 m) and MODIS MCD12Q2 phenometrics to derive start-of-season (SOS), end-of-season (EOS), length-of-season (LOS), and peak greenness. Temporal smoothing using harmonic regression and double-logistic models enabled robust extraction of phenological markers for croplands (rice, wheat, pulses), natural vegetation, and fallow systems. Historical climate variables—temperature extremes, monsoon onset variability, vapour pressure deficit (VPD), heatwave duration, and solar radiation—were obtained from IMD and NASA POWER datasets. Climate–phenology linkages were quantified using generalized additive models, Random Forest, LightGBM, and LSTM networks to capture nonlinear responses and climate lag effects.
Future projections were developed using five CMIP6 GCMs (ACCESS-CM2, MPI-ESM1-2-HR, MIROC6, NorESM2-LM, and FGOALS-g3) under SSP2-4.5 and SSP5-8.5 scenarios. Bias correction followed the ISIMIP3BAS protocol. ML-derived phenology models were then forced with CMIP6 futures to simulate phenological trajectories for the 2030s, 2050s, and 2080s. SHAP sensitivity analysis identified VPD, Tmax anomalies, and pre-monsoon rainfall deficits as dominant drivers controlling phenological timing in Bundelkhand’s water-limited environment. Results reveal region-wide advancement of SOS by 10–25 days, shortening of LOS by 6–20 days, and reductions in peak greenness due to compounded heat and moisture stress—particularly under SSP5-8.5. Projected declines in vegetation productivity range from 12–30%, with drought-prone districts (Tikamgarh, Chhatarpur, Mahoba, Hamirpur) emerging as phenological stress hotspots. These shifts threaten major rabi crops (wheat, gram, mustard) and already stressed natural vegetation. By integrating phenology–climate modelling, remote-sensing dynamics, and CMIP6 climate trajectories, this study provides a first-of-its-kind, high-resolution assessment of how Bundelkhand’s vegetation will respond to future climate change. The framework supports climate-smart agricultural planning, phenology-based early-warning systems, and long-term drought adaptation strategies in one of India’s most climate-vulnerable regions.
Keywords: Vegetation Phenology; Climate Change; Bundelkhand; CMIP6 Projections; SHAP sensitivity analysis
How to cite: Singh, P. N. and Pipil, S.: AI-Enabled Climate–Phenology Coupling and Future Productivity Assessment for Semi-Arid Bundelkhand under CMIP6 Forcings, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-22038, https://doi.org/10.5194/egusphere-egu26-22038, 2026.
The terrestrial biosphere plays a crucial role in regulating biogeochemical cycles, storing carbon, and providing essential ecosystem services, yet the land carbon sink remains a large source of uncertainty in future climate projections. Under the combined pressures of climate change, land use, pests, and fire, major biomes such as the Amazon and boreal forests face widespread degradation and may exhibit tipping behavior. The Tipping Points Modelling Intercomparison Project (TIPMIP) brings together the modelling community to develop unified simulation protocols for Earth System Models and standalone models, including Dynamic Global Vegetation Models. In this context, TIPMIP-BIO aims at running idealized experiments of climate overshoot and stabilizations scenarios, in order to assess climatic and deforestation thresholds that could trigger nonlinear biosphere responses. The framework provides opportunities to examine key processes and feedbacks, including CO₂ fertilization and physiological responses, and to explore emerging questions such as plant trait variability and the impact of compound climate events.
How to cite: Tardif, D., Harper, A. B., Hirota, M., Sakschewski, B., Georgievski, G., Licon Salaiz, J., Loriani, S., Dennis, D. P., Donges, J. F., and Winkemann, R.: TIPMIP-BIO: Towards a unified simulation protocol to assess potential tipping behavior in the biosphere and associated feedbacks and processes, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-15263, https://doi.org/10.5194/egusphere-egu26-15263, 2026.
Posters on site: Thu, 7 May, 08:30–10:15 | Hall X1
The atmosphere–land interaction is crucial to the climate and earth system through the exchange of energy, water, momentum and carbon among vegetation and atmosphere. In recent times, a great deal of variability in anthropogenic land use along with climate variability has greatly altered the terrestrial biosphere all around the globe. The global vegetation dynamics has garnered substantial attention due to its potential impact on food security, water cycle and terrestrial carbon sinks. The non-climatic factors have a very straightforward and regional impact on vegetation. However, there remains uncertainty regarding the response of the terrestrial ecosystems to climate change as vegetation-climate interactions is very intricate and intriguing. In the recent times, higher temperature (T) and evapotranspiration (ET) accompanied by insufficient precipitation (P) has depleted soil moisture (SM). We find temperature (T) is the dominant driver of global photosynthesis. Across biomes and land cover types, moisture availability (P and SM) is the key climatic control in tropical and arid but T in temperate and cold biomes. For croplands and forests, T is the predominant driver, but P is the key driver for grasses suggests Machine Learning (ML) based Random Forest (RF) model. However, there is decline in the control of temperature on photosynthesis due to saturation of boreal warming-induced greening and increasing dryness stress. The influence of water availability and energy has substantially grown on global photosynthesis. Interestingly, in regions where both increase in energy and decrease in water availability is present, the photosynthetic activity is largely moisture controlled. Therefore, the global photosynthesis is largely driven by moisture ahead of warmth and energy in the drying world.
How to cite: Kashyap, R. and Kuttippurath, J.: Unravelling the response of Global Vegetation to Climate Change during recent decades , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-958, https://doi.org/10.5194/egusphere-egu26-958, 2026.
On a global scale, human activities have significantly fragmented forest landscapes, resulting in over 20 % of the remaining forests being located within 100 meters of an edge. Forests adjacent to disturbances experience edge effects that can affect aboveground carbon storage through changes in forest structure. Our main objective was to assess aboveground carbon stocks and their drivers in edge forests across a large bioclimatic gradient of upland sites in the temperate and boreal forests of Eastern Canada, using powerline rights-of-way as a case study. We quantified the carbon stocks contained in living and dead trees of all sizes and measured tree growth at the stand and tree levels in forests adjacent to powerline rights-of-way. Compared with control forests (> 50 m from right-of-way), aboveground carbon stocks in edge forests (< 20 m) were up to 60 – 75 % higher in boreal spruce forests, 30 % higher in temperate forests, but only 2 % higher in boreal fir forests. Higher carbon stocks were linked to increased stand density, and thus a higher stand basal area, rather than larger tree diameters. Edge effects on tree characteristics (diameter, total height, crown length and area, and basal area increment) showed no clear pattern and depended on the characteristics of the forest. No edge effect was found in a stand with a recently established right-of-way (less than three years), suggesting that the magnitude of the edge effect varies over time. This study will improve the assessment of the carbon footprint of fragmented forest landscapes.
How to cite: Harel, A., Thiffault, E., Paré, D., Moreau, G., Achim, A., Leduc, F., Larochelle, M., and Chavaillaz, Y.: Higher aboveground carbon stocks in edge forests., EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-1507, https://doi.org/10.5194/egusphere-egu26-1507, 2026.
The terrestrial biosphere exchanges a large amount of CO2 with the atmosphere through photosynthesis and respiration, determining the magnitude of land carbon sink and consequently influencing the rate of global warming. The magnitudes of global photosynthesis and respiration, however, vary widely across models (100-200 PgC/year), constituting a key and persistent source of uncertainty in carbon cycle and climate modelling. Here, we argue that the uncertainty in the land carbon cycle modelling is largely attributable to the uncertainty in biogeography – the distribution of plant functional types (PFTs). Using an ensemble of dynamic global vegetation models (DGVMs), we find a strong dependence of total photosynthesis on total area for each PFT. The dependence allows us to reduce the spread of land carbon cycle estimates by ~75% using remote sensing-based PFT maps. We further find that 56 ± 21% of climate-driven changes in global photosynthesis modelled by DGVMs are caused by changes in PFT distribution in the last two decades. Our study identifies vegetation biogeography as a main controlling factor of uncertainty in land carbon cycle modelling and highlights the importance of biogeography-climate interactions in carbon cycle and climate studies.
How to cite: Luo, X., Zhao, R., Walker, A., Hoffman, F., and Koh, L. P.: Vegetation biogeography is a main source of uncertainty in modelling the land carbon cycle, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2488, https://doi.org/10.5194/egusphere-egu26-2488, 2026.
The Horizon Europe SELINA project (project-selina.eu) supports evidence-based decision-making for the sustainable use of natural capital by advancing the integration of information on biodiversity, ecosystem condition, and ecosystem services across Europe, contributing to the EU Biodiversity Strategy for 2030, the European Green Deal, and national ecosystem reporting.
As part of SELINA’s demonstration phase, advanced methods were tested to improve the spatial and temporal resolution of the three key ecosystem accounting components—extent, condition, and services — focusing on the capacity of forests, heathlands and peatlands to capture and store carbon on two test sites: São Miguel Island (Azores, Portugal) and Peloponnese, Greece.
Ecosystem extent mapping employed a dual approach: (i) national-centric, enhancing existing datasets with Copernicus Land Monitoring Service (CLMS) data; and (ii) vegetation-centric, classifying EUNIS habitats from remote-sensing-derived features. Forest condition was assessed using the PEOPLE-EA index, integrating multiple indicators relative to an optimal reference state. Carbon accounting also followed a dual approach: (i) a remote sensing–based method (GEDI, LiDAR, Sentinel-1) to map above-ground biomass, estimate carbon stock/fluxes, and detect deforestation; (ii) a GPP-based approach applying a Light Use Efficiency (LUE) model with Sentinel-2 and climate data, subsequently converted to biome-specific NPP.
These methods produced wall-to-wall 10 m resolution maps harmonized with the European ecosystem typology, enabling scalable, cost-effective, and policy-relevant ecosystem monitoring, particularly in typically underrepresented small-medium islands from the European outermost regions.
How to cite: Bernardo, F., Peters, A., Watson, M., Giagnacovo, L., Smets, B., Quiñones, M., Barbulo, D., Kooij, B., Gil, A., Dimopoulos, P., and Kokkoris, I.: Satellite data for carbon sequestration accounting: enhancing spatial and temporal resolution in harmony with the European ecosystem typology, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4586, https://doi.org/10.5194/egusphere-egu26-4586, 2026.
Forests are under increasing pressure from climate change, land-use transformation, and human activities, threatening their capacity to regulate climate, store carbon, and sustain biodiversity. Detecting long-term changes in forest canopy structure and productivity is therefore essential. Leaf Area Index (LAI) is a widely used proxy of vegetation structure and function and serves as a key indicator of forest productivity. It captures greening associated with canopy densification and potential carbon uptake, and browning linked to forest degradation, disturbance, or loss. However, global assessments of forest greening and browning and their spatial determinants remain limited.
Here, we present a global analysis of forest LAI trends over the last ~25 years using the High-Quality Reprocessed MODIS LAI dataset (HiQ-LAI; 8-day, 500 m; 2000–2024). The analysis is restricted to pure forest pixels derived from Hansen’s 30 m global forest cover map. Trends were quantified using Sen’s slope estimator and their significance assessed with the Mann–Kendall test. Globally, 21 % of forest areas exhibited significant greening, while 8 % showed browning, revealing strong regional contrasts. Browning patterns in tropical and subtropical forests are predominantly associated with land-use change, whereas in boreal regions they are largely driven by fire disturbances. In contrast, greening hotspots extend beyond climatic and CO₂ fertilization effects and strongly overlap with intensively managed forests, including plantation-dominated regions in China and Europe.
Our findings demonstrate that land-use change, forest management, and disturbance regimes are key spatial determinants of observed forest greening and browning patterns. However, these processes remain underrepresented in many Earth system models. By providing a spatially explicit global baseline, this study supports improved representation of human and disturbance processes in climate–vegetation modelling and informs conservation, climate mitigation, and sustainable forest management strategies under accelerating global change.
How to cite: Schauman, S., García, D., Aguilar, F., and Verger, A.: Global patterns of forest greening and browning: the imprint of land-use change, management, and fire, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9784, https://doi.org/10.5194/egusphere-egu26-9784, 2026.
Landsurface cover and hydrology is crucial information in modelling attempts on local to global scales. Arctic tundra regions are distinct in their plant communities and composition from Boreal or temperate regions and often underserved by globally available landcover data sets. Recent advances in mapping landcover units in Arctic tundra regions have allowed researchers to study and quantify landscape composition and wetness gradients on different scales. The Circumarctic Landcover Units (CALU) data set provides landcover information based on Sentinel – 1 and Sentinel – 2 at 10m resolution (ESA CCI+ Permafrost landcover). It also provides tundra specific landcover units not present in conventional data sets which improves the overall representation of tundra landscapes.
Global landcover data sets, such as the CCI Landcover, are available at coarser scales. To harness the potential of the CALU data set for usage in global applications, it is necessary to aggregate the 10m resolution to a coarser scale (300m in this case). In addition, the CCI Landcover includes different classes compared to the CALU data sets, which requires additional harmonization.
Here we present a harmonization effort between CALU and ESA CCI Landcover data sets. The result is a global landcover data set at 300m scale, with Arctic regions based on the dedicated CALU data. As a second step, novel classes in addition to the existing CCI Landcover classes are being suggested to improve representation of tundra landscapes in the harmonized data set. This harmonized landcover data set will allow for a more realistic representation of the Arctic landcover in land models which has the potential to improve the model estimate of the land carbon sink. Applications of the harmonized data set will be discussed.
How to cite: Bergstedt, H., Bartsch, A., Gruber, C., and Emmerichs, T.: Improved representation of Arctic tundra landcover in global data sets , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9880, https://doi.org/10.5194/egusphere-egu26-9880, 2026.
Variation in forest age can contribute to differences in biomass accumulation, making it useful for understanding forest recovery and carbon dynamics. In Europe, centuries of forest use, diverse management regimes, and disturbance histories have produced a heterogeneous age structure. Yet a temporally consistent, high-resolution, and spatially explicit forest-age dataset has been lacking for the continent. Here we reconstruct a long-term (1900–2019), annually resolved forest-age record for 38 European countries. To map high-stand forest age, we implement a 100 m backcasting framework initialized from a 2020 reference map and constrained by natural disturbances, rotation-based harvesting, and historical forest-area change. In parallel, we reconstruct coppice age at 5 km resolution for 1900–2010 to represent short-rotation management and its gradual conversion to high-stand forests. Validation against National Forest Inventory data shows good agreement for young and middle-aged forests (R² = 0.77 and 0.92, respectively). At the continental scale, high-stand forests experienced pronounced rejuvenation around the mid-twentieth century and have aged gradually since then. In contrast, former coppice forests shifted from widespread young stands (5–25 yr) in 1900–1950 to a smaller extent and an increasingly older age structure in later decades as coppice was progressively converted to high stand. This dataset provides the first Europe-wide, spatiotemporally consistent, annually resolved record of forest-age dynamics, supporting assessments of management legacies, recovery trajectories, and long-term carbon-cycle impacts.
How to cite: Ma, R., Ciais, P., Li, W., Pellissier-Tanon, A., Ritter, F., Xu, Y., Erb, K., Besnard, S., Xiao, J., Zhu, L., and Meng, N.: Reconstructing European forest age maps at 100 m resolution from 1900 to 2019, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10304, https://doi.org/10.5194/egusphere-egu26-10304, 2026.
Gross Primary Production (GPP) serves as a critical indicator of ecosystem function and its response to climate change. While numerous GPP estimates from different sources (model data, satellite derived, flux tower) exist, discrepancies between these datasets remain, emphasizing the need for new datasets and continuous improvement, particularly in understanding carbon-climate feedbacks and ecosystem resilience.
By refining the VODCA2GPP product [Wild et al., 2022], which estimates GPP using Vegetation Optical Depth (VOD) from microwave remote sensing, we present an enhanced version with four key methodological improvements.
First, we enhanced the spatial resolution from 0.25° to 0.1°, enabling finer scale detection of spatial heterogeneity in vegetation productivity and improving representation of local ecosystem dynamics. Additionally, we transitioned from X-band to Ku-band VOD observations due to their superior signal to noise ratio in the Mediterranean region, enhancing data quality while maintaining overall model performance.
Second, we integrated land cover information to improve model generalizability across different biomes, addressing the imbalanced distribution of in-situ validation stations and enhancing the model's ability to capture ecosystem specific carbon uptake patterns.
Third, we incorporated soil moisture data to account for water availability, which is the primary constraint on vegetation productivity in many biomes and is particularly crucial for understanding drought responses and ecosystem stress.
Fourth, we utilized ESA CCI Biomass observations to better capture biomass accumulation patterns.
The enhanced model was validated using an expanded set of in-situ measurements, including data from WARM Winter, AmeriFlux, JapanFlux, and CH4 datasets, which significantly extends our validation capabilities across different climatic zones and ecosystem types. Validation against FLUXNET in situ measurements and comparisons with leading datasets, including MODIS and FLUXCOM, demonstrate that the finer spatial resolution better captures local scale variability while maintaining strong model accuracy and reliability. This updated VODCA2GPP version offers a valuable resource for analyzing global vegetation dynamics, enabling better monitoring of ecosystem responses to environmental change and improving our understanding of the terrestrial carbon cycle.
This research has been funded through the GLANCE project.
References:
Bernhard Wild, Irene Teubner, Leander Moesinger, Ruxandra-Maria Zotta,
Matthias Forkel, and Wouter A Dorigo. Vodca2gpp–a new, global, long-
term (1988–2020) gross primary production dataset from microwave re-
mote sensing. Earth System Science Data, 14(3):1063–1082, 2022. doi:
10.5194/essd-14-1063-2022
How to cite: Müller, M., Zotta, R., Laluet, P., and Dorigo, W.: VODCA2GPP: High Resolution GPP Estimation in the Mediterranean basin from Vegetation Optical Depth Using Machine Learning, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10329, https://doi.org/10.5194/egusphere-egu26-10329, 2026.
A persistent challenge for models simulating carbon fluxes, such as gross primary productivity (GPP), is that the inter-annual variability (IAV) is currently not well-represented, often underestimating the peak GPP values, while also struggling with representing the onset and end of vegetation activity. We hypothesize that the difficulty with representing IAV can be attributed to temporally fixed model parameters, and yearly varying parameters can partially alleviate it. We test this hypothesis using two models: a simple light-use efficiency (LUE) model with response functions of solar radiation, air temperature, vapor pressure deficit, cloudiness, and soil water content, and an optimality-based model that includes parameter acclimation and drought stress. These functions have multiple parameters requiring calibration.
First, we calibrated all the model parameters per site-year and found that both models can simulate annual GPP better with annually calibrated parameters (median normalized Nash-Sutcliffe efficiency, viz. NNSE: 0.74 for the LUE model) compared to parameter calibration per site (median NNSE: 0.5) or per plant functional types (median NNSE: 0.23). Thereafter, we focused on calibrating parameters of one environmental response function as year-specific (one function at a time), while simultaneously calibrating year-invariant parameters for all other functions. These exercises were conducted for 198 eddy-covariance sites. The ability to represent IAV of GPP in arid sites was substantially improved when hydrological parameters were allowed to vary between years, both for herbaceous and forest ecosystems. However, for tropical, temperate and boreal climates, improvements in IAV emerged from parametric variability controlling the GPP responses to temperature, light or atmospheric dryness. Given the paucity of arid and semi-arid sites in the dataset, allowing year-specific parameters for vapor pressure deficit and atmospheric CO2 effects yielded a median annual NNSE of 0.73 across the whole dataset for the LUE model. These results challenge our perception on temporally static parameterizations, reflecting the need to learn the empirical relationships between observations and temporally-varying parameters, or improve the representation of missing state variables. It further suggests that these may be strongly linked to below-ground plant dynamics, largely unobserved in current Earth observation networks.
However, by analyzing mean absolute deviation of parameter values from per site and per site-year model calibrations, we found that temporal variation of parameters was lower than their spatial variation. For example, spatial variability of parameters, such as optimal temperature for photosynthesis, was 82.6% higher than temporal variability. Though we show that the temporal variability of model parameters is important to better capture the IAV of GPP flux, our analyses are currently limited to eddy-covariance sites, and only for the measurement periods at these sites.
As a next step, further research is needed to explain or statistically learn the temporal variability of model parameters using environmental variables, which can be used to predict the spatiotemporal variability of model parameters at sites with no observational data or predict the future temporal trend of model parameters. This, in turn, will likely improve the performance of simulated IAV of GPP and, consequently, enhance our ability to represent unknown linkages between IAV and longer time scales.
How to cite: De, R., Brenning, A., Reichstein, M., Šigut, L., Ruiz Reverter, B., Korkiakoski, M., Paul-Limoges, E., Blanken, P. D., Black, T. A., Gielen, B., Tagesson, T., Wohlfahrt, G., Montagnani, L., Wolf, S., Chen, J., Liddell, M., Desai, A. R., Koirala, S., and Carvalhais, N.: Towards Understanding the Inter-Annual Variation of Model Parameters Used to Simulate Gross Primary Productivity, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10408, https://doi.org/10.5194/egusphere-egu26-10408, 2026.
Forests play a central role in climate change mitigation through carbon sequestration and storage, yet spatial estimates remain highly uncertain due to variability in species composition, site conditions, and model representations of ecosystem processes. An example is presented as carbon stocks and sequestration rates are estimated for all forests in Germany using the process-based LandscapeDNDC model. This demonstrates the dependence on dominant tree species considering beech (Fagus sylvatica), oak (Quercus spp.), spruce (Picea abies), and pine (Pinus sylvestris), as well site conditions. Regarding the latter, the spatial resolution of 10 × 10 km enables to evaluation the role of soil fertility and water storage as well as precipitation and temperature effects.
The example also illustrates that the model is not able to represent the large carbon losses that have occurred during and after the extreme dry years 2018/2019 and neglects potential legacy effects. Therefore, it is suggested relating tree mortality and tree water deficit (TWD). Physically determined thresholds as well as theoretical concepts are available to principally derive the probability of mortality in cohort-based forest models. The new module provides a physically-based mechanism that very much depends on species-specific traits such as sapwood longevity and rooting intensity, but is robust against uncertainties in soil texture initialization. Furthermore, it is shown that it can easily be evaluated with micro-dendrometer data which are increasingly available.
How to cite: Grote, R., Moutahir, H., Ziegler, Y., Thurner, M., Kiese, R., and Ruehr, N.: Influence of Species Distribution, Site Conditions, and Model Structure on Carbon Stock and Sequestration Estimates in German Forests, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14847, https://doi.org/10.5194/egusphere-egu26-14847, 2026.
To mitigate damage from increasingly frequent and severe floods under climate change, the use of riparian corridors as flood buffers has been expanding. In particular, as part of Nature-based Solutions (NbS), growing efforts have focused on designing flood-buffer areas that reflect natural landscapes and geomorphic settings while actively incorporating natural attributes such as existing ecological conditions and vegetation distribution. In parallel, in response to carbon-neutrality policies, NbS-based flood-buffer spaces are increasingly being designed to include riparian vegetation communities—especially willows—so that these areas can also function as carbon sinks. Among willow species, Salix nipponica is the dominant riparian species in Korea; however, quantitative information on its age-dependent carbon dioxide uptake remains insufficient. Because carbon dioxide uptake can vary substantially with plant age, assessing uptake by growth stage or age class provides an important basis for characterizing carbon-sequestration potential and informing NbS design. In this study, we established a field-scale, grid-based cultivation system that enables annual destructive biomass sampling of Salix nipponica, with the ultimate objective of estimating age-dependent carbon dioxide uptake from the collected biomass data.
The field-scale grid-based cultivation system was established at the River Experiment Center of the Korea Institute of Civil Engineering and Building Technology using seeds of Salix nipponica collected from the Nakdonggang River. To identify growth characteristics, the basal diameter and height of sample trees were measured annually. Biomass for carbon dioxide uptake estimation was measured after uprooting sample trees and oven-drying to constant weight. Carbon dioxide uptake calculated from these directly measured biomass data corresponds to a measurement-based Tier 3 approach under the IPCC (Intergovernmental Panel on Climate Change) guideline. Associated with growth was estimated by applying the carbon fraction of dry matter and the mass ratio of CO₂ to C provided in the IPCC guideline.
The results of this study are presented for years 1–3 after planting in the cultivation system. Average individual biomass increased from 37 g (one-year) to 847 g (two-year) and 2,526 g (three-year), while average annual CO₂ uptake increased from 0.3 to 4 and 8 ton·ha⁻¹·yr⁻¹ over the same period. The increase in both biomass and average annual CO₂ uptake from one-year to two-year was larger than that from two-year to three-year. Compared with a previous study that estimated age-dependent CO₂ uptake for major forest tree species, CO₂ uptake in Salix nipponica was 1.2 times higher in two-year than that of same-aged softwood species and 1.3 times higher in three-year than that of same-aged hardwood species. The results quantitatively present the carbon sequestration potential of riparian willow stands in Korea. The identified age-dependent carbon dioxide uptake characteristics of Salix nipponica would demonstrate applicability to the development of river management strategies for climate change adaptation.
Acknowledge
This research was funded by the Korea Environment Industry & Technology Institute (KEITI) through the Smart Water-supply Service Research Program, funded by the Korea Ministry of Climate, Energy, Environment (MCEE)(RS-2022-KE002091).
How to cite: Park, Y., Ryu, J., and Ji, U.: Age-dependent CO2 uptake of Salix nipponica estimated from annual destructive biomass sampling in a field-scale grid-based cultivation system, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16156, https://doi.org/10.5194/egusphere-egu26-16156, 2026.
Wildfires are intensifying across boreal regions under climate warming, causing biomass losses, yet post-fire forest recovery and its drivers remain poorly quantified at large spatial scales in Russia. Here, we delineated forest fire disturbances by integrating multiple satellite-derived burned-area datasets with a consensus-based approach to reduce dataset uncertainties. We then applied a space-for-time substitution framework integrating forest fire disturbance, forest age, and aboveground biomass to quantify key forest recovery metrics for forest biomass recovery: AGBmax (potential maximum aboveground biomass), R30 (recovery rate at 0-30 yr), and T90 (recovery time to reach 90% of AGBmax). These metrics were validated against field measurements, and their environmental drivers were further explored using machine learning models. Forest fire hotspots were concentrated in central Russia, 151.3 Mha of cumulative burned area during 1985-2022. The median values of AGBmax, R30, and T90 were 99.7 Mg ha-1, 1.83 Mg ha-1 yr-1, and 127 yr, respectively. Spatial patterns were highly heterogeneous, with southern and western regions showing higher AGBmax and faster R30, while eastern and southwestern regions exhibited longer T90. Validation against field observations confirmed that the fitted curves closely reproduced observed recovery trajectories. Climate conditions, especially drought and low solar radiation, strongly constrain R30 and AGBmax, and while also prolonging T90. These findings enhance our understanding of post-fire forest resilience and the climatic controls on recovery across Russia, providing a robust foundation for future assessments of ecosystem carbon dynamics.
How to cite: meng, N., xu, Y., pellissier-tanon, A., ciais, P., and viovy, N.: Unraveling Post-fire Forest Recovery in Russia: Spatial Heterogeneity and Climatic Constraints, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18460, https://doi.org/10.5194/egusphere-egu26-18460, 2026.
Forest disturbances are fundamental drivers of accelerated ecosystem carbon turnover and massive carbon storage losses. At the landscape level, disturbance regimes characterized by their spatial extent (μ), frequency (α), and intensity (β) of disturbance histories, together with background mortality (Kb), are essential for understanding forest carbon sink dynamics. However, global patterns of these disturbance regimes and their climatic drivers, as well as their corresponding impacts on regional and global carbon budgets in the context of climate change, remain poorly understood.
Building upon our recently developed global dataset of these disturbance regimes data at 0.25° resolution, derived from high-resolution biomass observations (Wang et al., under review), this study aims to resolve the following three questions: (1) What disturbance characteristics have forests in different regions of the world experienced historically? (2) How do climate change and climate extremes influence these observed disturbance regimes? (3) To what extent do these disturbances directly or indirectly alter regional and global carbon budgets?
First, we applied K-Means clustering to the disturbance regime data using multivariate similarity. The optimal numbers of clusters were determined by the Elbow Method, allowing us to classify global ecosystems into 12 distinct disturbance-regime groups. The largest cluster (17.83%) is primarily distributed in temperate regions, while specific biomes, such as wet tropical forests, are dominated by 3 clusters (14%) characterized by relatively high extent and frequency but low-intensity disturbance regimes. These groups exhibit strong spatial coherence, closely mapping onto distinct biomes and climate zones.
Second, we developed cluster-specific Random Forest models to assess the primary climatic drivers associated with these regime types. Integrating ERA5 reanalysis data, we examined both long-term climatic means and a suite of extreme indices (e.g., heatwaves, precipitation anomalies). The feature importance of these variables reveals the distinct hierarchy of climatic drivers for each regime. This analysis differentiates the influence of baseline climatic conditions from extreme events, helping to identify the specific environmental factors most strongly associated with disturbance dynamics in different global regions.
Third, we plan to examine the potential implications of these regimes for the carbon cycle. We will analyze the regional carbon budgets from atmospheric inversions, linking them to disturbance regime characteristics, specifically investigating how shifts in disturbance intensity and frequency relate to regions transitioning between carbon sources and sinks.
This study systematically addresses how global ecosystems can be functionally grouped by their disturbance regimes, identifies the specific climatic factors driving these patterns, and quantifies the impact of regime shifts on the carbon budget. By linking these elements, our findings provide essential empirical constraints for Earth System Models (ESMs), particularly for representing the stochastic nature of disturbances and predicting their feedback to the global carbon cycle in a changing climate.
How to cite: Wang, S., Yang, H., and Carvalhais, N.: Global Disturbance Regimes: Patterns, Climatic Drivers, and Carbon Budget Implications, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20906, https://doi.org/10.5194/egusphere-egu26-20906, 2026.
Terrestrial biosphere models are central to quantifying the global land carbon sink, with the TRENDY ensemble of Dynamic Global Vegetation Models (DGVMs) providing the land-surface estimates underpinning the annual Global Carbon Budget. Despite their widespread use, TRENDY models exhibit well-documented biases in simulated leaf area index (LAI), including errors in both magnitude and phenological phase, which propagate into uncertainties in carbon and water flux estimates.
Deep-learning emulators of Earth system models are increasingly adopted to improve computational efficiency, yet their potential as a structured mechanism for integrating process-based and data-driven approaches remains underexplored. Here, we use a deep-learning emulator not only to reproduce TRENDY ensemble behaviour, but also as a controlled framework to correct inherited LAI biases using observations, without discarding underlying process relationships. We first pre-train an emulator on the TRENDY ensemble mean across 14 carbon- and water-related key variables, and subsequently apply transfer learning using satellite-derived LAI observations. Training across all four TRENDY factorial experiments isolates the causal effects of CO₂ fertilisation, climate change and variability, and land-use change, thereby expanding the training space and improving extrapolation potential. The emulator uses a transformer architecture and is formulated as a point model, run independently at each location, with temporal memory carried only through autoregressively propagated state variables.
Emulating the TRENDY ensemble mean is largely successful. High accuracy is achieved for non-disturbance, deterministic fluxes (mean R² = 0.94), including gross and net primary production, ecosystem respiration, evapotranspiration, and surface runoff. State variables, including carbon pools, soil moisture, and LAI, show a modest reduction in performance due to autoregressive drift, but remain well constrained (mean R² = 0.87). In contrast, disturbance-related fluxes—specifically fire and land-use change emissions—are reproduced with substantially lower skill (mean R² = 0.27). The emulator accurately reproduces the effects of CO₂ fertilisation and climate change and variability across scenarios.
Transfer learning substantially reduces LAI errors in magnitude, phase, and spatial distribution, decreasing the mean LAI bias from 1.13 in the TRENDY ensemble mean to 0.01, while maintaining performance across other variables. The resulting emulator provides a highly computationally efficient predictor of land-surface dynamics, with improved LAI–evapotranspiration and LAI–gross primary production relationships relative to observations. This work highlights the potential of deep learning as a controlled bridge between process-based and data-driven land-surface modelling, with potential to extend the work toward multiple observational constraints.
How to cite: O Cathain, E. N., Winkler, A., and Reimers, C.: TRENDY-Emulator: A Bias-Corrected Deep Learning Emulator of Terrestrial Carbon and Water Dynamics, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-21719, https://doi.org/10.5194/egusphere-egu26-21719, 2026.
Shrub vegetation has been expanding across Arctic regions in response to climate warming. However, its effects on terrestrial carbon cycling remain poorly understood. Moreover, shrubs are often underrepresented in land surface models. Here, we incorporated two shrub functional types—deciduous and evergreen, which exhibit distinct strategies from trees—into the nutrient-enabled QUINCY model.
The updated model simulates gross primary production (GPP) at site level in broad agreement with recent observations. Model simulations suggest that shrub expansion into needle-leaved forests or grasslands increases GPP by an average of 13% and 40%, respectively. Shrubs also produce substantial above-ground biomass—lower than in needle-leaved forests but higher than in grasslands—with these differences partly driven by both CO₂ and climate effects.
Our analyses reveal a strong model sensitivity to nitrogen availability. While the model applying unlimited nitrogen supply leads to an overestimation of maximum GPP at most study sites, activating the interactive nitrogen cycle suppresses GPP by up to 50% relative to flux-based observational constraints (ABCfluxnet). An additional sensitivity experiment, introducing permafrost nutrient inputs, improves soil carbon estimates but still results in GPP overestimation.
How to cite: Emmerichs, T., Lacroix, F., Brovkin, V., Gong, C., Zaehle, S., Voigt, C., Steenberg Larsen, K., Sjogersten, S., and Tuittila, E.-S.: Impacts of Shrub Coverage for modeled arctic Ecosystem Carbon Uptake and Storage, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14440, https://doi.org/10.5194/egusphere-egu26-14440, 2026.
Coastal forests are carbon-dense ecosystems providing critical services, including storm protection, water filtration, timber, and fisheries, to millions of people living in vulnerable low-lying regions. Increasing evidence suggests that tree mortality is accelerating across these habitats, driven by rising water tables (leading to anoxia), salinization, and extreme weather events such as droughts and storms. These stressors are triggering rapid shifts in coastal forest structure, function, and carbon balance. Yet, the spatial extent of coastal tree mortality remains poorly mapped due to the heterogeneity of coastal landscapes. Consequently, it becomes challenging to elucidate the pace and mechanisms behind such die-off patterns, particularly in areas beyond sea-level rise hotspots. Here, we present a deep learning–based approach for tracking individual tree mortality biennially between 2010 and 2023 using sub-meter aerial imagery across the coastal United States, spanning the Atlantic, Gulf, and Pacific coasts, as well as the Great Lakes region. By tracking over 200 million individual tree mortalities over the past decade, we captured signals of canopy stress and decline across at-risk forests, which enabled us to elucidate the underlying mortality drivers. We identified many mortality hotspots not captured by traditional remote sensing approaches or surveys. This approach offers a scalable framework for identifying emerging mortality hotspots and understanding how climate and hydrological stressors are reshaping forest resilience. Such insight is crucial for informing adaptive coastal management and anticipating ecosystem transformation under accelerating climate change.
How to cite: Yeung, C. H. and Yang, X.: Drivers of individual tree mortality across the US coasts , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2872, https://doi.org/10.5194/egusphere-egu26-2872, 2026.
Tropical forests play a pivotal role in the global carbon cycle, but the lack of long-term in-situ datasets renders our understanding of the specific carbon dynamics in tropical forests uncertain. In this study, we analyzed two decades (2003–2022) of eddy-covariance measurements from a primary tropical seasonal rainforest reserve in Xishuangbanna, southwest China, to characterize long-term trends in gross primary productivity (GPP), ecosystem respiration (Reco), and net ecosystem productivity (NEP). The protected rainforest functioned as a modest but steadily strengthening carbon sink (annual mean NEP = 157.9 ± 56.7 g C m⁻² year⁻¹, growth rate = 3.4% year⁻¹), consistent with an observed increase in carbon use efficiency (CUE) (annual mean CUE = 5.9% ± 1.8%, growth rate = 2.4% year-1), which reflects increasingly efficient carbon utilization and aligns with rising aboveground biomass. The enhancement of the interannual carbon sink was mainly driven by increasing GPP (mean = 2658.1 ± 254.5 g C m⁻² year⁻¹, growth rate = 1.0% year⁻¹). With the same 6-month duration, the tropical seasonal rainforest exhibited a stronger carbon sink during the dry season (148.3 g C m-1 season-1) than during the rainy season, with the dry season accounting for 93.9% of the annual carbon sink. The enhanced dry season radiation and precipitation throughout the two decades positively affected the upward trend of the carbon sink. Notably, the annual carbon sink showed a temporary decline approximately two years after droughts, suggesting a lagged ecosystem response to climatic disturbances. Overall, these findings underscore the long-term carbon sequestration potential of well-preserved tropical rainforests and provide critical empirical evidence for improving carbon budget assessments in tropical regions under ongoing climate change.
How to cite: Liu, Y., Jiao, L., Zhang, J., Li, X., Zheng, H., Sawasdchai, B., Chen, Y., Zhang, Y., Gnanamoorthy, P., and Song, Q.: Increase in carbon sink in a protected tropical seasonal rainforest in southwestern China over 20 years, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-527, https://doi.org/10.5194/egusphere-egu26-527, 2026.
Rising tree mortality threatens the forest carbon sink and may destabilize the global carbon cycle. Because tree mortality is driven by diverse disturbance agents (e.g., drought, insects, wind, fire), its impacts on above-ground biomass carbon (AGC) and its controlling factors vary widely and remain poorly constrained. Based on a consistent, multi-decadal record of biomass carbon from radar backscatter, we quantify AGC losses associated with specific disturbance agents in conterminous USA. We then develop agent-specific machine-learning models that relate mortality-driven AGC dynamics to climatic drivers (e.g., temperature, moisture deficits), vegetation characteristics, and soil properties. This framework could reveal how sensitivities and thresholds differ among agents and identify regions and conditions where forests are most vulnerable to mortality-related carbon losses. Our results support improved representation of tree mortality processes and associated carbon fluxes in Earth System Models, strengthening projections of present and future vegetation dynamics and carbon stocks under changing disturbance regimes.
How to cite: Liu, G., Tao, S., Eifler, L., Ciais, P., and Bastos, A.: Drivers and patterns of forest above-ground biomass carbon losses in the conterminous USA, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14168, https://doi.org/10.5194/egusphere-egu26-14168, 2026.
Posters virtual: Tue, 5 May, 14:00–18:00 | vPoster spot 2
EGU26-18930 | ECS | Posters virtual | VPS5
Improved Estimation of Gross Primary Productivity in Global Croplands Using a Transpiration-Based uWUE ModelTue, 05 May, 15:21–15:24 (CEST) vPoster spot 2
Accurate estimation of gross primary productivity (GPP) in croplands is essential for quantifying terrestrial carbon uptake and understanding carbon–water coupling under increasing agricultural water stress. Conventional Light Use Efficiency (LUE) models typically rely on evaporative fraction (EF), derived from total evapotranspiration (ET), which does not distinguish between productive transpiration and non-productive evaporation. In contrast, transpiration-based framework explicitly represent the physiological coupling between carbon assimilation and water loss regulated by stomatal conductance. In this study, transpiration is estimated using a Leaf Area Index (LAI)-based approach driven by remotely sensed MODIS data and environmental variables within the underlying Water Use Efficiency (uWUE) framework.
We evaluate the transpiration-based uWUE model against an EF-based LUE model for GPP estimation using eddy covariance observations from 51 globally distributed cropland sites. The dataset includes 6 sites from India (Flux Tower and INCOMPASS networks), 3 sites from Japan (AsiaFlux), 9 sites from Europe, and 33 sites from the United States (FLUXNET), spanning a wide range of hydro-climatic and management conditions. Model performance was assessed using the coefficient of determination (R²), root mean square error (RMSE), and bias.
The transpiration-based uWUE model showed overall better agreement with observed GPP than the EF-based LUE model across the global set of crop sites. Improvements were evident in both the strength of the relationship with observations and the reduction of estimation errors. At the site level, uWUE more frequently achieved higher R² together with lower RMSE, demonstrating consistent performance across multiple evaluation metrics at a larger number of sites. Superior performance was observed at 28 sites, driven by the model’s ability to capture coupled carbon–water dynamics under varying crop types, canopy structures, and climatic conditions. In contrast, the EF-based LUE model showed advantages at a limited number of sites characterized by distinct water stress regimes or vegetation properties.
Overall, the results highlight the critical role of transpiration dynamics in GPP estimation, with higher GPP values associated with dense canopies and favorable environmental conditions. By explicitly isolating transpiration from total evapotranspiration, the uWUE framework provides a more physically meaningful representation of carbon–water interactions than ET-based approaches. These findings demonstrate that incorporating transpiration-based constraints improves GPP estimation in croplands and has important implications for large-scale agricultural carbon cycle assessments under future climate scenarios characterized by increased water stress and drought frequency.
How to cite: Harde, S. and Rajasekaran, E.: Improved Estimation of Gross Primary Productivity in Global Croplands Using a Transpiration-Based uWUE Model, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18930, https://doi.org/10.5194/egusphere-egu26-18930, 2026.
