Global greening and browning, as evidenced by changes in leaf area index (LAI) derived from satellite observations, indicate how ecosystems respond to rising atmospheric CO 2 , climate change, and human interventions. However, uncertainties in satellite LAI records have led to conflicting conclusions about global trends and their drivers. Here, by developing a refined Advanced Very High-Resolution Radiometer LAI dataset with reduced radiometric and geometric uncertainties, we find a sustained global greening during 1982–2021 (0.035 m 2 m –2 decade –1 ). Global greening is largely driven by the continuous CO 2 fertilization (1982–2001: 72%; 2002–2021: 77%), while land-use management determines regional patterns. An ensemble of 15 Earth system models also simulates persistent greening (0.050 ± 0.044 m 2 m –2 decade –1 ), but fails to reproduce the observed spatial patterns of greening and drivers. These findings provide observational evidence for sustained global greening, with no clear signs of slowing CO 2 fertilization effect under current CO 2 and climatic conditions.

The sapwood area supporting a given leaf area (Huber value, vH) reflects the coupling between carbon uptake and water transport and loss at a whole-plant level. Geographic variation in vH presumably reflects plant strategic adaptations, but the lack of a general explanation for such variation hinders its representation in vegetation models and assessment of its impact on the global carbon and water cycles. Here we develop a simple hydraulic trait model to predict optimal vH by matching stem water supply and leaf water loss, and test its performance against two extensive plant hydraulic datasets. We show that our eco-evolutionary optimality-based model explains nearly 60% of global vH variation in response to light, vapour pressure deficit, temperature and sapwood conductivity. Enhanced hydraulic efficiency with warmer temperatures reduces the sapwood area required to support a given leaf area, whereas high irradiance (supporting increased photosynthetic capacity) and drier air increase it. This study thus provides a route to modelling variation in functional traits through the coordination of carbon uptake and water transport processes.

Vegetation cover regulates the exchanges of energy, water and carbon between land and atmosphere. Remotely-sensed fractional absorbed photosynthetically active radiation (fAPAR), a land-surface greenness metric, depends on carbon allocation to foliage while also controlling photon flux for photosynthesis. Greenness is thus both a driver and an outcome of gross primary production (GPP). An equation with just two (globally) fitted parameters describes annual maximum fAPAR (fAPARmax) as the smaller of a water-limited value, transpiring a constant fraction of annual precipitation, and an energy-limited value, maximizing annual plant growth. This minimalist description reproduces global greenness patterns, and the consistent temporal trends among remote-sensing products, as accurately as the best-performing dynamic global vegetation models. Widely observed greening is attributed to the influence of rising carbon dioxide on the light- and water-use efficiencies of GPP, augmented by wetting in some dry regions and warming in high latitudes. Limited regions show browning, attributed to drying.

There is growing concern over future trajectories of burning on Earth. One the one hand, some regions have seen the emergence of large and novel wildfires, whilst satellite observations continue to show declining burnt area globally, most notably in the tropics. Quantifying the response of global wildfire regimes to future changes in especially challenging given that wildfires are driven by climate, vegetation, and human activities, and that these different factors may have contrasting and opposing effects. Using global empirical models of burnt area, fire size and fire intensity we explore the trajectory of future fire regimes under high and low climate change mitigation efforts. The models are driven by lightning ignitions, climate, vegetation properties, topography, and human factors. Making use of a set of sensitivity analysis, we show a global shift in wildfire patterns by the end of the 21st century even with warming kept below 1.5°. Burning will generally be reduced in tropical regions but larger and more intense wildfires will occur in extra-tropical regions. Under low mitigation, increases in burnt area worldwide overwhelm the human-driven decline, with up to a 60% increase in burnt area by the end of the century. However, fire size and intensity will be increasingly limited by dryness and vegetation fragmentation. These results suggest that even under high climate change mitigation, fire management strategies must urgently be revised as current fire-suppression policies will no longer be effective in much of the world. Regional-level fire management, led by local stakeholders, should be encouraged. Wildfire risk and management must also be incorporated into mitigation scenarios that rely on extending forest area if these mitigation scenarios want to remain realistic.

Realistic projections of future wildfires need to account for both the stochastic nature of climate and the randomness of individual fire events. Here we adopt a probabilistic approach to predict current and future fire probabilities using a large ensemble of 1,600 modelled years representing different stochastic realisations of the climate during a modern reference period (2000–2009) and a future characterised by an additional 2°C global warming. This allows us to characterise the distribution of fire years for the contiguous United States, including extreme years when the number of fires or the length of the fire season exceeded those seen in the short observational record. We show that spread in the distribution of fire years in the reference period is higher in areas with a high mean number of fires, but that there is variation in this relationship with regions of proportionally higher variability in the Great Plains and southwestern United States. The principal drivers of variability in simulated fire years are related either to interannual variability in fuel production or atmospheric moisture controls on fuel drying, but there are distinct geographic patterns in which each of these is the dominant control. The ensemble also shows considerable spread in fire season length, with regions such as the southwestern United States being vulnerable to very long fire seasons in extreme fire years. The mean number of fires increases with an additional 2°C warming, but the spread of the distribution increases even more across three quarters of the contiguous United States. Warming has a strong effect on the likelihood of less fire-prone regions of the northern United States to experience extreme fire years. It also has a strong amplifying effect on annual fire occurrence and fire season length in already fire-prone regions of the western United States. The area in which fuel availability is the dominant control on fire occurrence increases substantially with warming. These analyses demonstrate the importance of taking account of the stochasticity of both climate and fire in characterising wildfire regimes, and the utility of large climate ensembles for making projections of the likelihood of extreme years or extreme fire seasons under future climate change.

From heightened canopy dieback to tree die-off, many forest ecosystems are showing signs of poorly coping with more severe, more frequent, or hotter droughts. Understanding forest resilience to drought has become paramount, and eco‐physiological optimisation approaches that test behavioural hypotheses have been proposed as a means to build this understanding in global terrestrial models. Here, we used a land-surface model that considers competing optimality principles to simulate canopy gas exchange and leaf nitrogen investments into the photosynthetic apparatus, whilst also accounting for sustained hydraulic impairment (Sabot et al., 2022). We applied this model to a pristine observational site of the Caatinga, Brazil’s drought-hardy, seasonally deciduous, and exceptionally diverse dry tropical forest. Six woody species dominate 80% of the study area whilst displaying contrasting functional strategies – for example, their respective P50s (the water potential at which 50% of a plant’s hydraulic conductivity is lost) range between -1 MPa and -5 MPa. Model predictions were assessed against species-specific leaf-level observations of stomatal conductance and photosynthetic uptake, as well as eddy covariance measurements of ecosystem carbon and water fluxes spanning a period with high interannual rainfall variability (and including a severe multi-year regional drought). We found that none of the six species could, in isolation, explain the magnitude and dynamics of the observed surface fluxes. However, taken together and accounting for their relative contribution to total ecosystem fluxes, they did. Further, our analysis emphasises the vital role of phenology in mitigating seasonal and inter-annual hydraulic risks, with foliage reductions triggered by a 10 to 20% loss of hydraulic conductivity in the canopy. On the whole, accounting for diverging species-level responses and their relative influence at the ecosystem-scale appears key to improving model predictions in functional diverse forests. Reference: Sabot, M.E.B., De Kauwe, M.G., Pitman, A.J., Ellsworth, D.S., Medlyn, B.E., Caldararu, S. et al. (2022) Predicting resilience through the lens of competing adjustments to vegetation function. Plant, Cell & Environment, 45, 2744–2761.

The TRY initiative (https://www.try-db.org) was established in 2007 on request from IGBP and DIVERSITAS to develop a joint database of in-situ measured plant traits supporting vegetation modelling and biodiversity research. Based on the mandate from the two initiatives, TRY has received significant contributions of original and integrated datasets from the global research community and achieved unprecedented coverage. The TRY database is regularly updated, and since 2019, data have been publicly available under a CC BY license. Trait data from the TRY database are frequently used to map plant traits at global scales. We here provide an overview of the TRY database, briefly summarise trait mapping approaches highlighting caveats, and provide an outlook on upcoming developments in the context of the TRY database.

Annual wildfire occurrences are associated with a high degree of variability. As well as arising from the inherent randomness of wildfire events, this is also due to variability in the climate factors affecting wildfire risk, such as to summer precipitation. With climate change linked to the emergence of regionally catastrophic fire years, understanding the probabilistic distribution of wildfires and the extent these are linked to predictable modes of climate variability (such as El Niño Southern Oscillation, ENSO, or the Atlantic Multidecadal Oscillation, AMO) is of increasing importance. We use a large climate ensemble (KNMI-LENTIS) together with probabilistic fire occurrence model accounting for human, vegetation type, vegetation growth, and weather effects to predict 1600 simulated fire years over the contiguous US in the modern climate (2000-2009) and for +2°C global warming. There is significant spread in the distribution of fire years in the modern ensemble, with interannual variability higher in regions with a high mean rate of fire activity. Controlling for the effect of the average fire rate, the southwestern US, the Great Plains and southern Florida have proportionally highest variability. Wildfire occurrence is strongly influenced by climate modes in all three of these regions in the ensemble - with greater wildfire occurrence associated with La Niña, negative Indian Ocean Dipole (IOD), and positive Tropical North Atlantic (TNA) years. The AMO, Pacific Decadal Oscillation and Pacific/North American oscillation all exert a significant influence on US wildfire in the modern and modern +2°C climates. Climate warming results in a considerable increase in annual wildfire occurrences across the US, including in less fire-prone regions of the northern and interior US, as well as a strong effect on the likelihood of extreme fire years and long fire seasons in the southwest. There is a strengthening effect of key climate modes on annual wildfires, especially from the AMO, IOD, TNA and ENSO. This analysis, in addition to specific findings concerning US wildfire, highlights the utility of large climate ensembles in characterising the variability of the wildfire regime and projecting wildfire under future climate change.

The time taken for ecosystems to recover after wildfire affects the rate of carbon sequestration, and this in turn impacts land-atmosphere exchanges and hydrological processes. Factors affecting post-fire recovery time have been investigated at site or regional scale, but there is comparatively little information about this at a global scale. In this study, we use solar-induced chlorophyll fluorescence (SIF) to estimate the recovery of photosynthetic activity after fire for more than 10,000 fires representing the range of ecosystems across the globe. We then examined the factors that influence post-fire recovery time, initially using the relaxed lasso technique to identify the most important factors and then using a linear regression model incorporating these factors. We show that vegetation characteristics, the characteristics of the fire, and post-fire climate all influence recovery time. Gross primary production (GPP) is the most important factor, with faster recovery in ecosystems with higher GPP. Fire properties which indicate substantial vegetation damage, such as fire intensity and duration, result in longer recovery times. Post-fire climate also affects recovery time: anomalous temperature and temperature seasonality, and higher than normal dry days increase recovery time while higher-than-average precipitation decreases recovery time. There is an additional impact of vegetation type (biome), which may reflect differences in plant adaptations to fire between biomes. We show that there is a clear relationship between the proportion of plants that resprout after fire in a biome and recovery time, with ecosystems characterised by higher abundance recovering faster.

Carbon allocation is a critical process that helps to optimize plant growth and significantly impacts ecosystem structure and function, with immediate implications for the global carbon cycle. Since leaves are the primary organs regulating the exchange of CO₂, energy, and water between terrestrial ecosystems and the atmosphere, accurate simulation of leaf carbon allocation is important. However, most land surface models (LSMs) lack detailed consideration of the leaf economics spectrum, which is represented by the coordination of leaf mass per area (LMA) and leaf longevity (LL) and how their relationship varies with the growth environment. Instead, LSMs commonly predict the proportion of biomass allocated to leaves either directly from the environment based on resource limitation theories, or from their functional relationships with other organs. A new theory based on eco-evolutionary optimality principles successfully predicts changes in LMA and LL with the environment by maximizing the average net carbon gain over the leaf life cycle. In addition, a prognostic, globally applicable Leaf Area Index (LAI) model has been developed recently, using climate data alone to capture LAI dynamics across biomes on the principle that the annual cycle of leaf display is closely related to the cycle of potential primary production by those leaves. Here, combining these two theoretical developments, we provide a universal expression for the proportion of biomass allocated to leaves. We successfully predict foliar carbon allocation as measured on a site basis and capture the plasticity of foliar carbon allocation with environmental change. The global average fraction of biomass production allocated to leaves is estimated as 0.31. The model also accounts for the different regulatory mechanisms of leaf carbon allocation by deciduous and evergreen plants.

Hydrology has been guided by establishing empirical relationships between the movement of water through landscapes and the application of the conservation of mass law in catchments. This has resulted in models with complex calibration frameworks that often overlook the physical and biochemical water-related processes linking plants to hydrological cycles. Studies have revealed that some of the empirical relationships in catchments might also reflect a potential ecosystem's coevolution with climate, driving catchments to optimise their supply and demand limits. This agrees with the eco-evolutionary optimality principles used in vegetation modelling that are based on the hypothesis that canopy conductance acclimates to environmental variations by balancing the costs of carbon assimilation and maintenance of transpiration rates. Here, we developed meaningful interfaces between simple models and approaches based on the use of optimality principles in vegetation modelling and hydrology. Our work is based on the application of the P-model to estimate to quantify gross primary productivity and transpiration and the use of a mass-balance approach to quantify the root zone storage. These integrations not only provide a more nuanced understanding of hydrological processes but also pave the way for more accurate and physically-informed models in hydrology. Our findings underscore the potential of using eco-evolutionary principles as a unifying framework in hydrological research, offering new insights for understanding and predicting water movement in catchments under varying climatic and ecological conditions.

Predicting the extent of massive wildfires once ignited is essential to reduce the subsequent socioeconomic losses and environmental damage, but challenging because of the complexity of fire behaviour. Existing physics-based models are limited in predicting large or long-duration wildfire events. Here, we develop a deep-learning-based predictive model, Fire-Image-DenseNet (FIDN), that uses spatial features derived from both near real-time and reanalysis data on the environmental and meteorological drivers of wildfire. We trained and tested this model using more than 300 individual wildfires that occurred between 2012 and 2019 in the western US. In contrast to existing models, the performance of FIDN does not degrade with fire size or duration. Furthermore, it predicts final burnt area accurately even in very heterogeneous landscapes in terms of fuel density and flammability. The FIDN model showed higher accuracy, with a mean squared error (MSE) about 82% and 67% lower than those of the predictive models based on cellular automata (CA) and the minimum travel time (MTT) approaches, respectively. Its structural similarity index measure (SSIM) averages 97%, outperforming the CA and FlamMap MTT models by 6% and 2%, respectively. Additionally, FIDN is approximately three orders of magnitude faster than both CA and MTT models. The enhanced computational efficiency and accuracy advancements offer vital insights for strategic planning and resource allocation for firefighting operations.

Leaf phenology, represented at the ecosystem scale by the seasonal dynamics of leaf area index (LAI), is a key control on the exchanges of CO2, energy, and water between the land and atmosphere. Robust simulation of leaf phenology is thus important for both dynamic global vegetation models (DGVMs) and land-surface representations in climate and Earth System models. There is no general agreement on how leaf phenology should be modeled. However, a recent theoretical advance posits a universal relationship between the time course of "steady-state" gross primary production (GPP) and LAI-that is, the mutually consistent LAI and GPP that would pertain if weather conditions were held constant. This theory embodies the concept that leaves should be displayed when their presence is most beneficial to plants, combined with the reciprocal relationship of LAI and GPP via (a) the Beer's law dependence of GPP on LAI, and (b) the requirement for GPP to support the allocation of carbon to leaves. Here we develop a global prognostic LAI model, combining this theoretical approach with a parameter-sparse terrestrial GPP model (the P model) that achieves a good fit to GPP derived from flux towers in all biomes and a scheme based on the P model that predicts seasonal maximum LAI as the lesser of an energy-limited rate (maximizing GPP) and a water-limited rate (maximizing the use of available precipitation). The exponential moving average method is used to represent the time lag between leaf allocation and modeled steady-state LAI. The model captures satellite-derived LAI dynamics across biomes at both site and global levels. Since this model outperforms the 15 DGVMs used in the TRENDY project, it could provide a basis for improved representation of leaf-area dynamics in vegetation and climate models.

Realistic simulation of leaf photosynthetic and respiratory processes is needed for accurate prediction of the global carbon cycle. These two processes systematically acclimate to long‐term environmental changes by adjusting photosynthetic and respiratory traits (e.g., the maximum photosynthetic capacity at 25°C ( V cmax,25 ) and the leaf respiration rate at 25°C ( R 25 )) following increasingly well‐understood principles. While some land surface models (LSMs) now account for thermal acclimation, they do so by assigning empirical parameterizations for individual plant functional types (PFTs). Here, we have implemented an Eco‐Evolutionary Optimality (EEO)‐based scheme to represent the universal acclimation of photosynthesis and leaf respiration to multiple environmental effects, and that therefore requires no PFT‐specific parameterizations, in a standard version of the widely used LSM, Noah MP. We evaluated model performance with plant trait data from a 5‐year experiment and extensive global field measurements, and carbon flux measurements from FLUXNET2015. We show that observed R 25 and V cmax,25 vary substantially both temporally and spatially within the same PFT ( C.V. >20%). Our EEO‐based scheme captures 62% of the temporal and 70% of the spatial variations in V cmax,25 (73% and 54% of the variations in R 25 ). The standard scheme underestimates gross primary production by 10% versus 2% for the EEO‐based scheme and generates a larger spread in r (correlation coefficient) across flux sites (0.79 ± 0.16 vs. 0.84 ± 0.1, mean ± S.D.). The standard scheme greatly overestimates canopy respiration (bias: ∼200% vs. 8% for the EEO scheme), resulting in less CO 2 uptake by terrestrial ecosystems. Our approach thus simulates climate‐carbon coupling more realistically, with fewer parameters.

The efflux of carbon dioxide (CO 2 ) from woody stems, a proxy for stem respiration, is a critical carbon flux from ecosystems to the atmosphere, which increases with temperature on short timescales. However, plants acclimate their respiratory response to temperature on longer timescales, potentially weakening the carbon-climate feedback. The magnitude of this acclimation is uncertain despite its importance for predicting future climate change. We develop an optimality-based theory dynamically linking stem respiration with leaf water supply to predict its thermal acclimation. We show that the theory accurately reproduces observations of spatial and seasonal change. We estimate the global value for current annual stem CO 2 efflux as 27.4 ± 5.9 PgC. By 2100, incorporating thermal acclimation reduces projected stem respiration without considering acclimation by 24 to 46%, thus reducing land ecosystem carbon emissions.

Land-atmosphere carbon exchanges and feedbacks constitute one of the largest uncertainties in future climate projections. Seasonal variations in atmospheric CO2 content depend on uptake by photosynthesis and release by autotrophic and heterotrophic respiration, providing an atmospheric signal of land ecosystem activity. Large increases in the seasonal cycle amplitude (SCA) of CO2 have occurred since the 1950s, especially in northern high latitudes. However, land surface and dynamic vegetation models have produced a wide range of magnitudes for the SCA, and have generally underestimated its increase. We explored the controls of the SCA by using a parameter-sparse eco-evolutionary optimality (EEO) model, the ‘P model’, combined with generic representations of plant and decomposer respiration, to simulate seasonal cycles and decadal trends of net ecosystem exchange (NEE). Simulated NEE fields were used to model near-surface CO2 concentrations during the satellite era, with the help of the atmospheric chemistry-transport model TM5. The P model has previously been shown to reproduce trends of gross primary production (GPP) at flux sites with long records. Our model set-up also generated a realistic simulation of global net terrestrial carbon uptake, comparable with results produced by more complex dynamic vegetation models; and allowed us to attribute causes to observed SCA increases at high-latitude CO2 monitoring stations.

Leaf photosynthetic and respiratory processes are important for the terrestrial carbon cycle. Leaf physiological traits, such as the maximum carboxylation rate and leaf respiration rate at 25˚C (Vcmax,25, R25), the key parameters affecting photosynthesis and respiration rate, acclimate to environmental changes. However, many land surface models (LSMs) assume a constant R25 and Vcmax,25 by plant functional types (PFTs) due to limited understanding of plant acclimation processes. Here, we incorporated the acclimation of photosynthesis and leaf respiration into a land surface model (Noah MP) using the Eco-Evolutionary Optimality principle (Noah MP-EEO), and evaluated the performances of the EEO and standard schemes to simulate photosynthesis and respiration using global plant trait measurements and data from FLUXNET. We demonstrate that R25 and Vcmax,25 varied temporally and spatially within the same PFT (C.V. >20%). This behaviour is captured by the EEO scheme (R2 =0.69 and 0.62 for temporal and spatial variations) but ignored by the standard scheme. At the FLUXNET sites, the standard scheme underestimates gross primary production (GPP) but this is reduced in the EEO scheme. The EEO scheme explains 66% of the variation of site-annual GPP compared to 55% in the standard scheme. The EEO scheme also simulates the variation of leaf respiration (Rleaves) better than the standard scheme (R2 increases from 0.45 to 0.77). The EEO scheme shows less temperature sensitivity than the standard scheme because of acclimation. This study indicates that adopting EEO approaches that do not require PFT-specific parameters improves carbon cycle predictions and could be used in Earth system models for better understanding the climate-carbon feedback.

Changes in climate have had major impacts on global vegetation during the Quaternary. However, variations in CO2 levels also play a role in shaping vegetation dynamics by influencing plant productivity and water-use efficiency and consequently the relative competitive success of the C3 and C4 photosynthetic pathways. We use an eco-evolutionary optimality (EEO)-based modelling approach to examine the impacts of climate fluctuations and CO2-induced alterations on gross primary production (GPP). We considered two contrasting periods, the Last Glacial Maximum (LGM; 21 000 years before present) and the mid-Holocene (MH; 6000 years before present) and compared both to pre-industrial (PI) conditions. The LGM, characterized by generally colder and drier climate, had a CO2 level close to the minimum for effective C3 plant operation. In contrast, the MH had warmer summers and increased monsoonal rainfall in the Northern Hemisphere, although with a CO2 level still below the PI. We simulated vegetation primary production at the LGM and the MH compared to the PI baseline using a light-use efficiency model that simulates GPP coupled to an EEO model that simulates leaf area index (LAI) and C3 / C4 competition. We found that low CO2 at the LGM was nearly as important as climate in reducing tree cover, increasing the abundance of C4 plants and lowering GPP. Global GPP in the MH was similar to the PI (although greater than the LGM), also reflecting CO2 constraints on plant growth despite the positive impacts of warmer and/or wetter climates experienced in the Northern Hemisphere and tropical regions. These results emphasize the importance of taking account of impacts of changing CO2 levels on plant growth to model ecosystem changes.

The coupling between carbon uptake and water loss through stomata implies that gross primary production (GPP) can be limited by soil water availability through reduced leaf area and/or stomatal conductance. Ecosystem and land-surface models commonly assume that GPP is highest under well-watered conditions and apply a stress function to reduce GPP as soil moisture declines. Optimality considerations, however, suggest that the stress function should depend on climatic aridity: ecosystems adapted to more arid climates should use water more conservatively when soil moisture is high, but maintain unchanged GPP down to a lower critical soil-moisture threshold. We use eddy-covariance flux data to test this hypothesis. We investigate how the light-use efficiency (LUE) of GPP depends on soil moisture across ecosystems representing a wide range of climatic aridity. 'Well-watered' GPP is estimated using the sub-daily P model, a first-principles LUE model driven by atmospheric data and remotely sensed vegetation cover. Breakpoint regression is used to relate daily β(θ) (the ratio of flux data-derived GPP to modelled well-watered GPP) to soil moisture estimated via a generic water balance model. The resulting piecewise function describing β(θ) varies with aridity, as hypothesised. Unstressed LUE, even when soil moisture is high, declines with increasing aridity index (AI). So does the critical soil-moisture threshold. Moreover, for any AI value, there exists a soil moisture level at which β(θ) is maximised. This level declines as AI increases. This behaviour is captured by universal non-linear functions relating both unstressed LUE and the critical soil-moisture threshold to AI. Applying these aridity-based functions to predict the site-level response of LUE to soil moisture substantially improves GPP simulation under both water-stressed and unstressed conditions, suggesting a route towards a robust, universal model representation of the effects of low soil moisture on leaf-level photosynthesis.