The widespread increase in vegetation productivity plays an important role in enhancing ecosystem carbon uptake. While it is well established that biodiversity increases ecosystem productivity, its influence on long-term changes in photosynthesis remains unclear. Here we integrate a high-resolution map of tree species richness with satellite-derived photosynthesis proxies during 2001–2020 to show that high richness not only enhances current levels of photosynthesis but also correlates with a greater increase in photosynthesis over time. This pattern is largely driven by an amplified CO2 fertilization effect (CFE) in species-rich forests. The ability of diverse forests to mitigate water and nutrient limitations probably contributes to the CFE enhancement and photosynthesis rise. Projections suggest that biodiversity losses by 2050 could reduce photosynthesis trends by 3–17%, representing a cumulative forest photosynthesis loss of 4.4–35.7 PgC. These findings underscore the critical need to integrate biodiversity conservation into climate mitigation strategies to safeguard the terrestrial carbon sink. The authors integrate tree species richness with satellite-derived photosynthesis proxies to show that richness correlates with greater current levels of photosynthesis and greater increases over time. Projected biodiversity loss by 2050 could lead to cumulative forest photosynthesis loss of 4.4–35.7 PgC.

Abstract Accurately representing photosynthetic optimum temperature (Topt) is essential for predicting terrestrial carbon uptake, yet conventional ecosystem-scale estimates based on peak gross primary productivity (GPP) are confounded by concurrent variations in radiation, moisture and phenology. Here, we develop a physiologically grounded approach that defines Topt as the air temperature where light use efficiency (LUE) reaches its maximum, thereby isolating the intrinsic thermal response of photosynthesis. We derived efficiency-based Topt (Topt-LUE) from 131 flux observation sites. By replacing biome-based Topt used in Vegetation Photosynthesis Model (VPM), we significantly improved GPP estimation (R2 = 0.71, RMSE = 1.93 gC m−2 d−1) compared to the biome-based approach (R2 = 0.62, RMSE = 2.84 gC m−2 d−1). We then used a Random Forest framework to generate global and time-varying Topt-LUE fields (2001–2020). The results revealed that biome-based Topt systematically overestimated Topt-LUE across ∼94% of global vegetated areas, with a mean bias of ∼10°C. The global Topt-LUE exhibit clear latitudinal gradients and biome-specific contrasts, providing evidence for widespread thermal acclimation of ecosystem photosynthesis. This acclimation is reflected in a mean increase in Topt-LUE of 0.021 ± 0.102 °C per year, underscoring a measurable response to long-term climate changes. When integrated into VPM, the dynamic Topt-LUE fields reshape the spatial pattern of simulated carbon uptake, mitigating overestimation in tropical areas (∼5 gC m−2 d−1) and enhancing underestimation in frigid areas and temperate regions including China, India and Europe. This study established a mechanistically grounded framework for quantifying ecosystem thermal acclimation, advancing the representation of temperature responses in terrestrial carbon cycle models.

2024 is the hottest year on record, accompanied by extreme precipitation, droughts and fires. The global atmospheric CO2 growth rate in 2024 reached a historic high of 3.73 ppm yr-1, significantly surpassing the previous record set during the 2015/16 El Niño event. Here, we investigate the causes and underlying mechanisms of this record-high growth rate by combining satellite-based atmospheric inversions and estimates of gross primary production and fire emissions. We find that the record-high CO2 growth rate is due to large reductions in the land CO2 sink. This is dominated by a dramatic increase in total ecosystem respiration, which occurred primarily in grass and shrub lands, owing to compound hot-wet climatic conditions in 2024. Given the projected increase in the frequency and intensity of compound pluvial-hot extremes under warming, changes in ecosystem respiration will become more drastic and cause positive feedback to climate warming.

Accurate accounting of regional methane (CH4) emissions and removals is essential for tracking national climate mitigation progress and assessing potential carbon-climate feedbacks. China is the world's largest CH4 emitter, yet comprehensive sub-national estimates remain limited, constraining the development of effective mitigation strategies. This study analyzes CH4 emissions across China from 2000 to 2019 using atmospheric inversion ensembles (top-down approach, TD) and process-based model estimates (bottom-up approach, BU). The datasets for both approaches were contributed by international research teams coordinated through the Global Carbon Project. The spatial distribution of various CH4 fluxes exhibits high spatial heterogeneity. Approximately 60% of national CH4 emissions come from three of the nine sub-national regions (North China, Southeast China, and Southwest China), which together account for <30% of China's land area. These emissions are dominated by the energy and agricultural sectors. Natural sources contribute 9%-16% of the total CH4 budget, but they have the largest relative uncertainties, reaching approximately 150%-170% of their estimated magnitudes. This high uncertainty partly reflects the limited number of studies on natural CH4 sources. The increase in anthropogenic emissions is the primary driver of the changes between the two decades, with increases of 10.4 [2.7-16.9] Tg CH4 a-1 (BU) and 6.1 [-2.6-10.7] Tg CH4 a-1 (TD). The largest increases occur in North China, Southwest China, Southeast China, and Northeast China. This study highlights the need for improved regional monitoring of CH4 emissions and sinks in China to support integrated and spatially targeted mitigation strategies.

Accurate measurement of canopy-scale solar-induced chlorophyll fluorescence (SIF) is essential for linking near-surface measurements with satellite observations and for reliably constraining terrestrial photosynthetic carbon uptake. However, the optimal ground observation scale (height and footprint) required to capture spatially representative SIF signals remains poorly defined. Here, we develop a physically based Optimal Observation Scale (OOS) model that integrates canopy height (Htoc) and fractional vegetation cover (FVC) to determine the optimal observation height (Hopt) and footprint size (Sopt) for maize. The model was parameterized using 3-D radiative transfer simulations (DART) and validated with a UAV-based hyperspectral SIF system across eight flight altitudes (5–50 m) over a full growing season. The results show that the spatial coefficient of variation (CV) of canopy SIF decreased significantly with increasing scale and stabilizes at Sopt of approximately 91.5 m2 (Hopt of approximately 26.6 m). Application of the OOS model reduced the CV from 17.3% to 7.9%, a 54.4% reduction in spatial uncertainty. The SIF–GPP (gross primary productivity) coupling strengthened with observation heights (Hobs) and approached saturation above the OOS-derived threshold (Hopt), highlighting the importance of observation scale for capturing photosynthetic dynamics. Validation against TROPOspheric Monitoring Instrument (TROPOMI) SIF further revealed that the strongest correlations (r ≈ 0.8) were achieved when UAV SIF observations were conducted at Hobs exceeding the Hopt calculated by the OOS model and spatially matched with satellite footprints. The OOS model provides a transferable, physically-based framework for optimizing ground SIF observations across scales. Its structure-based design also provides a pathway toward generalized SIF measurement strategies that can be extended to other ecosystems and future satellite validation efforts.

Large-scale photovoltaic (PV) deployment and afforestation represent vital climate change mitigation strategies, yet the climate mitigation potential of PV deployment in regions where afforestation is constrained remains understudied. Using the Community Earth System Model (CESM2.1.3), we quantitatively assess the carbon reduction and surface temperature effects of large-scale PV deployment and afforestation across global potential afforestation areas. Our results reveal regional limitations of afforestation and demonstrate strong climate benefits from PV deployment in these constrained regions. Our results demonstrate that PV deployment in potential afforestation regions could reduce carbon emissions by 3.47 Gt C yr-1 while decreasing global surface temperatures by 0.08 degrees C, whereas afforestation yields a smaller carbon sink (0.075 Gt C yr-1) and induces slight global warming (+0.02 degrees C). In arid zones, where solar radiation is high and bare land is prevalent, PV deployment mitigate 2.24 Gt C yr-1 (114 Mha) and cools the surface by -0.26 degrees C, whereas afforestation induces warming of +0.16 degrees C due to albedo reduction. In boreal regions, PV deployment cools surfaces by -0.26 degrees C, whereas afforestation leads to a +0.21 degrees C warming. These results demonstrate that in regions where afforestation is limited by water scarcity or induces biophysical warming, PV deployment can provide effective climate mitigation through stable carbon reduction and surface cooling.

Methane (CH4) is a potent greenhouse gas, with the oil and gas (O&G) sector being a major source of emissions, particularly in the Middle East and North Africa (MENA). However, a comprehensive assessment of point-source emissions in this region has remained limited. Here, we surveyed CH4 super-emitters across the MENA using hyperspectral remote sensing data from the GF-5, ZY-1, HJ-2, PRISMA, and EnMAP. We explored a detection framework that combines a Combo-matched filter algorithm with Chambolle total-variation denoising to enhance CH4 retrieval accuracy from satellite data. Between 2020 and 2023, we identified 136 CH4 plumes from 88 O&G facilities, with emission fluxes ranging from 221 +/- 112 to 27,600 +/- 7,100 kg hr(-1). The largest sources are concentrated in Iraq, Libya, and Turkmenistan, suggesting the spatial and statistical patterns consistent with previous findings that emissions are dominated by a small number of large emitters. Cross-platform comparisons, including publicly available EMIT detections, confirm major hotspot regions and extend spatial coverage beyond single-mission analyses. Based on the multi-satellite detections, the estimated contribution of super-emitters accounted for 4%-66% of the national O&G methane inventories. This suggests that intermittent or unreported releases are not fully reflected in the current inventory. We evaluated mitigation costs and societal benefits, showing that Turkmenistan offers the highest potential for cost-effective CH4 reduction, followed by Algeria and Iraq. These results demonstrate that multi-sensor satellite observations can bridge inventory gaps and support targeted methane mitigation with substantial societal and climate benefits.

Offshore oil and gas platforms constitute a significant but poorly constrained source of methane emission, largely due to their limited accessibility and episodic emission patterns. To address this, we present a multi-instrument remote sensing framework that integrates high-resolution airborne hyperspectral observations from AVIRIS-3 with spaceborne hyperspectral satellites (GF-5 and EnMAP) to detect and quantify methane emissions from offshore platforms in the Gulf of Mexico. Methane plumes are retrieved using an energy-partitioned matched-filter approach that accounts for sunglint-affected ocean reflectance, enabling robust detection over heterogeneous ocean surfaces. Airborne AVIRIS-3 observations are used to identify emitting platforms, while satellite observations acquired under favorable sunglint conditions extend spatial coverage and temporal sampling. Emission fluxes are quantified using the integrated mass enhancement method with instrument-specific wind parameterizations. Across seven days of AVIRIS-3 flights in June 2024, 72 methane plumes are detected and attributed to 21 offshore platforms, with flux rates ranging from 40.9 to 1481.3 kg/h. Furthermore, three satellite-detected plumes show good agreement with near-contemporaneous airborne estimates. Aggregating fluxes from all instruments and time periods, we find that platform-level emission rates range from 40.0 to 804.6 kg/h. Platform-level emissions exhibit strong temporal variability and show no clear relationship with reported production, while official inventories systematically underestimate observed emissions. These results demonstrate that combining airborne and satellite hyperspectral observations provides a robust and scalable approach for quantifying offshore methane emissions and capturing episodic release behavior.

Climatic and anthropogenic disturbances have led to intense small-scale tree cover loss in global forests. However, it remains unclear when forest attributes at a large scale (e.g., 0.05° resolution) will decline in response to such sub-grid (e.g., 30-m) tree cover losses within forest ecosystems. Utilizing global maps of forest attribute proxies, we discover that vegetation greenness, canopy structure, composition, and photosynthesis function can all increase under limited tree cover loss, indicating a widely existing safety margin in global forests that is primarily buffered by a positive edge effect of landscape fragmentation within forest ecosystems. The safety margin varies across biomes (tropical: 7.7%; temperate: 3.7%; boreal: 1.0%) and is often positively correlated with ecosystem resistance. In addition, about 35.7% of the remaining global forests have exceeded the safety margin. Our finding contrasts with the conventional perception that sub-grid tree cover losses are inevitably associated with declines in forest attributes and functions. It provides quantitative information for mitigating forest degradation and has strong implications for sustainable forest management practices.

Chlorophyll fluorescence (ChlF) is tightly linked to photosynthetic electron transport and informs gross primary productivity (GPP) across scales. Sun-induced ChlF (SIF) is typically retrieved at specific wavelengths (e.g., Fraunhofer lines, oxygen absorption bands) within a narrow field of view (solid angle), and given in power units. This spectral SIF radiance, denoted SIF lambda, is not directly photosynthetically relevant, rather, it is the integrated radiant exitance, SIFint, expressed in molar units and integrated over 660-800 nm and the full angular domain, that corresponds mechanistically to photosynthesis. It is generally assumed that the SIF lambda is proportional to SIFint and that this proportionality is spatially and temporally invariant. Here we tested this assumption with spectrally resolved SIF measured in three crop and six tree species at the leaf level. We found that while SIF lambda is significantly related to SIFint at individual Fraunhofer lines, this relationship varies with wavelength and leaf chlorophyll content (LCC). We therefore developed a model to predict SIFint from SIF lambda using wavelength and LCC as inputs. The model performed well across the ChlF emission band, particularly in the far-red region, enabling accurate conversion from observed SIF lambda to mechanistically relevant SIFint. As an exploratory extension, the model was applied at the canopy scale for C3 and C4 crops with the Mechanistic Light Response model, yielding encouraging agreement between modeled and observed GPP. The core contribution of this work is establishing the SIF lambda-SIFint relationship and a transfer model at the leaf scale, while the canopy application serves as an exploratory extension illustrating its scaling potential. Together, these findings provide a more mechanistically consistent basis for SIF-based GPP estimation and strengthen the application of fluorescence observations in carbon cycle research.

Accurate determination of the methane isotopic composition (delta 13CH4) is essential for attributing emission sources of methane (CH4). However, for measurements with optical instruments, spectral interference from water vapor and instrumental drift often introduce substantial biases in delta 13CH4 measurements, particularly for humid air measurements. Although multiple calibration strategies exist, a systematic evaluation of their performance under diverse field conditions remains lacking. Here, we evaluate two calibration strategies for a cavity ring-down spectrometer: a delta-based calibration for delta 13CH4 and an isotopologue-specific calibration for 12CH4 and 13CH4. We performed laboratory experiments over a water vapor range of 0.15 %-4.0 % to establish empirical correction functions, quadratic for 12CH4 and 13CH4, and linear for delta 13CH4, to remove humidity-induced biases. These correction functions were then applied to field measurements in both dried air at the SORPES stie and humid air at the Jurong site. At the SORPES site where air samples were dried using a Nafion (TM) dryer, the mean difference in delta 13CH4 between the two strategies was similar to 0.29 parts per thousand. In contrast, for humid air at the Jurong site, significant inter-method difference (Delta delta 13CH4) was observed, with which exhibiting a strong correlation with 1/CH4, indicating non-linear spectral effects are most pronounced at lower CH4 concentrations and compromise the performance of delta-based calibration. Notably, only the isotopologue-specific calibration, coupled with an explicit water vapor correction, delivered stable and accurate delta 13CH4 measurements across all conditions. This work underscores the need for robust calibration strategies to minimize bias in CH4 isotopic composition measurements.

Accurate estimation of gross primary productivity (GPP) is fundamental for understanding ecosystem carbon cycling. Solar-induced chlorophyll fluorescence (SIF) and carbonyl sulfide (COS) provide complementary proxies of photosynthesis, yet their relative performance across temporal scales and sky conditions remains uncertain. Using continuous eddy covariance COS fluxes and ground-based SIF observations in a rice paddy, we assess their relationships with GPP at half-hourly and daily scales under clear and cloudy conditions. SIF shows stronger correlations with GPP at the half-hourly scale, whereas COS–GPP associations strengthen after temporal aggregation, particularly under cloudy skies. Partial correlations indicate that SIF captures short-term variability, while COS becomes more informative at the daily scale. Integrating SIF and COS improves GPP estimation across conditions, especially at the daily scale and under cloudy skies. These results demonstrate scale-dependent complementarity and highlight the value of multi-proxy approaches for robust ecosystem GPP estimation.

Bottom-up coal mine methane (CMM) inventories rely on static or empirically derived emission factors (EFs), and therefore mine-level emissions are poorly constrained, limiting the use of these inventories for implementing detailed mitigation strategies. Here, we compiled 1418 satellite-detected methane plumes (2019 – 2025) and attributed them to 159 active underground coal mines in Shanxi province, China. We further derived observed mine-level EFs, calculated as mine-level emission rates divided by production data. We then compared these observed EFs with those from the State Administration of Coal Mine Safety (SACMS) and Global Coal Mine Tracker (GCMT), and developed a production-capacity-stratified bootstrap framework to upscale emissions from high-gas and outburst coal mines. Observed EFs were highly heterogeneous, right-skewed, temporally variable and negatively correlated with production capacity. Inventory comparisons revealed distinct biases: SACMS reproduced the overall EF magnitude but systematically underestimated EFs for small-capacity coal mines (production capacity <1.2 Mt yr ^−1 ), whereas GCMT overestimated EFs for medium- (1.2⩽ production capacity <3.0 Mt yr ^−1 ) and large-capacity coal mines (production capacity ⩾3.0 Mt yr ^−1 ). Using the production-capacity-stratified bootstrap upscaling framework, we estimated 2023 CMM emissions from high-gas and outburst coal mines in Shanxi to be 7.0 [5.6 – 8.9] Mt yr ^−1 . Total provincial CMM emissions were estimated at 11.2 [9.3 – 13.6] Mt yr ^−1 . These findings show that satellite-observed plumes can constrain mine-level EFs, reveal inventory biases, and support observation-based provincial CMM estimation.

Methane is a potent greenhouse gas with large emissions often arising from localized point sources in industrial facilities. MethaneSAT, launched in 2024, observes the 1,598-1,683 nm band with sub-nanometer resolution and 100 m & times; 400 m footprints, thereby bridging the gap between regional mapping and facility-scale monitoring and improving the detection of methane point sources from the energy and waste sectors. We introduce a Bias-Corrected Matched Filter (BCMF) with covariance exclusion and a statistical correction for linearization bias, mitigating the low bias observed in conventional matched-filter retrievals. Simulations and retrievals under MethaneSAT observing conditions show that BCMF mitigates the low bias observed in conventional matched-filter retrievals with algorithmic noise constrained to <= 40 ppb (1 sigma). Across multiple emission scenarios, 17 plumes were detected and cross-section emission flux estimates agree with MethaneSAT proxy products (r = 0.965, MAPE = 20.3%). These results demonstrate BCMF as a fast robust approach for quantifying methane emissions with next-generation high-resolution satellite sensors.

Abstract. Accurate estimation of coal mine methane (CMM) emissions in Shanxi Province, China's leading coal production hub, is essential for mitigating China's anthropogenic methane emissions. Hyperspectral remote sensing is an emerging method for real-time methane monitoring with significant potential for optimizing CMM emission factors. However, limited satellite revisit frequencies can introduce biases in CMM emission estimates. To address these issues, we developed a Hierarchical Bayesian Inversion Algorithm utilizing time-series observations from seven hyperspectral satellites in Shanxi (2019–2023), comprising 215 methane plumes from 26 coal mines, to estimate annual CMM emission rates with limited satellite revisit frequency. Subsequently, we integrated multi-source satellite observations with inventory data to estimate CMM emissions in Shanxi province. Our analysis yields a CMM emission factor of (7.9 ± 1.4)×10-3 Tg/Mt for Shanxi, with CMM emissions reaching 11 ± 2 Tg/yr in 2023. We demonstrate that CMM emissions follow a right-skewed distribution in Shanxi Province, where low-frequency extreme methane emission events (≥10000 kg/h) constitute approximately 25 % of all time-series observations. Additionally, our results reveal that capacity reduction policies initially decreased CMM emissions, but subsequent production recovery led to emission increases, with asymmetric responses to coal price fluctuations. Our findings establish a novel strategy for CMM accounting from hyperspectral satellite observations.

Climate-driven increases in wildfire frequency threaten forest resilience in fire-prone ecosystems. Although short-term studies indicate rapid recovery of eucalypt forests, such assessments may be overly optimistic, given the uncertain long-term alignment of structural and functional trajectories. This study integrated multi-source remote sensing data (2000-2020) with eddy covariance flux measurements to quantify post-fire recovery dynamics of eucalypt forests in southeastern Australia. We analyzed key structural metrics, including fraction of tree cover (FTC), the Normalized Difference Vegetation Index (NDVI), and leaf area index (LAI), as well as functional traits, including gross primary productivity (GPP), solar-induced fluorescence (SIF), evapotranspiration (ET), fuel moisture content (FMC), and vegetation optical depth (VOD). On average, all these traits recovered to pre-fire levels within 5.5 to 7.5 years, though some areas required over a decade or more. Recovery trajectories diverged notably both between the structural and functional categories and among individual traits within each category. These recovery rates varied across forest groups, modulated by disturbance severity, understory compositions, and fire frequency. Notably, functional traits such as GPP and SIF recovered rapidly, surpassing even the fastest structural metrics. At the Wallaby Creek flux site (burned in 2009 and rebuilt), GPP and SIF exhibited a distinct "peak-decline-stabilize" pattern, surpassing pre-fire levels within three years before declining as LAI recovered. In contrast, structural metrics recovered gradually, with LAI lagging behind FTC and NDVI. Our findings highlight the importance of integrating multi-dimensional data to comprehensively assess forest recovery, offering critical insights for ecosystem management and carbon cycle modeling in fire-prone regions.

Accurate representation of plant photosynthesis in terrestrial biosphere models (TBMs) is critical for reliable carbon-cycle simulations. Most TBMs employ the Farquhar-von Caemmerer-Berry (FvCB) model, which uses an empirical representation of electron transport that limits the simulation of gross primary productivity (GPP) and solar-induced chlorophyll fluorescence (SIF) under variable conditions. Here, we present BEPS-CB6F, an improved model that incorporates the mechanistic cytochrome b6f (Cyt b6f) scheme (CB6F) of Johnson and Berry into the Biosphere-atmosphere Exchange Process Simulator (BEPS). This implementation replaces empirical formulations with a process-based energy-allocation framework and links GPP and SIF through shared physiological parameters, including the maximum Cyt b6f activity (Vqmax) and the fraction of total leaf absorbance allocated to photosystem II (PSII) (β₂). BEPS-CB6F also integrates key photoprotective processes, including cyclic electron flow around photosystem I (CEF), non-photochemical quenching of photosystem II (NPQ), and photosynthetic control of Cyt b6f, within a two-leaf canopy scheme that differentiates sunlit and shaded responses. The results show that across flux-tower sites, BEPS-CB6F substantially improves SIF simulations, with RMSE and rRMSE reductions at more than 90% of sites, and yields moderate but consistent improvements in GPP, including higher R2 and reduced RMSE at over 80% of sites. The model alleviates GPP overestimation under low light, particularly in shaded leaves, and markedly reduces SIF overestimation under high irradiance in sunlit leaves. BEPS-CB6F further enhances performance during heat and high-VPD conditions. By explicitly representing temperature-responsive CEF and NPQ, it captures the strong midday suppression of GPP and SIF, including reductions in GPP and SIF during heatwaves. Sensitivity analyses indicate that Vqmax and β2 strongly influence GPP simulations, while β2 is the primary driver of SIF simulations. These results highlight the importance of mechanistic electron-transport representation and demonstrate the potential of CB6F to improve terrestrial biosphere model predictions of carbon uptake and fluorescence.