Explore the vast range of titles available.

A dynamic global vegetation model (DGVM) is applied in a probabilistic framework and benchmarking system to constrain uncertain model parameters by observations and to quantify carbon emissions from land-use and land-cover change (LULCC). Processes featured in DGVMs include parameters which are prone to substantial uncertainty. To cope with these uncertainties Latin hypercube sampling (LHS) is used to create a 1000-member perturbed parameter ensemble, which is then evaluated with a diverse set of global and spatiotemporally resolved observational constraints. We discuss the performance of the constrained ensemble and use it to formulate a new best-guess version of the model (LPX-Bern v1.4). The observationally constrained ensemble is used to investigate historical emissions due to LULCC (ELUC) and their sensitivity to model parametrization. We find a global ELUC estimate of 158 (108, 211) PgC (median and 90 % confidence interval) between 1800 and 2016. We compare ELUC to other estimates both globally and regionally. Spatial patterns are investigated and estimates of ELUC of the 10 countries with the largest contribution to the flux over the historical period are reported. We consider model versions with and without additional land-use processes (shifting cultivation and wood harvest) and find that the difference in global ELUC is on the same order of magnitude as parameter-induced uncertainty and in some cases could potentially even be offset with appropriate parameter choice.

Nitrous oxide (N₂O) is an important greenhouse gas and also an ozone-depleting substance that has both natural and anthropogenic sources. Large estimation uncertainty remains on the magnitude and spatiotemporal patterns of N₂O fluxes and the key drivers of N₂O production in the terrestrial biosphere. Some terrestrial biosphere models have been evolved to account for nitrogen processes and to show the capability to simulate N₂O emissions from land ecosystems at the global scale, but large discrepancies exist among their estimates primarily because of inconsistent input datasets, simulation protocol, and model structure and parameterization schemes. Based on the consistent model input data and simulation protocol, the global N₂O Model Intercomparison Project (NMIP) was initialized with 10 state-of-the-art terrestrial biosphere models that include nitrogen (N) cycling. Specific objectives of NMIP are to 1) unravel the major N cycling processes controlling N₂O fluxes in each model and identify the uncertainty sources from model structure, input data, and parameters; 2) quantify the magnitude and spatial and temporal patterns of global and regional N₂O fluxes from the preindustrial period (1860) to present and attribute the relative contributions of multiple environmental factors to N₂O dynamics; and 3) provide a benchmarking estimate of N₂O fluxes through synthesizing the multimodel simulation results and existing estimates from ground-based observations, inventories, and statistical and empirical extrapolations. This study provides detailed descriptions for the NMIP protocol, input data, model structure, and key parameters, along with preliminary simulation results. The global and regional N₂O estimation derived from the NMIP is a key component of the global N₂O budget synthesis activity jointly led by the Global Carbon Project and the International Nitrogen Initiative.

The observed global net land carbon sink is captured by current land models. All models agree that atmospheric CO2 and nitrogen deposition driven gains in carbon stocks are partially offset by climate and land-use and land-cover change (LULCC) losses. However, there is a lack of consensus in the partitioning of the sink between vegetation and soil, where models do not even agree on the direction of change in carbon stocks over the past 60 years. This uncertainty is driven by plant productivity, allocation, and turnover response to atmospheric CO2 (and to a smaller extent to LULCC), and the response of soil to LULCC (and to a lesser extent climate). Overall, differences in turnover explain ~70% of model spread in both vegetation and soil carbon changes. Further analysis of internal plant and soil (individual pools) cycling is needed to reduce uncertainty in the controlling processes behind the global land carbon sink.

Carbon dioxide removal (CDR) is discussed for offsetting residual greenhouse gas emissions or even reversing climate change. All emissions scenarios of the Intergovernmental Panel on Climate Change that meet the ‘well below 2 °C’ warming target of the Paris Agreement include CDR. Ocean alkalinity enhancement (OAE) may be one possible CDR where the carbon uptake of the ocean is increased by artificial alkalinity addition. Here, we investigate the effect of OAE on modelled carbon reservoirs and fluxes in two observationally-constrained large perturbed parameter ensembles. OAE is assumed to be technically successful and deployed as an additional CDR in the SSP5-3.4 temperature overshoot scenario. Tradeoffs involving feedbacks with atmospheric CO 2 result in a low efficiency of an alkalinity-driven atmospheric CO 2 reduction of −0.35 [−0.37 to −0.33] mol C per mol alkalinity addition (skill-weighted mean and 68% c.i.). The realized atmospheric CO 2 reduction, and correspondingly the efficiency, is more than two times smaller than the direct alkalinity-driven enhancement of ocean uptake. The alkalinity-driven ocean carbon uptake is partly offset by the release of carbon from the land biosphere and a reduced ocean carbon sink in response to lowered atmospheric CO 2 under OAE. In a second step we use the Bern3D-LPX model in CO 2 peak-decline simulations to address hysteresis and temporal lags of surface air temperature change (ΔSAT) in an idealized scenario where ΔSAT increases to ~2 °C and then declines to ~1.5 °C as result of CDR. ΔSAT lags the decline in CO 2 -forcing by 18 [14–22] years, depending close to linearly on the equilibrium climate sensitivity of the respective ensemble member. These tradeoffs and lags are an inherent feature of the Earth system response to changes in atmospheric CO 2 and will therefore be equally important for other CDR methods.

The Global Carbon Project estimates that the terrestrial biosphere has absorbed about one‐third of anthropogenic CO2 emissions during the 1959–2019 period. This sink‐estimate is produced by an ensemble of terrestrial biosphere models and is consistent with the land uptake inferred from the residual of emissions and ocean uptake. The purpose of our study is to understand how well terrestrial biosphere models reproduce the processes that drive the terrestrial carbon sink. One challenge is to decide what level of agreement between model output and observation‐based reference data is adequate considering that reference data are prone to uncertainties. To define such a level of agreement, we compute benchmark scores that quantify the similarity between independently derived reference data sets using multiple statistical metrics. Models are considered to perform well if their model scores reach benchmark scores. Our results show that reference data can differ considerably, causing benchmark scores to be low. Model scores are often of similar magnitude as benchmark scores, implying that model performance is reasonable given how different reference data are. While model performance is encouraging, ample potential for improvements remains, including a reduction in a positive leaf area index bias, improved representations of processes that govern soil organic carbon in high latitudes, and an assessment of causes that drive the inter‐model spread of gross primary productivity in boreal regions and humid tropics. The success of future model development will increasingly depend on our capacity to reduce and account for observational uncertainties. Plain Language Summary Earth's natural vegetation absorbs about one‐third of CO2 emissions caused by human activities. This value is produced by a group of models rather than through direct observations. Our study assesses how well models reproduce the processes that drive the CO2 exchange between land and atmosphere using a wide range of data sets that are mainly derived from field measurements and satellite images. These reference data sets are prone to errors that are not quantified in a consistent manner. To account for such errors, we first compare different reference data sets against each other. We then compare model output against reference data and assess whether the differences are comparable to the differences among the reference data sets. We conclude that the performance of models is encouraging given how uncertain reference data are, but that ample potential for improvements remains. Key Points Differences between model and observations are often similar compared to differences between independently derived observation‐based data We quantify differences between independently derived observations to disentangle model deficiencies from observational uncertainties Future work should address biases in soil organic carbon, leaf area index, and the large spread of gross primary productivity among models

The CO 2 efflux from soil (soil respiration (SR)) is one of the largest fluxes in the global carbon (C) cycle and its response to climate change could strongly influence future atmospheric CO 2 concentrations. Still, a large divergence of global SR estimates and its autotrophic (AR) and heterotrophic (HR) components exists among process based terrestrial ecosystem models. Therefore, alternatively derived global benchmark values are warranted for constraining the various ecosystem model output. In this study, we developed models based on the global soil respiration database (version 5.0), using the random forest (RF) method to generate the global benchmark distribution of total SR and its components. Benchmark values were then compared with the output of ten different global terrestrial ecosystem models. Our observationally derived global mean annual benchmark rates were 85.5 ± 40.4 (SD) Pg C yr −1 for SR, 50.3 ± 25.0 (SD) Pg C yr −1 for HR and 35.2 Pg C yr −1 for AR during 1982–2012, respectively. Evaluating against the observations, the RF models showed better performance in both of SR and HR simulations than all investigated terrestrial ecosystem models. Large divergences in simulating SR and its components were observed among the terrestrial ecosystem models. The estimated global SR and HR by the ecosystem models ranged from 61.4 to 91.7 Pg C yr −1 and 39.8 to 61.7 Pg C yr −1 , respectively. The most discrepancy lays in the estimation of AR, the difference (12.0–42.3 Pg C yr −1 ) of estimates among the ecosystem models was up to 3.5 times. The contribution of AR to SR highly varied among the ecosystem models ranging from 18% to 48%, which differed with the estimate by RF (41%). This study generated global SR and its components (HR and AR) fluxes, which are useful benchmarks to constrain the performance of terrestrial ecosystem models.

Natural wetlands constitute the largest and most uncertain source of methane (CH4) to the atmosphere and a large fraction of them are found in the northern latitudes. These emissions are typically estimated using process (“bottom-up”) or inversion (“top-down”) models. However, estimates from these two types of models are not independent of each other since the top-down estimates usually rely on the a priori estimation of these emissions obtained with process models. Hence, independent spatially explicit validation data are needed. Here we utilize a random forest (RF) machine-learning technique to upscale CH4 eddy covariance flux measurements from 25 sites to estimate CH4 wetland emissions from the northern latitudes (north of 45∘ N). Eddy covariance data from 2005 to 2016 are used for model development. The model is then used to predict emissions during 2013 and 2014. The predictive performance of the RF model is evaluated using a leave-one-site-out cross-validation scheme. The performance (Nash–Sutcliffe model efficiency =0.47) is comparable to previous studies upscaling net ecosystem exchange of carbon dioxide and studies comparing process model output against site-level CH4 emission data. The global distribution of wetlands is one major source of uncertainty for upscaling CH4. Thus, three wetland distribution maps are utilized in the upscaling. Depending on the wetland distribution map, the annual emissions for the northern wetlands yield 32 (22.3–41.2, 95 % confidence interval calculated from a RF model ensemble), 31 (21.4–39.9) or 38 (25.9–49.5) Tg(CH4) yr−1. To further evaluate the uncertainties of the upscaled CH4 flux data products we also compared them against output from two process models (LPX-Bern and WetCHARTs), and methodological issues related to CH4 flux upscaling are discussed. The monthly upscaled CH4 flux data products are available at https://doi.org/10.5281/zenodo.2560163 (Peltola et al., 2019).

Measurements of the stable carbon isotope ratio (δ13C) on annual tree rings offer new opportunities to evaluate mechanisms of variations in photosynthesis and stomatal conductance under changing CO2 and climate conditions, especially in conjunction with process-based biogeochemical model simulations. The isotopic discrimination is indicative of the ratio between the CO2 partial pressure in the intercellular cavities and the atmosphere (ci∕ca) and of the ratio of assimilation to stomatal conductance, termed intrinsic water-use efficiency (iWUE). We performed isotope-enabled simulations over the industrial period with the land biosphere module (CLM4.5) of the Community Earth System Model and the Land Surface Processes and Exchanges (LPX-Bern) dynamic global vegetation model. Results for C3 tree species show good agreement with a global compilation of δ13C measurements on leaves, though modeled 13C discrimination by C3 trees is smaller in arid regions than measured. A compilation of 76 tree-ring records, mainly from Europe, boreal Asia, and western North America, suggests on average small 20th century changes in isotopic discrimination and in ci∕ca and an increase in iWUE of about 27 % since 1900. LPX-Bern results match these century-scale reconstructions, supporting the idea that the physiology of stomata has evolved to optimize trade-offs between carbon gain by assimilation and water loss by transpiration. In contrast, CLM4.5 simulates an increase in discrimination and in turn a change in iWUE that is almost twice as large as that revealed by the tree-ring data. Factorial simulations show that these changes are mainly in response to rising atmospheric CO2. The results suggest that the downregulation of ci∕ca and of photosynthesis by nitrogen limitation is possibly too strong in the standard setup of CLM4.5 or that there may be problems associated with the implementation of conductance, assimilation, and related adjustment processes on long-term environmental changes.

Using high-precision and centennial-resolution ice core information on atmospheric nitrous oxide concentrations and its stable nitrogen and oxygen isotopic composition, we quantitatively reconstruct changes in the terrestrial and marine N2O emissions over the last 21 000 years. Our reconstruction indicates that N2O emissions from land and ocean increased over the deglaciation largely in parallel by 1.7±0.3 and 0.7±0.3 TgN yr−1, respectively, relative to the Last Glacial Maximum level. However, during the abrupt Northern Hemisphere warmings at the onset of the Bølling–Allerød warming and the end of the Younger Dryas, terrestrial emissions respond more rapidly to the northward shift in the Intertropical Convergence Zone connected to the resumption of the Atlantic Meridional Overturning Circulation. About 90 % of these large step increases were realized within 2 centuries at maximum. In contrast, marine emissions start to slowly increase already many centuries before the rapid warmings, possibly connected to a re-equilibration of subsurface oxygen in response to previous changes. Marine emissions decreased, concomitantly with changes in atmospheric CO2 and δ13C(CO2), at the onset of the termination and remained minimal during the early phase of Heinrich Stadial 1. During the early Holocene a slow decline in marine N2O emission of 0.4 TgN yr−1 is reconstructed, which suggests an improvement of subsurface water ventilation in line with slowly increasing Atlantic overturning circulation. In the second half of the Holocene total emissions remain on a relatively constant level, but with significant millennial variability. The latter is still difficult to attribute to marine or terrestrial sources. Our N2O emission records provide important quantitative benchmarks for ocean and terrestrial nitrogen cycle models to study the influence of climate on nitrogen turnover on timescales from several decades to glacial–interglacial changes.

Extreme climate events, such as droughts and heatwaves, can have large impacts on the environment. Disentangling their individual and combined effects is a difficult task, due to the challenges associated with generating controlled environments to study differences in their impacts. One approach to this problem is creating artificial climate forcing with varying magnitude of univariate and compound extremes, which can be applied to process‐based impact models. Here, we propose and describe a set of six 100‐year long climate scenarios with varying drought‐heat signatures that are derived from climate model simulations whose mean climate is comparable to present‐day climate conditions. The changes in extremes are most notable in the 3 months in which vegetation activity is highest and where arguably hot and dry extremes may have the largest impacts. Besides a control scenario representing natural variability (Control), one scenario has neither heat nor drought extremes (Noextremes), one has univariate extremes but no compound extremes (Nocompound), one has only heat extremes but few droughts (Hot), one has only droughts but few heatwaves (Dry), and one has many compound heat and drought extremes (Hotdry). These scenarios differ only moderately in their global mean climate (about 0.3°C in temperature and 6% in precipitation) and do not contain any long‐term trends. The data are provided on a daily timescale over land (except Antarctica and parts of Greenland) on a regular 1° × 1° grid. These scenarios were constructed primarily to investigate the impact of varying drought‐heat signatures on vegetation and the terrestrial carbon cycle. However, we believe that they may also prove useful to study the differential impacts of droughts and heatwaves in other areas, such as the occurrence of wildfires or crop failure. The data described here can be found on zenodo (https://doi.org/10.5281/zenodo.4385445, Tschumi et al., 2020). We created six climate scenarios with varying drought‐heat signatures for impact modelling. Besides a control scenario representing natural variability (Control), one scenario has neither heat nor drought extremes (Noextremes), one has univariate extremes but no compound extremes (Nocompound), one has only heat extremes but few droughts (Hot), one has only droughts but few heatwaves (Dry), and one has many compound heat and drought extremes (Hotdry). These datasets may proof useful to study the differential impacts of droughts and heatwaves in areas such as vegetation dynamics, occurrence of wildfires or crop failures.