Litter decomposition is an important ecosystem process and global carbon flux that has been shown to be controlled by climate, litter quality, and microbial communities. Process-based ecosystem models are used to predict responses of litter decomposition to climate change. While these models represent climate and litter quality effects on litter decomposition, they have yet to integrate empirical microbial community data into their parameterizations for predicting litter decomposition. To fill this gap, our research used a comprehensive leaf litterbag decomposition experiment at 10 temperate forest U.S. National Ecological Observatory Network (NEON) sites to calibrate (7 sites) and validate (3 sites) the MIcrobial-MIneral Carbon Stabilization (MIMICS) model. MIMICS was calibrated to empirical decomposition rates and to their empirical drivers, including the microbial community (represented as the copiotroph-to-oligotroph ratio). We calibrate to empirical drivers, rather than solely rates or pool sizes, to improve the underlying drivers of modeled leaf litter decomposition. We then validated the calibrated model and evaluated the effects of calibration under climate change using the SSP 3-7.0 climate change scenario. We find that incorporating empirical drivers of litter decomposition provides similar, and sometimes better (in terms of goodness-of-fit metrics), predictions of leaf litter decomposition but with different underlying ecological dynamics. For some sites, calibration also increased climate change-induced leaf litter mass loss by up to 5%, with implications for carbon cycle-climate feedbacks. Our work also provides an example for integrating data on the relative abundance of bacterial functional groups into an ecosystem model using a novel calibration method to bridge empiricism and process-based modeling, answering a call for the use of empirical microbial community data in process-based ecosystem models. We highlight that incorporating mechanistic information into models, as done in this study, is important for improving confidence in model projections of ecological processes like litter decomposition under climate change.

Increasing fine root carbon (FRC) inputs into soils has been proposed as a solution to increasing soil organic carbon (SOC). However, FRC inputs can also enhance SOC loss through priming. Here, we tested the broad-scale relationships between SOC and FRC at 43 sites across the US National Ecological Observatory Network. We found that SOC and FRC stocks were positively related with an across-ecosystem slope of 7 ± 3 kg SOC m-2 per kg FRC m-2, but this relationship was driven by grasslands. Grasslands had double the across-ecosystem slope while forest FRC and SOC were unrelated. Furthermore, deep grassland soils primarily showed net SOC accrual relative to FRC input. Conversely, forests had high variability in whether FRC inputs were related to net SOC priming or accrual. We conclude that while FRC increases could lead to increased SOC in grasslands, especially at depth, the FRC-SOC relationship remains difficult to characterize in forests.

ABSTRACT Permafrost is undergoing rapid changes due to climate warming, potentially exposing a vast reservoir of carbon to be released to the atmosphere, causing a positive feedback cycle. Despite the importance of this feedback, its specifics remain poorly constrained, because representing permafrost dynamics still poses a significant challenge for Earth System Models (ESMs). This review assesses the current state of permafrost representation in land surface models (LSMs) used in ESMs and offline permafrost models, highlighting both the progress made and the remaining gaps. We identify several key physical processes crucial for permafrost dynamics, including soil thermal regimes, freeze–thaw cycles, and soil hydrology, which are underrepresented in many models. While some LSMs have advanced significantly in incorporating these processes, others lack fundamental elements such as latent heat of freeze–thaw, deep soil columns, and Arctic vegetation dynamics. Offline permafrost models provide valuable insights, offering detailed process testing and aiding the prioritization of improvements in coupled LSMs. Our analysis reveals that while significant progress has been made in incorporating permafrost‐related processes into coupled LSMs, many small‐scale processes crucial for permafrost dynamics remain underrepresented. This is particularly important for capturing the complex interactions between physical and biogeochemical processes required to model permafrost carbon dynamics. We recommend leveraging advancements from offline permafrost models and progressively integrating them into LSMs, while recognizing the computational and technical challenges that may arise in coupled simulations. We highlight the importance of enhancing the representation of physical processes, including through improvements in model resolution and complexity, as this is a fundamental precursor to accurately incorporate biogeochemical processes and capture the permafrost carbon feedback.

Increasing root carbon inputs into soils has been proposed as a solution to increasing soil organic carbon (SOC). However, while fine root carbon (FRC) inputs can increase SOC accrual in soils, FRC can also enhance SOC loss by stimulating microbial respiration and cause a net loss of SOC through priming. It remains unclear how SOC varies as a function of FRC at broad spatial scales and across ecosystems and depths. Here, we tested the relationship of SOC and FRC using data from 43 sites across the US National Ecological Observatory Network (NEON). We found that total stocks of SOC and FRC in the top 2 meters of soil were positively related with an across-ecosystem slope of 7 ± 3 kg SOC m-2 per kg FRC m-2. However, grassland sites primarily drove this relationship. Grasslands had 15 ± 2 kg SOC m-2 per kg FRC m-2, which is double the across-ecosystem slope. We used deviations from the standardized 1:1 relationship between FRC and SOC to infer whether ecosystems were net priming (indicated by observed SOC being lower than the 1:1 line) or SOC accruing (higher SOC than the 1:1 line). Grassland soils and especially their deep soil layers (>30 cm) showed primarily SOC accrual with increasing fine root abundance. Meanwhile, forest soils had high variability in whether increasing fine roots were associated with net SOC priming or accrual across both shallow and deeper soil layers. We found that in grasslands, FRC inputs are strongly related to SOC accrual, especially at depth and at sites with high moisture and clay content. In contrast, SOC-FRC relationships in forests remain difficult to characterize. Nevertheless, deep grassland soils may serve as optimal environments in which increasing FRC could lead to meaningful increases in SOC stocks. ### Competing Interest Statement The authors have declared no competing interest.

Irrigation rapidly expanded during the 20th century, affecting climate via water, energy, and biogeochemical changes. Previous assessments of these effects predominantly relied on a single Earth System Model, and therefore suffered from structural model uncertainties. Here we quantify the impacts of historical irrigation expansion on climate by analysing simulation results from six Earth system models participating in the Irrigation Model Intercomparison Project (IRRMIP). Results show that irrigation expansion causes a rapid increase in irrigation water withdrawal, which leads to less frequent 2-meter air temperature heat extremes across heavily irrigated areas (≥4 times less likely). However, due to the irrigation-induced increase in air humidity, the cooling effect of irrigation expansion on moist-heat stress is less pronounced or even reversed, depending on the heat stress metric. In summary, this study indicates that irrigation deployment is not an efficient adaptation measure to escalating human heat stress under climate change, calling for carefully dealing with the increased exposure of local people to moist-heat stress. Although irrigation expansion during the 20th century masked or even reversed local warming trends over some intensely irrigated regions, the exposure to moist-heat extremes of local population has increased due to higher air humidity.

Comprehensive land models are subject to significant parametric uncertainty, which can be hard to quantify due to the large number of parameters and high model computational costs. We constructed a large parameter perturbation ensemble (PPE) for the Community Land Model version 5.1 with biogeochemistry configuration (CLM5.1-BGC). We performed more than 2,000 simulations perturbing 211 parameters across six forcing scenarios. This provides an expansive data set, which can be used to identify the most influential parameters on a wide range of output variables globally, by biome, or by plant functional type. We found that parameter effects can exceed scenario effects and that a small number of parameters explains a large fraction of variance across our ensemble. The most important parameters can differ regionally and also based on the forcing scenario. The software infrastructure developed for this experiment has greatly reduced the human and computer time needed for CLM PPEs, which can facilitate routine investigation of parameter sensitivity and uncertainty, as well as automated calibration.

Observations show an increase in the seasonal cycle amplitude of CO2 in northern latitudes over the past half century. Although multiple drivers contribute, observations and inversion models cannot quantitatively account for the factors contributing to the increased CO2 amplitude and older versions of Earth System Models (ESMs) do not simulate it. Here we show that several current generation ESMs are closer to the observed CO2 amplitude and highlight that in the Community Earth System Model (CESM) agricultural nitrogen (N) fertilization increases CO2 amplitude by 1-3 ppm throughout the Northern Hemisphere and up to 9 ppm in agricultural hotspots. While agricultural N fertilization is the largest contributor to the enhanced amplitude (45%) in Northern Hemisphere land-atmosphere carbon fluxes in CESM, higher CO2 concentrations and warmer temperatures also contribute, though to a lesser extent (40% and 18% respectively). Our results emphasize the fundamental role of agricultural management in Northern Hemisphere carbon cycle feedbacks and illustrate that agricultural N fertilization should be considered in future carbon cycle simulations.

Abstract Terrestrial, aquatic, and marine ecosystems regulate climate at local to global scales through exchanges of energy and matter with the atmosphere and assist with climate change mitigation through nature‐based climate solutions. Climate science is no longer a study of the physics of the atmosphere and oceans, but also the ecology of the biosphere. This is the promise of Earth system science: to transcend academic disciplines to enable study of the interacting physics, chemistry, and biology of the planet. However, long‐standing tension in protecting, restoring, and managing forest ecosystems to purposely improve climate evidences the difficulties of interdisciplinary science. For four centuries, forest management for climate betterment was argued, legislated, and ultimately dismissed, when nineteenth century atmospheric scientists narrowly defined climate science to the exclusion of ecology. Today's Earth system science, with its roots in global models of climate, unfolds in similar ways to the past. With Earth system models, geoscientists are again defining the ecology of the Earth system. Here we reframe Earth system science so that the biosphere and its ecology are equally integrated with the fluid Earth to enable Earth system prediction for planetary stewardship. Central to this is the need to overcome an intellectual heritage to the models that elevates geoscience and marginalizes ecology and local land knowledge. The call for kilometer‐scale atmospheric and ocean models, without concomitant scientific and computational investment in the land and biosphere, perpetuates the geophysical view of Earth and will not fully provide the comprehensive actionable information needed for a changing climate.

In the past few decades, there has been an evolution in our understanding of soil organic matter (SOM) dynamics from one of inherent biochemical recalcitrance to one deriving from plant-microbe-mineral interactions. This shift in understanding has been driven, in part, by influential conceptual frameworks which put forth hypotheses about SOM dynamics. Here, we summarize several focal conceptual frameworks and derive from them six controls related to SOM formation, (de)stabilization, and loss. These include: (1) physical inaccessibility; (2) mineral stabilization; (3) abiotic environmental limitation; (4) biochemical reactivity and diversity; (5) biodegradability of plant inputs; and (6) microbial properties. We then review the empirical evidence for these controls, their model representation, and outstanding knowledge gaps. We find relatively strong empirical support and model representation of abiotic environmental limitation but disparities between data and models for biochemical reactivity and diversity, mineral stabilization, and biodegradability of plant inputs, particularly with respect to SOM destabilization for the latter two controls. More empirical research on physical inaccessibility and microbial properties is needed to deepen our understanding of these critical SOM controls and improve their model representation. The SOM controls are highly interactive and also present some inconsistencies which may be reconciled by considering methodological limitations or temporal and spatial variation. Future conceptual frameworks must simultaneously refine our understanding of these six SOM controls at various spatial and temporal scales and within a hierarchical structure, while incorporating emerging insights. This will advance our ability to accurately predict SOM dynamics.

Abstract Nutrient limitation is widespread in terrestrial ecosystems. Accordingly, representations of nitrogen (N) limitation in land models typically dampen rates of terrestrial carbon (C) accrual, compared with C‐only simulations. These previous findings, however, rely on soil biogeochemical models that implicitly represent microbial activity and physiology. Here we present results from a biogeochemical model testbed that allows us to investigate how an explicit versus implicit representation of soil microbial activity, as represented in the MIcrobial‐MIneral Carbon Stabilization (MIMICS) and Carnegie‐Ames‐Stanford Approach (CASA) soil biogeochemical models, respectively, influence plant productivity, and terrestrial C and N fluxes at initialization and over the historical period. When forced with common boundary conditions, larger soil C pools simulated by the MIMICS model reflect longer inferred soil organic matter (SOM) turnover times than those simulated by CASA. At steady state, terrestrial ecosystems experience greater N limitation when using the MIMICS‐CN model, which also increases the inferred SOM turnover time. Over the historical period, however, warming‐induced acceleration of SOM decomposition over high latitude ecosystems increases rates of N mineralization in MIMICS‐CN. This reduces N limitation and results in faster rates of vegetation C accrual. Moreover, as SOM stoichiometry is an emergent property of MIMICS‐CN, we highlight opportunities to deepen understanding of sources of persistent SOM and explore its potential sensitivity to environmental change. Our findings underscore the need to improve understanding and representation of plant and microbial resource allocation and competition in land models that represent coupled biogeochemical cycles under global change scenarios.

Soil carbon-nitrogen (C:N) stoichiometry acts as a control over decomposition and soil organic matter formation and loss, making it a key soil property for understanding ecosystem dynamics and projected ecosystems responses to global environmental change. However, the controls of soil C:N and how they respond to increasing pressures from global change agents are not fully understood. The “foundational” controls on soil C:N, namely plant and microbial C:N, have been used to predict soil C:N, but fail to accurately simulate all ecosystems and may be insufficient for predictions under global environmental change. We present an “emerging” representation of controls of soil C:N that includes plant-microbe-mineral feedbacks that have been shown to regulate soil C:N. We argue that including representation of these emerging drivers in process-based terrestrial biogeochemistry models, which include biological N fixation, mycorrhizae, priming, root exudation of organic acids, and mineralogy (including soil texture, mineral composition, and aggregation), will improve mechanistic representation of soil C:N and associated processes. Such improvements will produce models that will better simulate a variety of ecological states and predict soil C:N when global changes modify plant-microbe-mineral interactions. Here, we align our empirical understanding of controls of soil C:N with those controls represented in models, identifying contexts where emerging drivers might be particularly important to represent (e.g., priming and root exudation in nutrient-limited conditions) and areas of future work. Additionally, we show that implementing emerging drivers of soil C:N results in different simulated outcomes at steady state and in response to elevated atmospheric CO2. Our review and preliminary simulations support the need to incorporate emerging drivers of soil C:N into process-based terrestrial biogeochemistry models, allowing for both theoretical exploration of mechanisms and potentially more accurate predictions of land biogeochemical responses to global change.

Soil organic matter decomposition and its interactions with climate depend on whether the organic matter is associated with soil minerals. However, data limitations have hindered global-scale analyses of mineral-associated and particulate soil organic carbon pools and their benchmarking in Earth system models used to estimate carbon cycle–climate feedbacks. Here we analyse observationally derived global estimates of soil carbon pools to quantify their relative proportions and compute their climatological temperature sensitivities as the decline in carbon with increasing temperature. We find that the climatological temperature sensitivity of particulate carbon is on average 28% higher than that of mineral-associated carbon, and up to 53% higher in cool climates. Moreover, the distribution of carbon between these underlying soil carbon pools drives the emergent climatological temperature sensitivity of bulk soil carbon stocks. However, global models vary widely in their predictions of soil carbon pool distributions. We show that the global proportion of model pools that are conceptually similar to mineral-protected carbon ranges from 16 to 85% across Earth system models from the Coupled Model Intercomparison Project Phase 6 and offline land models, with implications for bulk soil carbon ages and ecosystem responsiveness. To improve projections of carbon cycle–climate feedbacks, it is imperative to assess underlying soil carbon pools to accurately predict the distribution and vulnerability of soil carbon.

Climate simulation uncertainties arise from internal variability, model structure, and external forcings. Model intercomparisons (such as the Coupled Model Intercomparison Project; CMIP) and single-model large ensembles have provided insight into uncertainty sources. Under the Community Earth System Model (CESM) project, large ensembles have been performed for CESM2 (a CMIP6-era model) and CESM1 (a CMIP5-era model). We refer to these as CESM2-LE and CESM1-LE. The external forcing used in these simulations has changed to be consistent with their CMIP generation. As a result, differences between CESM2-LE and CESM1-LE ensemble means arise from changes in both model structure and forcing. Here we present new ensemble simulations which allow us to separate the influences of these model structural and forcing differences. Our new CESM2 simulations are run with CMIP5 forcings equivalent to those used in the CESM1-LE. We find a strong influence of historical forcing uncertainty due to aerosol effects on simulated climate. For the historical period, forcing drives reduced global warming and ocean heat uptake in CESM2-LE relative to CESM1-LE that is counteracted by the influence of model structure. The influence of the model structure and forcing vary across the globe, and the Arctic exhibits a distinct signal that contrasts with the global mean. For the 21st century, the importance of scenario forcing differences (SSP3–7.0 for CESM2-LE and RCP8.5 for CESM1-LE) is evident. The new simulations presented here allow us to diagnose the influence of model structure on 21st century change, despite large scenario forcing differences, revealing that differences in the meridional distribution of warming are caused by model structure. Feedback analysis reveals that clouds and their impact on shortwave radiation explain many of these structural differences between CESM2 and CESM1. In the Arctic, albedo changes control transient climate evolution differences due to structural differences between CESM2 and CESM1.

Internal climate variability (ICV) creates a range of climate trajectories, which are superimposed upon the forced response. A single climate model realization may not represent forced change alone and may diverge from other realizations, as well as observations, due to ICV. We use an initial-condition large ensemble of simulations with the Community Earth System Model (CESM2) to show that ICV produces a range of outcomes in the terrestrial carbon cycle. Trends in gross primary production (GPP) from 1991 to 2020 differ among ensemble members due to the different climate trajectories resulting from ICV. We quantify how ICV imparts on GPP trends and apply our methodology to the observational record. Observed changes in GPP at two long-running eddy covariance flux towers are consistent with ICV, challenging the understanding of forced changes in the carbon cycle at these locations. A probabilistic framework that accounts for ICV is needed to interpret carbon cycle trends.

Unabated 21st-century climate change will accelerate Arctic-Subarctic permafrost thaw which can intensify microbial degradation of carbon-rich soils, methane emissions, and global warming. The impact of permafrost thaw on future Arctic-Subarctic wildfires and the associated release of greenhouse gases and aerosols is less well understood. Here we present a comprehensive analysis of the effect of future permafrost thaw on land surface processes in the Arctic-Subarctic region using the CESM2 large ensemble forced by the SSP3-7.0 greenhouse gas emission scenario. Analyzing 50 greenhouse warming simulations, which capture the coupling between permafrost, hydrology, and atmosphere, we find that projected rapid permafrost thaw leads to massive soil drying, surface warming, and reduction of relative humidity over the Arctic-Subarctic region. These combined processes lead to nonlinear late-21st-century regime shifts in the coupled soil-hydrology system and rapid intensification of wildfires in western Siberia and Canada. In a warming world, future rapid permafrost thawing can cause significant soil drying, surface warming, and atmospheric drying in the Arctic-Subarctic region, leading to an intensification of peat and forest fires in western Siberia and Canada.

This paper analyzes fi ndings from semistructured interviews and focus groups with 31 farmers in the Willamette Valley in which farmers were asked about their needs for climate data and about the usability of a range of outputs from the Community Earth System Model, version 2 (CESM2), for their soil management practices. Findings indicate that climate and soils data generated from CESM and other Earth system models (ESMs), despite their coarse spatial scale resolutions, can inform farmers' long-term decisions, but that the data would be more usable if the outputs were provided in a format that allowed farmers to choose the variables and thresholds relevant to their particular needs and if ESMs incorporated farmer practices including residue removal, cover cropping, and tillage levels into the model operations so that farmers could better understand the impacts of their decisions. Findings also suggest that although there is a signifi- fi- cant gap in the spatial resolution at which these global ESMs generate data and the spatial resolution needed by farmers to make most decisions, farmers are adept at making scalar adjustments to apply coarse-resolution data to the specifics fi cs of their own farm's microclimate. Thus, our fi ndings suggest that, to support agricultural decision-making, development priorities for ESMs should include developing better representations of agricultural management practices within the models and creating interactive data dashboards or platforms.

Climate change jeopardizes human health, global biodiversity, and sustainability of the biosphere. To make reliable predictions about climate change, scientists use Earth system models (ESMs) that integrate physical, chemical, and biological processes occurring on land, the oceans, and the atmosphere. Although critical for catalyzing coupled biogeochemical processes, microorganisms have traditionally been left out of ESMs. Here, we generate a "top 10" list of priorities, opportunities, and challenges for the explicit integration of microorganisms into ESMs. We discuss the need for coarse-graining microbial information into functionally relevant categories, as well as the capacity for microorganisms to rapidly evolve in response to climate-change drivers. Microbiologists are uniquely positioned to collect novel and valuable information necessary for next-generation ESMs, but this requires data harmonization and transdisciplinary collaboration to effectively guide adaptation strategies and mitigation policy.

Changes in the nitrogen (N) status of forest ecosystems can directly and indirectly influence their carbon (C) sequestration potential by altering soil organic matter (SOM) decomposition, soil enzyme activity, and plant–soil interactions. However, model representations of linked C–N cycles and SOM decay are not well validated against experimental data. Here, we use extensive data from the Fernow Experimental Forest long-term whole-watershed N fertilization study to compare the response to N perturbations of two soil models that represent decomposition dynamics differently (first-order decay versus microbially explicit reverse Michaelis–Menten kinetics). These two soil models were coupled to a common vegetation model which provided identical input data. Key responses to N additions measured at the study site included a shift in plant allocation to favor woody biomass over belowground carbon inputs, reductions in soil respiration, accumulation of particulate organic matter (POM), and an increase in soil C:N ratios. The vegetation model did not capture the often-observed shift in plant C allocation with N additions, which resulted in poor predictions of the soil responses. We modified the parameterization of the plant C allocation scheme to favor wood production over fine-root production with N additions, which significantly improved the vegetation and soil respiration responses. Additionally, to elicit an increase in the soil C stocks and C:N ratios with N additions, as observed, we modified the decay rates of the POM in the soil models. With these modifications, both models captured negative soil respiration and positive soil C stock responses in line with observations, but only the microbially explicit model captured an increase in soil C:N. Our results highlight the need for further model development to accurately represent plant–soil interactions, such as rhizosphere priming, and their responses to environmental change.

Vegetation and soils are taking up approximately 30% of anthropogenic carbon dioxide emissions because of small imbalances in large gross carbon exchanges from productivity and turnover that are poorly constrained. We combined a new budget of radiocarbon produced by nuclear bomb testing in the 1960s with model simulations to evaluate carbon cycling in terrestrial vegetation. We found that most state-of-the-art vegetation models used in the Coupled Model Intercomparison Project underestimated the radiocarbon accumulation in vegetation biomass. Our findings, combined with constraints on vegetation carbon stocks and productivity trends, imply that net primary productivity is likely at least 80 petagrams of carbon per year presently, compared with the 43 to 76 petagrams per year predicted by current models. Storage of anthropogenic carbon in terrestrial vegetation is likely more short-lived and vulnerable than previously predicted.

Abstract Increased plant growth under elevated carbon dioxide (CO2) slows the pace of climate warming and underlies projections of terrestrial carbon (C) and climate dynamics. However, this important ecosystem service may be diminished by concurrent changes to vegetation carbon‐to‐nitrogen (C:N) ratios. Despite clear observational evidence of increasing foliar C:N under elevated CO2, our understanding of potential ecological consequences of foliar stoichiometric flexibility is incomplete. Here, we illustrate that when we incorporated CO2‐driven increases in foliar stoichiometry into the Community Land Model the projected land C sink decreased two‐fold by the end of the century compared to simulations with fixed foliar chemistry. Further, CO2‐driven increases in foliar C:N profoundly altered Earth's hydrologic cycle, reducing evapotranspiration and increasing runoff, and reduced belowground N cycling rates. These findings underscore the urgency of further research to examine both the direct and indirect effects of changing foliar stoichiometry on soil N cycling and plant productivity.