Search

An exponential rise in the atmospheric vapour pressure deficit (VPD) is among the most consequential impacts of climate change in terrestrial ecosystems. Rising VPD has negative and cascading effects on nearly all aspects of plant function including photosynthesis, water status, growth and survival. These responses are exacerbated by land–atmosphere interactions that couple VPD to soil water and govern the evolution of drought, affecting a range of ecosystem services including carbon uptake, biodiversity, the provisioning of water resources and crop yields. However, despite the global nature of this phenomenon, research on how to incorporate these impacts into resilient management regimes is largely in its infancy, due in part to the entanglement of VPD trends with those of other co‐evolving climate drivers. Here, we review the mechanistic bases of VPD impacts at a range of spatial scales, paying particular attention to the independent and interactive influence of VPD in the context of other environmental changes. We then evaluate the consequences of these impacts within key management contexts, including water resources, croplands, wildfire risk mitigation and management of natural grasslands and forests. We conclude with recommendations describing how management regimes could be altered to mitigate the otherwise highly deleterious consequences of rising VPD. Summary statement: Rising atmospheric vapour pressure deficit (or VPD) is one of the most widespread and significant consequences of climate warming for terrestrial ecosystems. This article reviews the mechanistic bases of these usually deleterious impacts and synthesises that information into a set of management recommendations to mitigate them. [ABSTRACT FROM AUTHOR]

Machine learning (ML)‐based parameterizations have been developed for Earth System Models (ESMs) with the goal to better represent subgrid‐scale processes or to accelerate computations. ML‐based parameterizations within hybrid ESMs have successfully learned subgrid‐scale processes from short high‐resolution simulations. However, most studies used a particular ML method to parameterize the subgrid tendencies or fluxes originating from the compound effect of various small‐scale processes (e.g., radiation, convection, gravity waves) in mostly idealized settings or from superparameterizations. Here, we use a filtering technique to explicitly separate convection from these processes in simulations with the Icosahedral Non‐hydrostatic modeling framework (ICON) in a realistic setting and benchmark various ML algorithms against each other offline. We discover that an unablated U‐Net, while showing the best offline performance, learns reverse causal relations between convective precipitation and subgrid fluxes. While we were able to connect the learned relations of the U‐Net to physical processes this was not possible for the non‐deep learning‐based Gradient Boosted Trees. The ML algorithms are then coupled online to the host ICON model. Our best online performing model, an ablated U‐Net excluding precipitating tracer species, indicates higher agreement for simulated precipitation extremes and mean with the high‐resolution simulation compared to the traditional scheme. However, a smoothing bias is introduced both in water vapor path and mean precipitation. Online, the ablated U‐Net significantly improves stability compared to the non‐ablated U‐Net and runs stable for the full simulation period of 180 days. Our results hint to the potential to significantly reduce systematic errors with hybrid ESMs. Plain Language Summary: Due to their computational costs, it is currently not feasible to run more accurate high‐resolution climate models on a global domain on climate (century) time‐scales. However, high‐accuracy climate simulations are needed for more robust and detailed projections of our future climate. Here, we develop and evaluate various machine learning‐based convection parameterizations learned on reconstructed and coarse‐grained high‐resolution subgrid fluxes to solve this problem, and benchmark their performance. The data set is chosen from simulations of the Icosahedral Non‐hydrostatic modeling framework (ICON) in a realistic setting of the tropical Atlantic and at storm‐resolving resolutions. We focus only on convective subgrid fluxes that are isolated from other components. We improve the best ML algorithms further by excluding variables that cause unphysical correlations. Finally, we explain the learned relations of the best data‐driven schemes based on physical process understanding, test their performance when coupled to the ICON model, and achieve stable coupled simulations for 180 days as well as improved precipitation predictions. Key Points: We train/benchmark machine learning models on convective fluxes derived from realistic coarse‐grained data of storm‐resolving simulationsShapley values reveal that the best offline model, a U‐Net, learns non‐causal links to precipitation and shows poor online performanceA model, without non‐causal precipitation connections, runs more stable coupled to ICON and indicates better precipitation predictions [ABSTRACT FROM AUTHOR]

Multiple lines of evidence confirm a widespread increase in vegetation growth across China over the past few decades. The relationship between vegetation growth and water availability is thought to be becoming stronger under climate change, that is, water constraints on vegetation growth have been increasing. However, our understanding of how water constraints have influenced these vegetation greening trends, especially those climate change‐driven ones, remains limited. Here, we conduct a comprehensive evaluation of recent water constraints and their implications for vegetation growth in China between 1982 and 2015. By analyzing the spatiotemporal patterns of the relationship between vegetation growth and water availability, we reveal recent water constraints and their changes hidden within an overall greening trend in China. Further analysis demonstrates that two climate change‐related categories, defined broadly as "climate" (e.g., air temperature, precipitation, and so on) and "CO2" (i.e., atmospheric carbon dioxide), have exerted varying levels of importance in regulating vegetation growth across different water constraints. With increasing water constraints, the proportion of the climate‐dominated area has significantly risen, while that of the CO2‐dominated area has sharply declined. Our findings highlight that water constraints can mediate the dominance of climate and atmospheric CO2 on vegetation growth. This has the great potential to exacerbate the uncertainty surrounding current and future sustainable vegetation greening trends. Plain Language Summary: China's vegetation growth has shown a broad increase under climate change and human activities over the past few decades. At the same time, changes in vegetation water demand, terrestrial water availability, and climatic factors are expected to lead to increasing water constraints on vegetation growth. However, whether water constraints have influenced climate change‐driven vegetation greening trends remains unclear. In this study, we provide a comprehensive assessment of recent water constraints and their implications for vegetation growth in China between 1982 and 2015. To do this, we analyze the spatiotemporal patterns of the relationship between vegetation growth and water availability to reveal recent water constraints hidden within China's vegetation greening. Further analysis shows that climate and atmospheric CO2 have exerted varying levels of importance in regulating vegetation growth across different water constraints. With increasing water constraints, more regions become climate‐dominated and fewer become CO2‐dominated. Overall, our findings highlight that water constraints can mediate the dominance of climate and atmospheric CO2 on vegetation growth. This will help decision‐makers recognize the uncertainty of sustainable vegetation greening trends. Key Points: We reveal recent water constraints and their changes hidden within China's widespread vegetation greeningClimate and atmospheric CO2 have exerted varying levels of importance in regulating vegetation growth across different water constraintsWith increasing water constraints, more of China's vegetated regions become climate‐dominated and fewer become CO2‐dominated [ABSTRACT FROM AUTHOR]

A promising method for improving the representation of clouds in climate models, and hence climate projections, is to develop machine learning‐based parameterizations using output from global storm‐resolving models. While neural networks (NNs) can achieve state‐of‐the‐art performance within their training distribution, they can make unreliable predictions outside of it. Additionally, they often require post‐hoc tools for interpretation. To avoid these limitations, we combine symbolic regression, sequential feature selection, and physical constraints in a hierarchical modeling framework. This framework allows us to discover new equations diagnosing cloud cover from coarse‐grained variables of global storm‐resolving model simulations. These analytical equations are interpretable by construction and easily transferable to other grids or climate models. Our best equation balances performance and complexity, achieving a performance comparable to that of NNs (R2 = 0.94) while remaining simple (with only 11 trainable parameters). It reproduces cloud cover distributions more accurately than the Xu‐Randall scheme across all cloud regimes (Hellinger distances < 0.09), and matches NNs in condensate‐rich regimes. When applied and fine‐tuned to the ERA5 reanalysis, the equation exhibits superior transferability to new data compared to all other optimal cloud cover schemes. Our findings demonstrate the effectiveness of symbolic regression in discovering interpretable, physically‐consistent, and nonlinear equations to parameterize cloud cover. Plain Language Summary: In climate models, cloud cover is usually expressed as a function of coarse, pixelated variables. Traditionally, this functional relationship is derived from physical assumptions. In contrast, machine learning (ML) approaches, such as neural networks, sacrifice interpretability for performance. In our approach, we use high‐resolution climate model output to learn a hierarchy of cloud cover schemes from data. To bridge the gap between simple statistical methods and ML algorithms, we employ a symbolic regression method. Unlike classical regression, which requires providing a set of basis functions from which the equation is composed of, symbolic regression only requires mathematical operators (such as +, ×) that it learns to combine. By using a genetic algorithm, inspired by the process of natural selection, we discover an interpretable, nonlinear equation for cloud cover. This equation is simple, performs well, satisfies physical principles, and outperforms other algorithms when applied to new observationally‐informed data. Key Points: We systematically derive and evaluate cloud cover parameterizations of various complexity from global storm‐resolving simulation outputUsing symbolic regression combined with physical constraints, we find a new interpretable equation balancing performance and simplicityOur data‐driven cloud cover equation can be retuned with few samples, facilitating transfer learning to generalize to other realistic data [ABSTRACT FROM AUTHOR]

Climate models are essential to understand and project climate change, yet long‐standing biases and uncertainties in their projections remain. This is largely associated with the representation of subgrid‐scale processes, particularly clouds and convection. Deep learning can learn these subgrid‐scale processes from computationally expensive storm‐resolving models while retaining many features at a fraction of computational cost. Yet, climate simulations with embedded neural network parameterizations are still challenging and highly depend on the deep learning solution. This is likely associated with spurious non‐physical correlations learned by the neural networks due to the complexity of the physical dynamical system. Here, we show that the combination of causality with deep learning helps removing spurious correlations and optimizing the neural network algorithm. To resolve this, we apply a causal discovery method to unveil causal drivers in the set of input predictors of atmospheric subgrid‐scale processes of a superparameterized climate model in which deep convection is explicitly resolved. The resulting causally‐informed neural networks are coupled to the climate model, hence, replacing the superparameterization and radiation scheme. We show that the climate simulations with causally‐informed neural network parameterizations retain many convection‐related properties and accurately generate the climate of the original high‐resolution climate model, while retaining similar generalization capabilities to unseen climates compared to the non‐causal approach. The combination of causal discovery and deep learning is a new and promising approach that leads to stable and more trustworthy climate simulations and paves the way toward more physically‐based causal deep learning approaches also in other scientific disciplines. Plain Language Summary: Climate models have biases compared to observations that have been present for a long time because certain processes, like convection, are only approximated using simplified methods. Neural networks can better represent these processes, but often learn incorrect connections leading to unreliable results and climate model crashes. To solve this, we used a new method that informs neural networks with causal drivers, therefore, respecting the underlying physical processes. By doing so, we developed more reliable and trustworthy neural networks, allowing us to accurately represent the climate of the original high‐resolution simulation on which these neural networks were trained. Key Points: Causal discovery unveils causal drivers of subgrid‐scale processes across climatesThe causally‐informed hybrid model runs stably and generates a climate close to the original high‐resolution simulationSpurious correlations are evident in the non‐causal parameterization, leading to underestimate feature importance of physical drivers [ABSTRACT FROM AUTHOR]

Summary: Accounting for water limitation is key to determining vegetation sensitivity to drought. Quantifying water limitation effects on evapotranspiration (ET) is challenged by the heterogeneity of vegetation types, climate zones and vertically along the rooting zone.Here, we train deep neural networks using flux measurements to study ET responses to progressing drought conditions. We determine a water stress factor (fET) that isolates ET reductions from effects of atmospheric aridity and other covarying drivers. We regress fET against the cumulative water deficit, which reveals the control of whole‐column moisture availability.We find a variety of ET responses to water stress. Responses range from rapid declines of fET to 10% of its water‐unlimited rate at several savannah and grassland sites, to mild fET reductions in most forests, despite substantial water deficits. Most sensitive responses are found at the most arid and warm sites.A combination of regulation of stomatal and hydraulic conductance and access to belowground water reservoirs, whether in groundwater or deep soil moisture, could explain the different behaviors observed across sites. This variety of responses is not captured by a standard land surface model, likely reflecting simplifications in its representation of belowground water storage. [ABSTRACT FROM AUTHOR]

Although there is mounting concern about how high vapor pressure deficit (VPD) and low soil moisture (SM) affect ecosystem productivity, their relative importance is still under debate. Here, we comprehensively quantified the relative impacts of these two factors on ecosystem gross primary production (GPP) using observations from a global network of eddy‐covariance towers and two approaches (sensitivity analysis and linear regression model). Both approaches agree that a higher percentage of sites experience GPP reduction from high VPD than from low SM over the growing season. However, the constraint of high VPD and low SM on GPP reduction is tightly linked with climates and plant functional types. Humid and mesic ecosystems including forests and grasslands are dominated by VPD, while the semi‐arid and arid ecosystems including shrublands and savannas are dominated by SM. The varying dominant role of these two factors on GPP is closely related to plant stomatal behavior, as predicted by a stomatal conductance model. Additionally, we highlight the non‐linear impact of SM on GPP during droughts and the possible underestimation of the SM effects for deep‐rooted plants when only using surface‐layer SM. Our results shed light on a better understanding of the impacts of VPD and SM on vegetation productivity, with important implications for modeling the response and feedback of ecosystem dynamics to current and future climates. Plain Language Summary: Vapor pressure deficit (VPD) has increased over recent decades and is expected to continue to rise in the future with climate change. Projected changes in precipitation and soil moisture (SM) are more uncertain with higher spatial variability and larger projected intermodel spread. Given the global environmental changes underway, it is of paramount importance to independently assess the effects of high VPD and low SM on ecosystem productivity. Here, we assessed the relative impacts of these two factors on ecosystem gross primary production (GPP) using observations from a global network of eddy‐covariance towers. We find that reductions in GPP are more attributable to high VPD rather than low SM, on average. The constraint of these two factors on GPP reduction is tightly linked with climates and plant functional types. We further emphasize the amplified impact of SM on GPP during droughts and the need to consider using deep SM. Overall, our findings provide a more comprehensive and quantitative attribution of the impact of increasing atmospheric dryness and changes in soil water deficit on ecosystem productivity in a changing climate. Key Points: Reductions in gross primary production (GPP) are more attributable to high vapor pressure deficit (VPD) rather than low soil moisture (SM) across all sitesThe constraint of high VPD and low SM on GPP reduction is tightly linked with climates and plant functional typesOur results highlight that the relative and non‐linear impacts of VPD and SM on GPP should be adequately considered [ABSTRACT FROM AUTHOR]

The increasing frequency and intensity of climate extremes and complex ecosystem responses motivate the need for integrated observational studies at low latency to determine biosphere responses and carbon‐climate feedbacks. Here, we develop a satellite‐based rapid attribution workflow and demonstrate its use at a 1–2‐month latency to attribute drivers of the carbon cycle feedbacks during the 2020–2021 Western US drought and heatwave. In the first half of 2021, concurrent negative photosynthesis anomalies and large positive column CO2 anomalies were detected with satellites. Using a simple atmospheric mass balance approach, we estimate a surface carbon efflux anomaly of 132 TgC in June 2021, a magnitude corroborated independently with a dynamic global vegetation model. Integrated satellite observations of hydrologic processes, representing the soil–plant–atmosphere continuum (SPAC), show that these surface carbon flux anomalies are largely due to substantial reductions in photosynthesis because of a spatially widespread moisture‐deficit propagation through the SPAC between 2020 and 2021. A causal model indicates deep soil moisture stores partially drove photosynthesis, maintaining its values in 2020 and driving its declines throughout 2021. The causal model also suggests legacy effects may have amplified photosynthesis deficits in 2021 beyond the direct effects of environmental forcing. The integrated, observation framework presented here provides a valuable first assessment of a biosphere extreme response and an independent testbed for improving drought propagation and mechanisms in models. The rapid identification of extreme carbon anomalies and hotspots can also aid mitigation and adaptation decisions. [ABSTRACT FROM AUTHOR]

Monitoring and estimating drought impact on plant physiological processes over large regions remains a major challenge for remote sensing and land surface modeling, with important implications for understanding plant mortality mechanisms and predicting the climate change impact on terrestrial carbon and water cycles. The Orbiting Carbon Observatory 3 (OCO‐3), with its unique diurnal observing capability, offers a new opportunity to track drought stress on plant physiology. Using radiative transfer and machine learning modeling, we derive a metric of afternoon photosynthetic depression from OCO‐3 solar‐induced chlorophyll fluorescence (SIF) as an indicator of plant physiological drought stress. This unique diurnal signal enables a spatially explicit mapping of plants' physiological response to drought. Using OCO‐3 observations, we detect a widespread increasing drought stress during the 2020 southwest US drought. Although the physiological drought stress is largely related to the vapor pressure deficit (VPD), our results suggest that plants' sensitivity to VPD increases as the drought intensifies and VPD sensitivity develops differently for shrublands and grasslands. Our findings highlight the potential of using diurnal satellite SIF observations to advance the mechanistic understanding of drought impact on terrestrial ecosystems and to improve land surface modeling. [ABSTRACT FROM AUTHOR]

The effects of precipitation (Pr) and evapotranspiration (ET) on surface soil moisture (SSM) play an essential role in the land‐atmosphere system. Here we evaluate multimodel differences of these effects within the Coupled Model Intercomparison Project Phase 5 (CMIP5) compared to Soil Moisture Active Passive (SMAP) products and ECMWF Reanalysis v5 (ERA5) as references in a frequency domain. The variability of SSM, Pr, and ET within three frequency bands (1/7 ∼ 1/30 days−1, 1/30 ∼ 1/90 days−1, and 1/90 ∼ 1/365 days−1) after normalization is quantified using Fourier transform. We analyze the impact of ET and Pr on SSM variability based on a transfer function assuming that these variables form a linear time‐invariant (LTI) system. For the total effects of ET and Pr on SSM variability, the CMIP5 estimations are smaller than the reference data in the two higher frequency bands and are larger than the reference data in the lowest frequency band. Besides, the effects on SSM by Pr and ET are found to be different across the three frequency bands. In each frequency band, the variability of the factor that dominates SSM (i.e., Pr or ET) from CMIP5 is smaller than that from the references. This study identifies the spatiotemporal distribution of differences between CMIP5 models and references (SMAP and ERA5) in simulating ET and Pr effects on SSM within three frequency bands. This study provides insightful information on how soil moisture variability is affected by varying precipitation and evapotranspiration at different time scales within Earth System Models. Plain Language Summary: Climate is influenced by the interactions between the land surface and atmosphere boundary, and soil moisture is a key component of these physical processes. Precipitation and evapotranspiration, as two major variables involved in these interactions, have been largely regarded as essential processes affecting soil moisture dynamics. However, Earth System Models have large uncertainties in simulating these effects. This study compares the average performance of 14 Earth System Models in capturing the effects of precipitation and evapotranspiration on surface soil moisture variability. We find that (a) soil moisture is mainly affected by precipitation at weekly to seasonal time scales and by evapotranspiration at seasonal to annual time scales; (b) compared to two largely used reference data, the total effects of precipitation and evapotranspiration on soil moisture is smaller at weekly to seasonal time scales and are larger at seasonal to annual time scale; and (c) spatially, models tend to simulate less variability of precipitation or evapotranspiration as a major control on surface soil moisture. Key Points: The effects of precipitation and evapotranspiration on soil moisture variability can be analyzed in a frequency domainPrecipitation dominates weekly to seasonal variability and evapotranspiration dominates seasonal to annual variability of soil moistureEarth System Models shall be improved in simulating soil moisture temporal variabilities [ABSTRACT FROM AUTHOR]

A promising approach to improve cloud parameterizations within climate models and thus climate projections is to use deep learning in combination with training data from storm‐resolving model (SRM) simulations. The ICOsahedral Non‐hydrostatic (ICON) modeling framework permits simulations ranging from numerical weather prediction to climate projections, making it an ideal target to develop neural network (NN) based parameterizations for sub‐grid scale processes. Within the ICON framework, we train NN based cloud cover parameterizations with coarse‐grained data based on realistic regional and global ICON SRM simulations. We set up three different types of NNs that differ in the degree of vertical locality they assume for diagnosing cloud cover from coarse‐grained atmospheric state variables. The NNs accurately estimate sub‐grid scale cloud cover from coarse‐grained data that has similar geographical characteristics as their training data. Additionally, globally trained NNs can reproduce sub‐grid scale cloud cover of the regional SRM simulation. Using the game‐theory based interpretability library SHapley Additive exPlanations, we identify an overemphasis on specific humidity and cloud ice as the reason why our column‐based NN cannot perfectly generalize from the global to the regional coarse‐grained SRM data. The interpretability tool also helps visualize similarities and differences in feature importance between regionally and globally trained column‐based NNs, and reveals a local relationship between their cloud cover predictions and the thermodynamic environment. Our results show the potential of deep learning to derive accurate yet interpretable cloud cover parameterizations from global SRMs, and suggest that neighborhood‐based models may be a good compromise between accuracy and generalizability. Plain Language Summary: Climate models, such as the ICOsahedral Non‐hydrostatic climate model, operate on low‐resolution grids, making it computationally feasible to use them for climate projections. However, physical processes –especially those associated with clouds– that happen on a sub‐grid scale (inside a grid box) cannot be resolved, yet they are critical for the climate. In this study, we train neural networks that return the cloudy fraction of a grid box knowing only low‐resolution grid‐box averaged variables (such as temperature, pressure, etc.) as the climate model sees them. We find that the neural networks can reproduce the sub‐grid scale cloud fraction on data sets similar to the one they were trained on. The networks trained on global data also prove to be applicable on regional data coming from a model simulation with an entirely different setup. Since neural networks are often described as black boxes that are therefore difficult to trust, we peek inside the black box to reveal what input features the neural networks have learned to focus on and in what respect the networks differ. Overall, the neural networks prove to be accurate methods of reproducing sub‐grid scale cloudiness and could improve climate model projections when implemented in a climate model. Key Points: Neural networks can accurately learn sub‐grid scale cloud cover from realistic regional and global storm‐resolving simulationsThree neural network types account for different degrees of vertical locality and differentiate between cloud volume and cloud area fractionUsing a game theory based library we find that the neural networks tend to learn local mappings and are able to explain model errors [ABSTRACT FROM AUTHOR]

Previous studies seldom consider humidity when examining heat‐related extremes, and none have explored the effects of humidity on concurrent extremes of high heat stress and low river streamflow. Here, we present the first global picture of projected changes in compound lethal heat stress (Th)‐drought hazards (CHD) across 11,637 catchments. Our observational datasets show that atmospheric conditions (e.g., energy and vapor flux) play an important role in constraining the heat extremes, and that Th (32% ± 11%) yields a higher coincidence rate of global CHD than wet‐bulb temperature (28% ± 11%), driven by lower relative humidity (RH) and thus air dryness in Th extremes. Our large model ensemble projects a 10‐fold intensification of bivariate CHD risks by 2071–2100, mainly driven by increases in heat extremes. Future declines in RH, wind, snow, and precipitation in many regions are likely to exacerbate such water and weather‐related hazards (e.g., drought and CHD). Plain Language Summary: Water and weather‐related extremes such as droughts and heatwaves typically occur simultaneously, posing larger risks to humans and ecosystem than any individual hazard. Extreme high lethal heat stress, which is governed by both temperature and humidity, inhibits the evaporative cooling function of the human body. However, changes in lethal heat stress under past and future climate are poorly understood at the global scale, and no study has yet explored their influences on the risk of compound heat stress‐drought events. Here, we provide the first global picture of the concurrent high heat stress–low streamflow extremes and their projected changes. The fraction of lethal heat extremes accompanied by drought hazards is projected to rise markedly throughout the 21st century. The global magnitude and occurrence of compound hazards is projected to intensify by 4–10 times by 2071–2100, mainly due to the increasing severity of heat extremes. Our study reveals an increasing global risk of compounding hazards, highlighting the need to better prepare adaptation and mitigation solutions in the Anthropocene. Key Points: We present the first global assessment of compound extremes of high lethal heat stress and low river streamflowLethal heat stress yields a higher coincidence rate of compound hazards than wet‐bulb temperatureFuture lethal heat stress‐drought events are mainly exacerbated by changes in heat extremes [ABSTRACT FROM AUTHOR]

Deep learning can accurately represent sub‐grid‐scale convective processes in climate models, learning from high resolution simulations. However, deep learning methods usually lack interpretability due to large internal dimensionality, resulting in reduced trustworthiness in these methods. Here, we use Variational Encoder Decoder structures (VED), a non‐linear dimensionality reduction technique, to learn and understand convective processes in an aquaplanet superparameterized climate model simulation, where deep convective processes are simulated explicitly. We show that similar to previous deep learning studies based on feed‐forward neural nets, the VED is capable of learning and accurately reproducing convective processes. In contrast to past work, we show this can be achieved by compressing the original information into only five latent nodes. As a result, the VED can be used to understand convective processes and delineate modes of convection through the exploration of its latent dimensions. A close investigation of the latent space enables the identification of different convective regimes: (a) stable conditions are clearly distinguished from deep convection with low outgoing longwave radiation and strong precipitation; (b) high optically thin cirrus‐like clouds are separated from low optically thick cumulus clouds; and (c) shallow convective processes are associated with large‐scale moisture content and surface diabatic heating. Our results demonstrate that VEDs can accurately represent convective processes in climate models, while enabling interpretability and better understanding of sub‐grid‐scale physical processes, paving the way to increasingly interpretable machine learning parameterizations with promising generative properties. Plain Language Summary: Deep neural nets are hard to interpret due to their hundred thousand or million trainable parameters without further postprocessing. We demonstrate in this paper the usefulness of a network type that is designed to drastically reduce this high dimensional information in a lower‐dimensional space to enhance the interpretability of predictions compared to regular deep neural nets. Our approach is, on the one hand, able to reproduce small‐scale cloud related processes in the atmosphere learned from a physical model that simulates these processes skillfully. On the other hand, our network allows us to identify key features of different cloud types in the lower‐dimensional space. Additionally, the lower‐order manifold separates tropical samples from polar ones with a remarkable skill. Overall, our approach has the potential to boost our understanding of various complex processes in Earth System science. Key Points: A Variational Encoder Decoder (VED) can predict sub‐grid‐scale thermodynamics from the coarse‐scale climate stateThe VED's latent space can distinguish convective regimes, including shallow/deep/no convectionThe VED's latent space reveals the main sources of convective predictability at different latitudes [ABSTRACT FROM AUTHOR]

Tree resistance to extreme droughts and post‐drought recovery are sensitive to the drought timing. However, how the bioclimatic sensitivity of tree growth may vary with the timing and order of extreme droughts and wetness is still poorly understood. Here, we quantified the bioclimatic sensitivity of tree growth in the period of 1951–2013 under different seasonal extreme drought/wetness regimes over the extra‐tropical Northern Hemisphere, using 1,032 tree ring chronologies from 121 gymnosperm and angiosperm species. We found a negative asymmetry in tree growth under regimes with seasonal extreme droughts. With extreme drought, tree growth in arid and temperate dry regions is more negatively impacted by pre‐growing‐season (PGS) extreme droughts. Clade‐wise, angiosperms are more sensitive to PGS water availability, and gymnosperms to legacy effects of the preceding tree growth conditions in temperate dry and humid regions. Our finding of divergent bioclimatic legacy effects underscores contrasting trends in forest responses to drought across different ecoregions and functional groups in a more extreme climate. Plain Language Summary: The occurrence patterns of seasonal extreme drought and wetness events are dramatically shifting with climate warming. However, how will different seasonal extreme climate regimes affect the bioclimatic sensitivity of tree growth remains poorly understood. In this study, we investigated the sensitivity of tree growth to different seasonal climate factors and preceding tree growth conditions during 1951–2013 under different seasonal extreme drought/wetness regimes, using 1,032 tree ring chronologies covering 121 gymnosperm and angiosperm species. We found the magnitude in tree growth reduction caused by seasonal extreme drought events is much larger than that in tree growth stimulation by seasonal extreme wetness events in arid and temperate regions. Tree growth in arid and temperate dry regions is more negatively impacted by extreme drought events in pre‐growing‐seasons (PGSs) than in growing‐seasons. We further found that angiosperms are more sensitive to PGS water availability, while gymnosperms are more sensitive to legacy effects of the preceding tree growth conditions in temperate dry and humid regions. These findings highlight divergent bioclimatic legacy effects on tree growth under different seasonal extreme climate regimes, and provide valuable insights into the future trajectories of forest growth across diverse ecoregions and functional groups in a more extreme climate. Key Points: Negative asymmetry in tree growth was observed under regimes with seasonal extreme droughtsTree growth in arid and temperate dry regions is more vulnerable to extreme droughts in pre‐growing‐seasons than in growing‐seasonsAngiosperms are more sensitive to pre‐growing‐season water availability, and gymnosperms to legacy effects of the preceding tree growth [ABSTRACT FROM AUTHOR]

In large‐eddy simulations, subgrid‐scale (SGS) processes are parameterized as a function of filtered grid‐scale variables. First‐order, algebraic SGS models are based on the eddy‐viscosity assumption, which does not always hold for turbulence. Here we apply supervised deep neural networks (DNNs) to learn SGS stresses from a set of neighboring coarse‐grained velocity from direct numerical simulations of the convective boundary layer at friction Reynolds numbers Reτ up to 1243 without invoking the eddy‐viscosity assumption. The DNN model was found to produce higher correlation between SGS stresses compared to the Smagorinsky model and the Smagorinsky‐Bardina mixed model in the surface and mixed layers and can be applied to different grid resolutions and various stability conditions ranging from near neutral to very unstable. The DNN model can capture key statistics of turbulence in a posteriori (online) tests when applied to large‐eddy simulations of the atmospheric boundary layer. Plain Language Summary: Using grid‐scale state variables to represent subgrid‐scale (SGS) processes is a major step in large‐eddy simulations of turbulence, which typically rely on some physically based assumptions that do not always apply. Here we replace physically based assumptions with deep learning in SGS modeling and show the latter performs better in a priori (offline) tests of atmospheric turbulence. In addition, the deep neural networks model well captures key statistics of turbulence in a posteriori (online) tests when applied to in large‐eddy simulations of atmospheric boundary layers. Key Points: Deep learning‐based subgrid‐scale models outperform traditional eddy‐viscosity models in a priori (offline) testsDeep learning‐based subgrid‐scale models can account for density stratification effects in boundary‐layer turbulenceThe DNN model can capture key turbulence statistics in a posteriori (online) tests when applied to large‐eddy simulations of the atmospheric boundary layer [ABSTRACT FROM AUTHOR]

Vegetation is a key component in the global carbon cycle as it stores ~450 GtC as biomass, and removes about a third of anthropogenic CO2 emissions. However, in some regions, the rate of plant carbon uptake is beginning to slow, largely because of water stress. Here, we develop a new observation‐based methodology to diagnose vegetation water stress and link it to environmental drivers. We used the ratio of remotely sensed land surface to near surface atmospheric temperatures (LST/Tair) to represent vegetation water stress, and built regression tree models (random forests) to assess the relationship between LST/Tair and the main environmental drivers of surface energy fluxes in the tropical Americas. We further determined ecosystem traits associated with water stress and surface energy partitioning, pinpointed critical thresholds for water stress, and quantified changes in ecosystem carbon uptake associated with crossing these critical thresholds. We found that the top drivers of LST/Tair, explaining over a quarter of its local variability in the study region, are (1) radiation, in 58% of the study region; (2) water supply from precipitation, in 30% of the study region; and (3) atmospheric water demand from vapor pressure deficits (VPD), in 22% of the study region. Regions in which LST/Tair variation is driven by radiation are located in regions of high aboveground biomass or at high elevations, while regions in which LST/Tair is driven by water supply from precipitation or atmospheric demand tend to have low species richness. Carbon uptake by photosynthesis can be reduced by up to 80% in water‐limited regions when critical thresholds for precipitation and air dryness are exceeded simultaneously, that is, as compound events. Our results demonstrate that vegetation structure and diversity can be important for regulating surface energy and carbon fluxes over tropical regions. [ABSTRACT FROM AUTHOR]

Increasing atmospheric CO2 and associated global warming are expected to alter the global hydrological cycle, thereby posing widespread threats to freshwater availability. However, future hydrological projections differ greatly between models, particularly over the tropical regions. The large difference between model projections directly limits policy planning efforts, and the responsible modeling processes remain unclear. Here, we identify the primary processes accounting for model differences in tropical hydrological changes using multiple CO2 sensitivity experiments in the Coupled Model Intercomparison Project. We show that differences in projected changes to tropical evapotranspiration, precipitation, and surface water availability mainly arise from model representations of vegetation cover and stomatal conductance responses to elevated CO2 and associated changes in atmospheric moisture and circulation. Atmospheric responses to sea surface warming contribute additionally to the divergence in hydrological projections. Given the importance of vegetation responses to elevated CO2 and associated atmosphere feedbacks, our results underscore the need to improve representations of the vegetation physiological response to rising CO2 and its coupling to the atmosphere, to provide reliable tropical hydrological projections. Plain Language Summary: Future hydrological changes pose great challenges to global freshwater and food security. However, hydrological projections diverge greatly between models, especially over tropical lands, which directly impedes planning for adaptation to climate change. Projection divergence is largely thought to be caused by differences in the representation of sea surface warming effect. Our study shows, however, that large divergence in tropical hydrological projections occurs immediately, in response to abrupt CO2 quadrupling, before slower sea surface warming effect comes into play. Our results identify differences in the representation of fast vegetation physiological responses to elevated CO2 and associated atmospheric moisture and circulation changes, as the primary cause of divergent long‐term hydrological projections over the tropical regions. Key Points: Future hydrological projections diverge greatly between models, especially over tropical landsThe divergence mainly arises from fast vegetation physiological responses to elevated CO2 and associated atmosphere feedbacksAtmospheric responses to slow sea surface warming play a second role in the divergence of hydrological projections [ABSTRACT FROM AUTHOR]

There are significant biases and uncertainties in the simulated soil moisture with land surface models. Here we evaluate multimodel differences in Coupled Model Intercomparison Project Phase 5 (CMIP5) compared to Soil Moisture Active Passive (SMAP) products on different time scales. The variability of surface soil moisture (SSM) within three frequency bands (7–30 days, 30–90 days, and 90–365 days) after normalization is quantified using Fourier transform for the evaluation. Compared to the SMAP observations, the simulated SSM variability within CMIP5 is underestimated in the two higher frequency bands (by 72% and 56%, respectively) and overestimated in the lowest frequency band (by 113%). In addition, these differences concentrate in regions with larger SSM. Finally, these multimodel differences are found to be significantly correlated with mean climate conditions rather than soil texture. This study identifies the spatiotemporal distribution of the model deficiencies within CMIP5 and finds they are systematic in the long‐term simulation on a global scale. Plain Language Summary: Soil moisture has been largely regarded as a key variable in Earth system and plays an important role in climate prediction. However, land surface models have large uncertainties in simulating soil moisture. This study identifies that (a) land surface models underestimate soil moisture variability on weekly to seasonal time scales and overestimate it on seasonal to annual time scales compared to a remote sensing observation, (b) both the underestimation and overestimation are concentrated in the wetter regions, and (c) the differences between these models and the observation are more closely related to vegetation condition and surface temperature than soil sand and clay content. Using satellite observed data, this study reveals the deficiencies of land surface models in simulating temporal variability of soil moisture, which will help improve the soil moisture predictability of these models. Key Points: Land surface models tend to underestimate weekly to seasonal variability and overestimate seasonal to annual variability of soil moistureSimulated surface soil moisture variability is more closely related to climate conditions than soil textureModel‐based soil moisture data may need to be improved to capture soil moisture variability on various frequency bands [ABSTRACT FROM AUTHOR]

Our limited understanding of the impacts of drought on tropical forests significantly impedes our ability in accurately predicting the impacts of climate change on this biome. Here, we investigated the impact of drought on the dynamics of forest canopies with different heights using time‐series records of remotely sensed Ku‐band vegetation optical depth (Ku‐VOD), a proxy of top‐canopy foliar mass and water content, and separated the signal of Ku‐VOD changes into drought‐induced reductions and subsequent non‐drought gains. Both drought‐induced reductions and non‐drought increases in Ku‐VOD varied significantly with canopy height. Taller tropical forests experienced greater relative Ku‐VOD reductions during drought and larger non‐drought increases than shorter forests, but the net effect of drought was more negative in the taller forests. Meta‐analysis of in situ hydraulic traits supports the hypothesis that taller tropical forests are more vulnerable to drought stress due to smaller xylem‐transport safety margins. Additionally, Ku‐VOD of taller forests showed larger reductions due to increased atmospheric dryness, as assessed by vapor pressure deficit, and showed larger gains in response to enhanced water supply than shorter forests. Including the height‐dependent variation of hydraulic transport in ecosystem models will improve the simulated response of tropical forests to drought. [ABSTRACT FROM AUTHOR]

Documenting the large scale variability of tropical forest structure and function is needed for improved understanding of the carbon and water cycles. The seasonal and diurnal cycles of passive and active microwave satellite observations are jointly analyzed for the first time, using the Global Precipitation Mission (GPM). Collocated backscattering coefficients from the GPM Dual‐frequency Precipitation Radar (DPR) and emissivities derived from the GPM Microwave Imager (GMI) are studied for 4 yr over the Tropics, at the available multiple frequencies, angles, and polarizations. Our analysis confirms the diurnal patterns already observed with radars, with a maximum backscatter in the early morning attributed to the water xylem refill at night. At the same time, early in the morning, we evidence a minimum emissivity over most of tropical forests: the backscatter and emissivity diurnal cycles tend to be in phase opposition, as well as their seasonal cycles. However, during the dry season in the south eastern Amazonian forest where precipitation can be low for several months, the emissivity diurnal cycle changes, implying a different behavior of the forest in this area, related to the state of the canopy. Plain Language Summary: The tropical forest plays an important role in the global water and carbon cycles, and it can be analyzed with satellite observations. The seasonal and diurnal cycles of passive and active microwave satellite observations are studied here simultaneously for the first time, thanks to the radar and radiometer on board the Global Precipitation Mission. Four years of observations are collected and analyzed over the Tropics. Our analysis shows that the passive and active observations vary both seasonally and diurnally in tropical forests, with similar amplitude within a day than over the year. The radar and radiometer signals tend to vary in phase opposition, seasonally and diurnally, over most of the tropical forests. However, during the dry season in the south eastern Amazonian forest where precipitation can be low for several months in a row, changes in the seasonal and diurnal cycles of the radiometer observations are evidenced, implying a different behavior of the forest in this area, related to the state of the canopy. Key Points: Strong diurnal cycles over tropical forests, for both microwave backscatter and emissivity, with amplitude similar to seasonal variationsSeasonal and diurnal cycles of backscattering and emissivity temporally anti‐correlated over most parts of the tropical forestsOver Amazonia, seasonal cycle of backscatter and emissivity in phase in dry regions and diurnal cycle also in phase there during dry season [ABSTRACT FROM AUTHOR]

Droughts in a warming climate have become more common and more extreme, making understanding forest responses to water stress increasingly pressing. Analysis of water stress in trees has long focused on water potential in xylem and leaves, which influences stomatal closure and water flow through the soil‐plant‐atmosphere continuum. At the same time, changes of vegetation water content (VWC) are linked to a range of tree responses, including fluxes of water and carbon, mortality, flammability, and more. Unlike water potential, which requires demanding in situ measurements, VWC can be retrieved from remote sensing measurements, particularly at microwave frequencies using radar and radiometry. Here, we highlight key frontiers through which VWC has the potential to significantly increase our understanding of forest responses to water stress. To validate remote sensing observations of VWC at landscape scale and to better relate them to data assimilation model parameters, we introduce an ecosystem‐scale analog of the pressure–volume curve, the non‐linear relationship between average leaf or branch water potential and water content commonly used in plant hydraulics. The sources of variability in these ecosystem‐scale pressure‐volume curves and their relationship to forest response to water stress are discussed. We further show to what extent diel, seasonal, and decadal dynamics of VWC reflect variations in different processes relating the tree response to water stress. VWC can also be used for inferring belowground conditions—which are difficult to impossible to observe directly. Lastly, we discuss how a dedicated geostationary spaceborne observational system for VWC, when combined with existing datasets, can capture diel and seasonal water dynamics to advance the science and applications of global forest vulnerability to future droughts. [ABSTRACT FROM AUTHOR]

Currently available evapotranspiration (ET) products have not provided us with an accurate estimation for global irrigated land area. Thermal energy balance (EB) model has the potential to solve this problem. The accurate estimation of aerodynamic resistances is a major complication in most remote sensing ET models. An EB model using a column canopy‐air turbulent heat diffusion method was developed to more realistically depict dynamic changes in aerodynamic resistance. In order to estimate global ET and land surface fluxes for all weather conditions, Moderate Resolution Imaging Spectroradiometer (MODIS) Aqua and Terra land surface temperature fields were combined and a nearest‐evaporative‐fraction gap‐filling method was merged into the EB model. A global ET product covering the period 2003–2017 was produced using the EB model. By combining thermal and optional information from MODIS satellites and surface meteorological forcing data from ERA‐Interim reanalysis data, the EB model provides a 5 × 5 km resolution estimate of daily ET without spatio‐temporal gaps at the global scale. Assessment of the daily EB ET at 238 flux sites showed a mean bias of 0.05 mm/day and an RMSE of 1.56 mm/day. Analysis of global precipitation minus ET demonstrated that EB ET has a relatively higher potential for agriculture water resource management than currently available global ET products, such as Landflux, GLEAM, MOD16, GLDAS, and ERA‐Interim ET. In addition, the EB model developed in this study can be applied to both polar and geostationary satellite thermal sensors. Plain Language Summary: A new global evapotranspiration (ET) product was derived from satellite data using an energy balance (EB) model. The product has a daily scale and at a high spatial resolution (5 km for a global scale) for 2003–2017, which is particularly useful for agricultural studies. The model makes use of satellite remotely sensed land surface temperature (LST) and roughness length to calculate instantaneous sensible and latent heat fluxes. These instantaneous fluxes were further used to estimate daily ET. To provide a spatially and temporally continuous map of daily ET globally, two gap‐filling methods were employed to correct for the gaps in LST and evaporative fraction products under cloudy sky conditions. The daily ET product has been successfully evaluated with ground based flux tower data. The global ET dataset has a relatively higher potential for agriculture water resource management than currently available global ET products. In addition, the EB model developed in this study can be applied to other polar and geostationary satellites. Key Points: An energy balance (EB) model was developed to take into account more accurate dynamic changes of aerodynamic resistance than other evapotranspiration (ET) modelsThe thermal EB model can estimate daily ET for all weather conditionsThe EB ET data has a higher potential for application in agriculture water resource management than most of the global ET products [ABSTRACT FROM AUTHOR]

Plants employ a range of strategies to modulate the impact of water stress, including changes to rooting depth and hydraulic conductance (e.g., xylem conductance). However, it is still poorly understood how these strategies vary in relation to climate and land cover types and how they could coordinate globally. Based on daily microwave vegetation optical depth (VOD) from AMSR‐E and AMSR2 over 2002–2011, we estimate two proxies for stress regulation: (a) an effective plant rooting depth (Zr∗) and (b) the effective plant hydraulic conductance (Kplant∗) to delineate two strategies: deep rooting and strong physiological regulation. We find that plants with deeper Zr∗ (e.g., evergreen/deciduous broadleaf forest) are mostly distributed in warm or wet regions, and maintain a relatively steady nighttime VOD because of access to deeper water. Taller plants exhibit greater drop in daytime VOD due to their greater physiological vulnerability. While physiological regulation appears to be the dominant water stress regulation strategy at Northern high latitudes where open shrubland and (woody) savannas are distributed, this physiological regulation is coupled with deep rooting in forest and (woody) savanna areas in the tropics, Eastern US, and Southeastern China. Meanwhile, grasslands in the Western US, Central Asia, Northeastern China, and Mongolia Plateau may be the regions most susceptible to water stress impact because neither water stress mitigation strategy is present. This new framework paves the road for a better understanding of plant water stress strategies at the global scale, and for enhancing large‐scale drought prediction and drought impact assessment in Earth system models by improving plant water stress response. Key Points: Plants with deeper roots maintain relatively steady nighttime vegetation optical depth (VOD) and taller plants drop daytime VOD morePhysiological regulation is the dominant strategy in Northern high latitudes, but it couples with deep rooting in the tropics, Eastern USGrassland areas in the Western US, Central Asia, Northeastern China, and Mongolia Plateau lack strategies for water stress [ABSTRACT FROM AUTHOR]

Amazonian evergreen forests show distinct canopy phenology and photosynthetic seasonality but the climatic triggers are not well understood. This imposes a challenge for modeling leaf phenology and photosynthesis seasonality in land surface models (LSMs) across Amazonian evergreen forest biome. On continental scale, we tested two climatic triggers suggested by site observations, vapor pressure deficit (VPD), and short‐wave incoming radiation (SW) for defining leaf shedding and incorporated VPD‐ and SW‐triggered new canopy phenology modules in the ORCHIDEE LSM (hereafter VPD‐AP and SW‐AP versions). Our results show that both VPD and SW are plausible precursors of large scale litterfall seasonality across the basin by comparing against in situ data from 14 sites. Specially, both VPD‐AP and SW‐AP correctly capture the increases in litterfall during the early dry season, followed by a flush of new leaves with increasing photosynthetic rates during the later dry season. The VPD‐AP version performs better than the SW‐AP version in capturing a dry‐season increase of photosynthesis across the wet Amazonia areas where mean annual precipitation exceeds 2,000 mm yr−1, consistent with previous satellite data analysis. Both VPD‐AP and SW‐AP model versions perform well in northern, central and southern Amazon regions where the SW seasonality is unimodal, but miss the seasonality of satellite GPP proxies in the eastern region off the coast of Guyana shield where SW seasonality is bimodal. Our findings imply that atmospheric dryness and sunlight availability likely explain the seasonality of leaf shedding and leaf flush processes, respectively, and consequently control canopy photosynthesis in Amazonian evergreen forests. Key Points: The new canopy phenology module captures the large‐scale spatial and seasonal patterns of Amazonian forest photosynthesisVapor pressure deficits and sunlight explain seasonality of Amazonian forest phenology and photosynthesisThis study for the first time integrates site‐level process knowledge to match diagnostic evidence at large scale across Amazonian forests [ABSTRACT FROM AUTHOR]

Water deficit in the atmosphere and soil are two key interactive factors that constrain transpiration and vegetation productivity. It is not clear which of these two factors is more important for the water and carbon flux response to drought stress in ecosystems. In this study, field data and numerical modeling were used to isolate their impact on evapotranspiration (ET) and gross primary productivity (GPP) at a tropical forest site in Barro Colorado Island (BCI), Panama, focusing on their response to the drought induced by the El Niño event of 2015–2016. Numerical simulations were performed using a plant hydrodynamic scheme (HYDRO) and a heuristic approach that ignores stomatal sensitivity to leaf water potential in the Energy Exascale Earth System Model (E3SM) Land Model (ELM). The sensitivity of canopy conductance (Gs) to vapor pressure deficit (VPD) obtained from eddy‐covariance fluxes and measured sap flux shows that, at both ecosystem and plant scale, soil water stress is more important in limiting Gs than VPD at BCI during the El Niño event. The model simulations confirmed the importance of water stress limitation on Gs, but overestimated the VPD impact on Gs compared to that estimated from the observations. We also found that the predicted soil moisture is less sensitive to the diversity of plant hydraulic traits than ET and GPP. During the dry season at BCI, seasonal ET, especially soil evaporation at VPD > 0.42 kPa, simulated using HYDRO and ELM, were too strong and will require alternative parameterizations. Plain Language Summary: Plants close stomata to regulate water loss through transpiration when stressed by the dry soil or high atmospheric water demand (vapor pressure deficit, VPD) or both. Tropical forests have experienced periodic droughts in the past. Recent drought‐related plant mortality has been attributed to increasing VPD associated with climate change. To evaluate whether water stress from dry soil or VPD has more limiting effect on plant transpiration, we analyzed the results from statistical models fitted to two types of field observation data and a land surface model which can simulate water movement in the soil and the water transport within the plant at a tropical forest site in Panama. We found dry soil is more important in limiting plant water loss at the site during the drought of the El Niño event of 2015–2016 and identified future model improvement need. Key Points: Canopy conductance (Gs) sensitivity to vapor pressure deficit (VPD) was analyzed from observations using multiple methodsEstimated Gs indicates stronger limitation by soil water stress than VPD at Barro Colorado Island during the 2015 El Niño eventWith a plant hydrodynamics scheme in Energy Exascale Earth System Model (E3SM) Land Model, variation of plant hydraulic traits has greater effects on Gs limitation than on soil moisture [ABSTRACT FROM AUTHOR]

We explore the potential of feed‐forward deep neural networks (DNNs) for emulating cloud superparameterization in realistic geography, using offline fits to data from the superparameterized community atmospheric model. To identify the network architecture of greatest skill, we formally optimize hyperparameters using ∼250 trials. Our DNN explains over 70% of the temporal variance at the 15‐min sampling scale throughout the mid‐to‐upper troposphere. Autocorrelation timescale analysis compared against DNN skill suggests the less good fit in the tropical, marine boundary layer is driven by neural network difficulty emulating fast, stochastic signals in convection. However, spectral analysis in the temporal domain indicates skillful emulation of signals on diurnal to synoptic scales. A closer look at the diurnal cycle reveals correct emulation of land‐sea contrasts and vertical structure in the heating and moistening fields, but some distortion of precipitation. Sensitivity tests targeting precipitation skill reveal complementary effects of adding positive constraints versus hyperparameter tuning, motivating the use of both in the future. A first attempt to force an offline land model with DNN emulated atmospheric fields produces reassuring results further supporting neural network emulation viability in real‐geography settings. Overall, the fit skill is competitive with recent attempts by sophisticated Residual and Convolutional Neural Network architectures trained on added information, including memory of past states. Our results confirm the parameterizability of superparameterized convection with continents through machine learning and we highlight the advantages of casting this problem locally in space and time for accurate emulation and hopefully quick implementation of hybrid climate models. Plain Language Summary: Machine learning methods have been previously used to replace parameterizations (approximations) of atmospheric convection under very idealized scenarios (aqua‐planets). The hope is that these machine learning emulators can help power the next generation of climate models with similar accuracy but at a fraction of the computational cost. But important questions remain about how learnable more realistic convection (over both land and ocean) is. Recently, the first attempt at machine learning replicated convection was made under these Earth‐like conditions. But it required a highly specialized neural network as well as memory of the previous behavior of the atmosphere. This design would make using these machine learning emulators with climate models very difficult. This motivates learning convection under realistic geography with a simpler network. Our results are reassuring because our simple neural network learns realistic convection over land as well as a more complicated model. But even harder tests involving full coupling with a host climate model will be needed to truly test this method's potential. Key Points: Feed‐forward neural networks can emulate explicit convection including geographic complexityAfter tuning, our neural network can fit 70% of the temporal variance in the mid‐to‐upper troposphereDeep convection over land is parameterizable in neural networks locally in time [ABSTRACT FROM AUTHOR]

Turbulence parameterizations for convective boundary layer in coarse‐scale atmospheric models usually consider a combination of the eddy‐diffusive transport and a non‐local transport, typically in the form of a mass flux term, such as the widely adopted eddy‐diffusivity mass‐flux (EDMF) approach. These two types of turbulent transport are generally considered to be independent of each other. Using results from large‐eddy simulations, here, we show that a Taylor series expansion of the updraft and downdraft mass‐flux transport can be used to approximate the eddy‐diffusivity transport in the atmospheric surface layer and the lower part of the mixed layer, connecting both eddy‐diffusivity and mass‐flux transport theories in convective conditions, which also quantifies departure from the Monin‐Obukhov similarity (MOS) in the surface layer. This study provides a theoretical support for a unified EDMF parameterization applied to both the surface layer and mixed layer and highlights important correction required for surface models relying on MOS. Plain Language Summary: The atmospheric boundary layer is a key layer connecting the atmosphere and the surface. This boundary layer during daytime can be split into the lower surface layer where large gradients in momentum and scalars occur, overlaid by the mixed layer. In the surface layer, the concept of eddy‐diffusivity transport is typically assumed and used, but it does not apply in the mixed layer, where non‐local transport is essential to maintain transport under vanishing vertical gradients. These two types of transport are generally considered to be independent of each other and are usually represented in coarse‐scale atmospheric models with a combination of the eddy‐diffusive transport and a non‐local transport, typically in the form of a mass flux term, such as the widely adopted eddy‐diffusivity mass‐flux (EDMF) approach. Here, we show that a Taylor series expansion of the mass‐flux transport can be used to approximate the eddy‐diffusivity transport in the surface layer and lower part of the mixed layer, connecting both eddy‐diffusivity and mass‐flux transport theories in convective conditions. This study provides a theoretical support for a unified EDMF parameterization applied to both the surface layer and mixed layer and highlights the need to account for non‐local transport for Monin‐Obukhov similarity‐based surface‐layer models. Key Points: The linearized mass‐flux transport contributes to eddy‐diffusivity in the surface and lower mixed layersThis contribution quantifies departure from the Monin‐Obukhov similarity scaling under increasing convective conditionsIt implies that a unified eddy‐diffusivity mass‐flux parameterization can be applied to both the surface and mixed layers [ABSTRACT FROM AUTHOR]

Global satellite precipitation estimation at high spatiotemporal resolutions is crucial for hydrological and meteorological applications but is still a challenging task. One major challenge is that the microwave data are discontinuous in space and time. We present a novel approach to merge incomplete passive microwave (PMW) precipitation estimates using the conditional information provided by complete infrared (IR) precipitation estimates based on the generative adversarial network (GAN), and name the algorithm PrecipGAN. PrecipGAN decomposes the precipitation system into content and evolution subspaces to propagate PMW estimates to regions outside the orbit coverage of PMW sensors. PrecipGAN can skillfully simulate the spatiotemporal changes of precipitation events, and produce precipitation estimates with overall better statistical performance than the baseline product Integrated MultisatellitE Retrievals for GPM (IMERG) Uncalibrated over the Continental US. PrecipGAN provides an alternative of accurate and computationally efficient algorithm that can be implemented globally to produce satellite‐based precipitation estimates. Plain Language Summary: Producing accurate precipitation estimates globally using satellite data at high spatial and temporal resolutions remains a major challenge for the remote sensing community. Uncertain precipitation products limit our capacity in various applications, such as hydrology, meteorology, or climatology. One difficult part in satellite precipitation estimation lies in combining infrared with passive microwave data effectively, as infrared data suffers from low detection accuracy but is superior in terms of spatiotemporal coverage while passive microwave data suffers from low spatiotemporal coverage but does excellent jobs at detecting precipitation accurately. Here, we present a novel deep learning model based on generative adversarial network to merge infrared and passive microwave data with a perspective based on the physical process of precipitation. The performance of the new algorithm is verified through comprehensive comparison with a state‐of‐the‐art precipitation product. The results show that this new algorithm has the potential to be a powerful alternative for generating global precipitation product merging infrared and microwave data. Key Points: A deep learning model called PrecipGAN is developed to merge passive microwave with infrared signals for seamless precipitation estimationPrecipGAN is designed based on the physical process of precipitation by decomposing it into content and evolution subspacesPrecipGAN can produce precipitation estimates with overall better statistical performance compared to an operational satellite product [ABSTRACT FROM AUTHOR]

A constellation of satellites is now in orbit providing information about terrestrial carbon and water storage and fluxes. These combined observations show that the tropical biosphere has changed significantly in the last 2 decades from the combined effects of climate variability and land use. Large areas of forest have been cleared in both wet and dry forests, increasing the source of carbon to the atmosphere. Concomitantly, tropical fire emissions have declined, at least until 2016, from changes in land‐use practices and rainfall, increasing the net carbon sink. Measurements of carbon stocks and fluxes from disturbance and recovery and of vegetation photosynthesis show significant regional variability of net biosphere exchange and gross primary productivity across the tropics and are tied to seasonal and interannual changes in water fluxes and storage. Comparison of satellite based estimates of evapotranspiration, photosynthesis, and the deuterium content of water vapor with patterns of total water storage and rainfall demonstrate the presence of vegetation‐atmosphere interactions and feedback mechanisms across tropical forests. However, these observations of stocks, fluxes and inferred interactions between them do not point unambiguously to either positive or negative feedbacks in carbon and water exchanges. These ambiguities highlight the need for assimilation of these new measurements with Earth System models for a consistent assessment of process interactions, along with focused field campaigns that integrate ground, aircraft and satellite measurements, to quantify the controlling carbon and water processes and their feedback mechanisms.Plain Language Summary: Changes to the carbon sequestered in tropical forests and soils, as a result of human activities and changes in rainfall, temperature, and CO2 concentrations, have a substantial impact on Earth's climate. This review summarizes recent results highlighting how the constellation of satellites now in orbit are providing new understanding of the tropical carbon cycle and how it interacts with climate variability through the water cycle, and how satellite data can be used to improve our process description of the Earth System.Key Points: Almost two decades are available of satellite measurements of multiple carbon and water tracers for quantifying tropical carbon and water states, fluxes, and interactionsSubstantial progress has been made in quantifying the role of water variability on carbon variability in the wet and dry tropicsSatellite data provide evidence of the role of vegetation on the water cycleNew measurements are needed of key states and fluxes controlling C/W interactions (e.g. soil moisture, evapotranspiration, soil carbon and respiration) to reduce uncertainty in the tropical carbon sink and its evolutionModels that can integrate this wealth of data are needed to infer the hidden processes controlling carbon and water [ABSTRACT FROM AUTHOR]

Atmospheric moisture holding capacity increases with temperature by about 7% per °C according to the Clausius‐Clapeyron relationship. Thermodynamically then, precipitation intensity should exponentially intensify and thus worsen flood conditions as the climate warms. However, regional and global analyses often report a nonmonotonic (hook) scaling of precipitation and runoff, in which extremes strengthen with rising temperature up to a maximum or peak point (Tpp) and decline thereafter. The underlying cause of this hook structure is not yet well‐understood, and whether it may shift and/or regulate storm runoff extremes under anthropogenic climate warming remains unknown. Here, we examine temperature scaling of precipitation and storm runoff extremes under different climate conditions using observations and large ensemble hydroclimatic simulations over mainland China. In situ observations suggest a spatially homogeneous, negative response of relative humidity to warming climates over 34.6% of the land area, and the remaining hook‐dominated regions usually show a colder Tpp than that of precipitation or storm runoff extremes. The precipitation and streamflow series over mainland China's catchments throughout the 21st century are projected by a model cascade chain under a high‐end emission scenario (RCP 8.5), which involves 31 CMIP5 climate models, 11 CMIP6 climate members, a daily bias correction method, and four lumped conceptual hydrological models. The CMIP5 ensemble projects that the hook structures shift toward warmer temperature bins, resulting in 10%–30% increases in storm runoff extremes over mainland China, while the CMIP6 ensemble projects more severe flood conditions in future warming climates. Plain Language Summary: China is threatened by river floods and has frequently suffered from significant ecosystem damage and economic loss under a changing environment. Storm runoff extremes, an important indicator of flooding regime, have received increasing attention, especially regarding their response to a warming climate. Previous studies have observed a nonmonotonic relationship (called a hook structure) between storm runoff extremes and atmospheric temperatures, in which the extremes strengthen with warming up to a peak point and decline thereafter in a hotter environment. Here, we investigate the ability of such behavior to constrain the intensification of future storm runoff extremes, as well as the underlying mechanisms. We find that the hook structures are not fixed under climate warming but peak at warmer temperatures and higher storm runoff rates. As a result, future flooding risks in mainland China will not be constrained by the hook structures but progressively increase by roughly 10%–30%. Key Points: The temperature scaling of precipitation, storm runoff, and relative humidity are characterized over mainland China's catchmentsPrecipitation and storm runoff usually exhibit a hook structure, with a warmer peak point temperature than that of relative humidityThe hook structures are not fixed under climate change but shift toward warmer temperatures and higher extreme rainfall and runoff rates [ABSTRACT FROM AUTHOR]

The terrestrial biosphere is currently slowing down global warming by absorbing about 30% of human emissions of carbon dioxide (CO2). The largest flux of the terrestrial carbon uptake is gross primary production (GPP) defined as the production of carbohydrates by photosynthesis. Elevated atmospheric CO2 concentration is expected to increase GPP ("CO2 fertilization effect"). However, Earth system models (ESMs) exhibit a large range in simulated GPP projections. In this study, we combine an existing emergent constraint on CO2 fertilization with a machine learning approach to constrain the spatial variations of multimodel GPP projections. In a first step, we use observed changes in the CO2 seasonal cycle at Cape Kumukahi to constrain the global mean GPP at the end of the 21st century (2091–2100) in Representative Concentration Pathway 8.5 simulations with ESMs participating in the Coupled Model Intercomparison Project Phase 5 (CMIP5) to 171 ± 12 Gt C yr−1, compared to the unconstrained model range of 156–247 Gt C yr−1. In a second step, we use a machine learning model to constrain gridded future absolute GPP and gridded fractional GPP change in two independent approaches. For this, observational data are fed into the machine learning algorithm that has been trained on CMIP5 data to learn relationships between present‐day physically relevant diagnostics and the target variable. In a leave‐one‐model‐out cross‐validation approach, the machine learning model shows superior performance to the CMIP5 ensemble mean. Our approach predicts an increased GPP change in northern high latitudes compared to regions closer to the equator. Plain Language Summary: About a quarter of human emissions of carbon dioxide (CO2) is absorbed by vegetation and soil on the Earth's surface and hence does not contribute to global warming caused by CO2 in the atmosphere. Thus, in order to better define the remaining carbon budgets left to meet particular warming targets like the 1.5°C of the Paris Agreement, it is important to accurately quantify the carbon uptake by plants in the future. Currently, this is modeled by Earth system models yet with great uncertainties. In this work, we present an alternative machine learning approach to reduce spatial uncertainties of vegetation carbon uptake in future climate projections using observations of today's conditions. Key Points: An emergent constraint on CO2 seasonal cycle amplitude changes reduces uncertainties in global mean gross primary production projectionsA machine learning model with multiple predictors can further constrain the spatial distribution of gross primary productionHigh‐latitude ecosystems show higher gross primary production increase over the 21st century compared to regions closer to the equator [ABSTRACT FROM AUTHOR]

Mesoscale convection generates the majority of extreme precipitation in tropical regions. Changes to these precipitation intensities, P, with long‐term modes of climate variability have been hard to assess because they are not well represented in current climate models. Here we stratify a satellite climatology of convective systems by El Niño phase and cloud top temperature. We find that gains (losses) in high precipitation intensity (P˙> 10 mm hr−1) are largest for the deepest (least deep) systems during El Niño relative to La Niña. The surface temperature and wind changes that define El Niño manifest as surface flux changes but are not sufficient to explain these P˙ trends. We explore also the dynamical component of precipitation generation with a vertical momentum budget. Midtropospheric drying in the vicinity of the deepest systems boosts instability and ascent rates during El Niño, while the strengthened large‐scale ascent minimizes the drag force on their updrafts. Plain Language Summary: In the tropics, the majority of high‐intensity precipitation comes from the organization of multiple thunderstorms into a convective system. These systems are not yet well represented in global models, and studies of how their precipitation changes with long‐term modes of climate variability like the El Niño–Southern Oscillation have been limited. Using long‐term satellite data, we find here that the most intense precipitation becomes two times more probable for the deepest systems and 20% less probable for the least deep systems during El Niño. With a budget for the ascent rate within these systems, we illustrate how favorable moisture structure acts as a buoyancy source in the first case and unfavorable circulation acts as a buoyancy sink in the latter case. Key Points: For the deepest convective systems, precipitation rates >10 mm hr−1 become two times more likely during El NiñoThe distribution of net gross moist stability shifts to lower values for these deepest systems during El NiñoDecreased net gross moist stability can be explained by strengthened large‐scale ascent and free tropospheric drying [ABSTRACT FROM AUTHOR]

Stratification can cause turbulence spectra to deviate from Kolmogorov's isotropic −5/3 power law scaling in the universal equilibrium range at high Reynolds numbers. However, a consensus has not been reached with regard to the exact shape of the spectra. Here we propose a shape of the turbulent kinetic energy and temperature spectra in horizontal wavenumber for the equilibrium range that consists of three regimes at small Froude number: the buoyancy subrange, a transition region, and the isotropic inertial subrange through dimensional analysis and substantial revision of previous theoretical approximation. These spectral regimes are confirmed by various observations in the atmospheric boundary layer. The representation of the transition region in direct numerical simulations will require large‐scale separation between the Dougherty‐Ozmidov scale and the Kolmogorov scale for strongly stratified turbulence at high Reynolds numbers, which is still challenging computationally. In addition, we suggest that the failure of Monin‐Obukhov similarity theory in the very stable atmospheric boundary layer is due to the fact that it does not consider the buoyancy scale that characterizes the transition region. Key Points: Turbulent kinetic energy and temperature spectra in horizontal wavenumber consist of three regions in the stable ABLLarge‐scale separation is required for numerical simulations to resolve the transition regionMonin‐Obukhov similarity theory may not apply in the very stable ABL as the buoyancy scale is not considered [ABSTRACT FROM AUTHOR]

Dry deposition of ozone is an important sink of ozone in near‐surface air. When dry deposition occurs through plant stomata, ozone can injure the plant, altering water and carbon cycling and reducing crop yields. Quantifying both stomatal and nonstomatal uptake accurately is relevant for understanding ozone's impact on human health as an air pollutant and on climate as a potent short‐lived greenhouse gas and primary control on the removal of several reactive greenhouse gases and air pollutants. Robust ozone dry deposition estimates require knowledge of the relative importance of individual deposition pathways, but spatiotemporal variability in nonstomatal deposition is poorly understood. Here we integrate understanding of ozone deposition processes by synthesizing research from fields such as atmospheric chemistry, ecology, and meteorology. We critically review methods for measurements and modeling, highlighting the empiricism that underpins modeling and thus the interpretation of observations. Our unprecedented synthesis of knowledge on deposition pathways, particularly soil and leaf cuticles, reveals process understanding not yet included in widely used models. If coordinated with short‐term field intensives, laboratory studies, and mechanistic modeling, measurements from a few long‐term sites would bridge the molecular to ecosystem scales necessary to establish the relative importance of individual deposition pathways and the extent to which they vary in space and time. Our recommended approaches seek to close knowledge gaps that currently limit quantifying the impact of ozone dry deposition on air quality, ecosystems, and climate.Plain Language Summary: The removal of tropospheric ozone at Earth's surface (often called dry deposition) is important for our understanding of air pollution, ecosystem health, and climate. Several processes contribute to dry deposition of ozone. While we have basic knowledge of these processes, we lack the ability to robustly estimate changes in ozone dry deposition through time and from one place to another. Here we review ozone deposition processes, measurements, and modeling and propose steps necessary to close gaps in understanding. A major conclusion revealed by our review is that most deposition processes can be fairly well described from a theoretical standpoint, but the relative importance of the various processes remains uncertain. We suggest that progress can be made by establishing multiyear measurements of ozone dry deposition at a limited set of sites around the world and coordinating these measurements with laboratory and field experiments that can be integrated with theory through carefully designed modeling studies.Key Points: Ozone dry deposition through pathways other than plant stomata is critical for describing the total terrestrial ozone sinkProcess‐level knowledge of ozone deposition pathways is missing from the models used to quantify deposition impacts on the Earth systemLong‐term ozone flux and related measurements are key for establishing relative importance of individual pathways [ABSTRACT FROM AUTHOR]

Estimating ecosystem evapotranspiration (ET) is important to understanding the global water cycle and to study land‐atmosphere interactions. We developed a physics constrained machine learning (ML) model (hybrid model) to estimate latent heat flux (LE), which conserves the surface energy budget. By comparing model predictions with observations at 82 eddy covariance tower sites, our hybrid model shows similar performance to the pure ML model in terms of mean metrics (e.g., mean absolute percent errors) but, importantly, the hybrid model conserves the surface energy balance, while the pure ML model does not. A second key result is that the hybrid model extrapolates much better than the pure ML model, emphasizing the benefits of combining physics with ML for increased generalizations. The hybrid model allows inferring the structural dependence of ET and surface resistance (rs), and we find that vegetation height and soil moisture are the main regulators of ET and rs. Plain Language Summary: A physics constrained machine learning model is developed using the FLUXNET2015 Tier 1 data set. This new approach is able to effectively retrieve latent heat flux while constraining energy conservation in the surface energy budget. This hybrid model has better performance in extrapolation than a pure machine learning model. Key Points: A physics‐constrained machine learning model of evapotranspiration (hybrid model) is developed and trained using the FLUXNET 2015 data setThe evapotranspiration retrieved by the hybrid model is as accurate as pure machine learning model and also conserves surface energy balanceThe hybrid model better reproduces extremes and thus better extrapolates compared to the pure machine learning approach [ABSTRACT FROM AUTHOR]

Increasing vapor pressure deficit (VPD) increases atmospheric demand for water. While increased evapotranspiration (ET) in response to increased atmospheric demand seems intuitive, plants are capable of reducing ET in response to increased VPD by closing their stomata. We examine which effect dominates the response to increasing VPD: atmospheric demand and increases in ET or plant response (stomata closure) and decreases in ET. We use Penman-Monteith, combined with semiempirical optimal stomatal regulation theory and underlying water use efficiency, to develop a theoretical framework for assessing ET response to VPD. The theory suggests that depending on the environment and plant characteristics, ET response to increasing VPD can vary from strongly decreasing to increasing, highlighting the diversity of plant water regulation strategies. The ET response varies due to (1) climate, with tropical and temperate climates more likely to exhibit a positive ET response to increasing VPD than boreal and arctic climates; (2) photosynthesis strategy, with C3 plants more likely to exhibit a positive ET response than C4 plants; and (3) plant type, with crops more likely to exhibit a positive ET response, and shrubs and gymniosperm trees more likely to exhibit a negative ET response. These results, derived from previous literature connecting plant parameters to plant and climate characteristics, highlight the utility of our simplified framework for understanding complex land-atmosphere systems in terms of idealized scenarios in which ET responds to VPD only. This response is otherwise challenging to assess in an environment where many processes coevolve together. [ABSTRACT FROM AUTHOR]

Convective organization has a large impact on precipitation and feeds back on larger‐scale circulations in the tropics. The degree of this convective organization changes with modes of climate variability like the El Niño–Southern Oscillation (ENSO), but because organization is not represented in current climate models, a quantitative assessment of these shifts has not been possible. Here, we construct multidecade satellite climatologies of occurrence of tropical convective organization and its properties and assess changes with ENSO phase. The occurrence of organized deep convection becomes more concentrated, increasing threefold in the eastern and central Pacific during El Niño and decreasing twofold outside of these regions. Both horizontal extent of the cold cloud shield and convective depth increase in regions of positive sea surface temperature anomaly (SSTa); however, the regions of greatest convective deepening are those of large‐scale ascent, rather than those of warmest SSTa. Extent decreases with SSTa at a rate of about 20 km/K, while the SSTa dependence of depth is only about 0.2 K/K. We introduce two values to describe convective changes with ENSO more succinctly: (1) an information entropy metric to quantify the clustering of convective system occurrences and (2) a growth metric to quantify deepening relative to spreading over the system lifetime. Finally, with collocated precipitation data, we see that rainfall attributable to convective organization jumps up to 5% with warming. Rain intensity and amount increase for a given system size during El Niño, but a given rain amount may actually fall with higher intensity during La Niña. Plain Language Summary: Tropical storms at an intermediate scale between individual thunderstorms and cyclones or hurricanes are not represented by climate models. As the satellite data record now spans several decades, we can use these observations to understand how large storms change during El Niño periods. Here, we find that warmer waters in the central and eastern Pacific during El Niño cause a dramatic increase in storm occurrence. Storms are more likely to occur near to other storms and to become larger in both the vertical and horizontal. Changes to large‐scale winds are more influential on these structural factors than locally warmer ocean water. The hydrological impact of these storms is also important because they bring the majority of precipitation in some regions. We see that a storm of a certain extent brings larger volume and intensity of rain during El Niño. But a certain rain volume may actually fall with greater intensity during La Niño. Key Points: Convective organization becomes more frequent throughout the tropics, and especially over warmer ocean waters, during El NiñoThe vertical depth, horizontal extent, and attributable precipitation of tropical organized convection increase during El NiñoWe present new metrics that quantify the relative importance of convective system deepening to spreading and their degree of organization [ABSTRACT FROM AUTHOR]

Summary: Reducing uncertainties in the response of tropical forests to global change requires understanding how intra‐ and interannual climatic variability selects for different species, community functional composition and ecosystem functioning, so that the response to climatic events of differing frequency and severity can be predicted.Here we present an extensive dataset of hydraulic traits of dominant species in two tropical Amazon forests with contrasting precipitation regimes – low seasonality forest (LSF) and high seasonality forest (HSF) – and relate them to community and ecosystem response to the El Niño–Southern Oscillation (ENSO) of 2015.Hydraulic traits indicated higher drought tolerance in the HSF than in the LSF. Despite more intense drought and lower plant water potentials in HSF during the 2015‐ENSO, greater xylem embolism resistance maintained similar hydraulic safety margin as in LSF. This likely explains how ecosystem‐scale whole‐forest canopy conductance at HSF maintained a similar response to atmospheric drought as at LSF, despite their water transport systems operating at different water potentials.Our results indicate that contrasting precipitation regimes (at seasonal and interannual time scales) select for assemblies of hydraulic traits and taxa at the community level, which may have a significant role in modulating forest drought response at ecosystem scales. [ABSTRACT FROM AUTHOR]

Evapotranspiration (ET) plays a central role in the water, energy, and carbon cycles but is difficult to model and estimate due to its dependence on the heterogeneous land surface. Recent studies suggest that standard weather station atmospheric observations alone may be sufficient to estimate ET. This is surprising since ET over land is often strongly constrained by the land surface, for instance, by water limitation. While these studies have been empirically successful, a physical explanation for why this is possible has been lacking. Here, we provide a physical explanation for why one of these approaches—the ET from Relative Humidity at Equilibrium (ETRHEQ) method—works, using a simple model of a steady‐state idealized atmospheric boundary layer. We show that, across a wide range of plausible parameter values, this model reproduces ETRHEQ, suggesting that it contains the essential physics that lead to ETRHEQ. We derive a closed‐form expression for ETRHEQ at steady state and use it to show that ETRHEQ can be explained in terms of the near‐surface relative humidity (RH) budget of the idealized model: in particular, it is equivalent to assuming a balance between surface moistening and heating terms in the RH budget. Negative feedbacks between surface fluxes (constrained by the surface energy budget) and atmospheric temperature and humidity mean that these terms typically balance, explaining the empirical success of ETRHEQ over a wide range of conditions. We define this state—in which the moistening and heating terms balance in the RH budget—as "surface flux equilibrium." Key Points: Empirical work suggests that ET can be estimated from atmospheric observations alone, even under water‐limited conditionsWe develop an idealized equilibrium ET‐ABL model that couples the surface and atmospheric statesThe model reproduces and explains the observed phenomenon [ABSTRACT FROM AUTHOR]

Over the past decade, the concept of isohydry or anisohydry, which describes the link between soil water potential (ΨS), leaf water potential (ΨL), and stomatal conductance (gs), has soared in popularity. However, its utility has recently been questioned, and a surprising lack of coordination between the dynamics of ΨL and gs across biomes has been reported. Here, we offer a more expanded view of the isohydricity concept that considers effects of vapour pressure deficit (VPD) and leaf area index (AL) on the apparent sensitivities of ΨL and gs to drought. After validating the model with tree‐ and ecosystem‐scale data, we find that within a site, isohydricity is a strong predictor of limitations to stomatal function, though variation in VPD and leaf area, among other factors, can challenge its diagnosis. Across sites, the theory predicts that the degree of isohydricity is a good predictor of the sensitivity of gs to declining soil water in the absence of confounding effects from other drivers. However, if VPD effects are significant, they alone are sufficient to decouple the dynamics of ΨL and gs entirely. We conclude with a set of practical recommendations for future applications of the isohydricity framework within and across sites. We expand the mathematical basis of the popular but controversial isohydric–anisohydric framework for plant drought response to include variations in atmospheric aridity and drought deciduousness. We demonstrate that aridity effects are critical to consider when applying the framework across sites. [ABSTRACT FROM AUTHOR]

Extreme temperatures are responsible for damages to society and ecosystems. There is evidence that severe episodes of extreme heat have been occurring more frequently and more severely in recent periods. Driven primarily by oceanic and atmospheric effects as well as land‐climate feedbacks, those extreme events are expected to increase with climate change. Vegetation, which regulates the energy, water, and carbon cycles, is a key player of land‐atmosphere interactions that has been proven to be determinant in recent extreme events. Using an ensemble of Earth System Models simulations, we show that physiological effects globally increase the annual daily maximum temperature (Txx) with rising [CO2], accounting globally for around 13% of the full Txx trend. Due to physiological effects, Txx can reinforce (e.g., central Europe) or reduce (e.g., central North America) the mean temperature increase. Plain Language Summary: Heat waves cause damages to society and natural environment. We observed a trend of increasing frequency and intensity of extreme temperature episodes in recent years. Climate change on one side and the complex feedbacks of the climate with the land on the other side are the main drivers of this trend. Vegetation is an important factor of the feedback as it connects the land surface and the atmosphere, regulating the exchanges of water and heat. Using an ensemble of Earth System Models simulations, we show that the effects of vegetation globally increase extreme temperatures with rising [CO2], accounting globally for around 13% of the full trend. This effect can, however, regionally be larger (e.g., central Europe) or smaller (e.g., central North America) than the mean temperature increase. Key Points: Physiological effects increase the annual daily maximum temperature (Txx) with rising [CO2], accounting globally for around 15% of its trendThe Txx increasing trends can be reinforced by vegetation physiological feedbacks, especially in regions like Europe or the AmazonIn several regions, such as North America, vegetation physiological feedbacks can oppositely mitigate the radiatively driven increase in Txx [ABSTRACT FROM AUTHOR]

Version 5 of the Community Land Model (CLM5) introduces the plant hydraulic stress (PHS) configuration of vegetation water use, which is described and compared with the corresponding parameterization from CLM4.5. PHS updates vegetation water stress and root water uptake to better reflect plant hydraulic theory, advancing the physical basis of the model. The new configuration introduces prognostic vegetation water potential, modeled at the root, stem, and leaf levels. Leaf water potential replaces soil potential as the basis for stomatal conductance water stress, and root water potential is used to implement hydraulic root water uptake, replacing a transpiration partitioning function. Point simulations of a tropical forest site (Caxiuanã, Brazil) under ambient conditions and partial precipitation exclusion highlight the differences between PHS and the previous CLM implementation. Model description and simulation results are contextualized with a list of benefits and limitations of the new model formulation, including hypotheses that were not testable in previous versions of the model. Key results include reductions in transpiration and soil moisture biases relative to a control model under both ambient and exclusion conditions, correcting excessive dry season soil moisture stress in the control model. PHS implements hydraulic gradient root water uptake, which allows hydraulic redistribution and compensatory root water uptake and results in PHS utilizing a larger portion of the soil column to buffer shortfalls in precipitation. The new model structure, which bases water stress on leaf water potential, could have significant implications for vegetation‐climate feedbacks, including increased sensitivity of photosynthesis to atmospheric vapor pressure deficit. Key Points: An updated soil‐plant‐atmosphere continuum model based on hydraulic theory is implemented in the Community Land Model (version 5)Prognostic leaf water potential replaces soil matric potential as the basis for stomatal conductance water stressPrognostic root water potential is used to implement hydraulic root water uptake, replacing a "soil wilting point" approach [ABSTRACT FROM AUTHOR]

Droughts and heatwaves cause agricultural loss, forest mortality, and drinking water scarcity, especially when they occur simultaneously as combined events. Their predicted increase in recurrence and intensity poses serious threats to future food security. Still today, the knowledge of how droughts and heatwaves start and evolve remains limited, and so does our understanding of how climate change may affect them. Droughts and heatwaves have been suggested to intensify and propagate via land–atmosphere feedbacks. However, a global capacity to observe these processes is still lacking, and climate and forecast models are immature when it comes to representing the influences of land on temperature and rainfall. Key open questions remain in our goal to uncover the real importance of these feedbacks: What is the impact of the extreme meteorological conditions on ecosystem evaporation? How do these anomalies regulate the atmospheric boundary layer state (event self‐intensification) and contribute to the inflow of heat and moisture to other regions (event self‐propagation)? Can this knowledge on the role of land feedbacks, when available, be exploited to develop geo‐engineering mitigation strategies that prevent these events from aggravating during their early stages? The goal of our perspective is not to present a convincing answer to these questions, but to assess the scientific progress to date, while highlighting new and innovative avenues to keep advancing our understanding in the future. Droughts and heatwaves cause agricultural loss, forest mortality and drinking water scarcity, especially when they occur simultaneously as combined events. Their predicted increase in recurrence and intensity poses serious threats to future food security. Still today, the knowledge of how droughts and heatwaves start and evolve remains limited, and so does our understanding of how climate change may affect them. Our goal is to revisit the role of land feedbacks on the evolution of these extreme events, and highlight new and innovative avenues to keep advancing our understanding how droughts and heatwaves develop and propagate. [ABSTRACT FROM AUTHOR]

Near‐surface turbulent flux and radiation divergence field observations are analysed over a grass‐covered surface located at the Wageningen observatory, the Netherlands. Net radiative flux divergence appears to be a large component of the energy budget near the surface, accounting for a cooling rate of several tens of degrees per day. Long‐wave radiation divergence dominates this radiation divergence flux. We show here that long‐wave flux divergence near the surface is strongly coupled to sensible heat flux, except in high relative humidity (>90%) and foggy conditions. The net long‐wave radiative flux divergence exhibits a sharp gradient within the first 20 m above the surface. This flux divergence is itself strongly coupled to sensible heat flux through adjustments in surface‐layer profiles. Nonetheless, no systematic effect of radiation is witnessed on the turbulent temperature spectrum so that the main effect of near‐surface radiation is on the mean heat budget. As typical flux‐gradient relationships are derived based on observations taken in the near‐surface boundary‐layer region where sharp long‐wave divergence is present, we suspect that those relationships implicitly represent some of the long‐wave divergence term. A modification of Monin–Obukhov Similarity Theory to include the effect of radiative divergence is proposed and discussed. This calls for independent measurements of turbulent fluxes and radiative flux divergence near the surface to re‐derive turbulent flux‐gradient relationships independently of radiative effects. Schematics showing the position of the different. instruments [ABSTRACT FROM AUTHOR]

Abstract: Cloud shading limits surface radiation, thus reducing vegetation water stress and, presumably, flammability. Since the early 1970s, cloud observations from airfields in coastal Southern California (CSCA) indicate reductions of ~25–50% in warm‐season frequency of daytime stratus clouds at many sites, including fire‐prone wildland‐urban interface zones. We use 10 years of meteorological, surface radiation, and cloud observations to statistically model the effects of clouds on warm‐season surface energy fluxes in CSCA. Forcing our model with cloud observations, we estimate that reduced warm‐season cloud shading since the 1970s significantly enhanced daytime solar radiation and evaporative demand throughout much of CSCA, particularly in greater Los Angeles and northern San Diego. Correlation with burned area and live fuel moisture implicates stratus cloud shading as an important driver of warm‐season wildfire activity in CSCA. Large reductions in cloud shading have likely enhanced warm‐season wildfire potential in many CSCA areas when and where fuels are not limiting. [ABSTRACT FROM AUTHOR]

Light‐use efficiency (LUE), which quantifies the plants' efficiency in utilizing solar radiation for photosynthetic carbon fixation, is an important factor for gross primary production estimation. Here we use satellite‐based solar‐induced chlorophyll fluorescence as a proxy for photosynthetically active radiation absorbed by chlorophyll (APARchl) and derive an estimation of the fraction of APARchl (fPARchl) from four remotely sensed vegetation indicators. By comparing maximum LUE estimated at different scales from 127 eddy flux sites, we found that the maximum daily LUE based on PAR absorption by canopy chlorophyll (εmaxchl ), unlike other expressions of LUE, tends to converge across biome types. The photosynthetic seasonality in tropical forests can also be tracked by the change of fPARchl, suggesting the corresponding εmaxchl to have less seasonal variation. This spatio‐temporal convergence of LUE derived from fPARchl can be used to build simple but robust gross primary production models and to better constrain process‐based models. [ABSTRACT FROM AUTHOR]

Abstract: In the southwestern USA, recent large‐scale die‐offs of conifers raise the question of their resilience and mortality under droughts. To date, little is known about the interannual structural response to droughts. We hypothesized that piñon pines (Pinus edulis) respond to drought by reducing the drop of leaf water potential in branches from year to year through needle morphological adjustments. We tested our hypothesis using a 7‐year experiment in central New Mexico with three watering treatments (irrigated, normal, and rain exclusion). We analyzed how variation in “evaporative structure” (needle length, stomatal diameter, stomatal density, stomatal conductance) responded to watering treatment and interannual climate variability. We further analyzed annual functional adjustments by comparing yearly addition of needle area (LA) with yearly addition of sapwood area (SA) and distance to tip (d), defining the yearly ratios SA:LA and SA:LA/d. Needle length (l) increased with increasing winter and monsoon water supply, and showed more interannual variability when the soil was drier. Stomatal density increased with dryness, while stomatal diameter was reduced. As a result, anatomical maximal stomatal conductance was relatively invariant across treatments. SA:LA and SA:LA/d showed significant differences across treatments and contrary to our expectation were lower with reduced water input. Within average precipitation ranges, the response of these ratios to soil moisture was similar across treatments. However, when extreme soil drought was combined with high VPD, needle length, SA:LA and SA:LA/d became highly nonlinear, emphasizing the existence of a response threshold of combined high VPD and dry soil conditions. In new branch tissues, the response of annual functional ratios to water stress was immediate (same year) and does not attempt to reduce the drop of water potential. We suggest that unfavorable evaporative structural response to drought is compensated by dynamic stomatal control to maximize photosynthesis rates. [ABSTRACT FROM AUTHOR]

Abstract: The concept of isohydry/anisohydry describes the degree to which plants regulate their water status, operating from isohydric with strict regulation to anisohydric with less regulation. Though some species level measures of isohydry/anisohydry exist at a few locations, ecosystem‐scale information is still largely unavailable. In this study, we use diurnal observations from active (Ku‐Band backscatter from QuikSCAT) and passive (X‐band vegetation optical depth (VOD) from Advanced Microwave Scanning Radiometer on EOS Aqua) microwave satellite data to estimate global ecosystem isohydry/anisohydry. Here diurnal observations from both satellites approximate predawn and midday plant canopy water contents, which are used to estimate isohydry/anisohydry. The two independent estimates from radar backscatter and VOD show reasonable agreement at low and middle latitudes but diverge at high latitudes. Grasslands, croplands, wetlands, and open shrublands are more anisohydric, whereas evergreen broadleaf and deciduous broadleaf forests are more isohydric. The direct validation with upscaled in situ species isohydry/anisohydry estimates indicates that the VOD‐based estimates have much better agreement than the backscatter‐based estimates. The indirect validation with prior knowledge suggests that both estimates are generally consistent in that vegetation water status of anisohydric ecosystems more closely tracks environmental fluctuations of water availability and demand than their isohydric counterparts. However, uncertainties still exist in the isohydry/anisohydry estimate, primarily arising from the remote sensing data and, to a lesser extent, from the methodology. The comprehensive assessment in this study can help us better understand the robustness, limitation, and uncertainties of the satellite‐derived isohydry/anisohydry estimates. The ecosystem isohydry/anisohydry has the potential to reveal new insights into spatiotemporal ecosystem response to droughts. [ABSTRACT FROM AUTHOR]

Elevated atmospheric CO2 concentrations are expected to enhance photosynthesis and reduce stomatal conductance, thus increasing plant water use efficiency. A recent study based on eddy covariance flux observations from Northern Hemisphere forests showed a large increase in inherent water use efficiency (IWUE). Here we used an updated version of the same data set and robust uncertainty quantification to revisit these contemporary IWUE trends. We tested the hypothesis that the observed IWUE increase could be attributed to interannual trends in plant functional traits, potentially triggered by environmental change. We found that IWUE increased by ~1.3% yr−1, which is less than previously reported but still larger than theoretical expectations. Numerical simulations with the Tethys-Chloris ecosystem model using temporally static plant functional traits cannot explain this increase. Simulations with plant functional trait plasticity, i.e., temporal changes in model parameters such as specific leaf area and maximum Rubisco capacity, match the observed trends in IWUE. Our results show that trends in plant functional traits, equal to 1.0% yr−1, can explain the observed IWUE trends. Thus, at decadal or longer time scales, trait plasticity could potentially influence forest water, carbon, and energy fluxes with profound implications for both the monitoring of temporal changes in plant functional traits and their representation in Earth system models. [ABSTRACT FROM AUTHOR]

While the breakdown in similarity between turbulent transport of heat and momentum (or Reynolds analogy) is not disputed in the atmospheric surface layer (ASL) under unstably stratified conditions, the causes of this breakdown are still debated. One reason for the breakdown is differences between how coherent structures transport heat and momentum, and their differing responses to increasing instability. Monin-Obukhov Similarity Theory (MOST), which hypothesizes that only local length-scales play a role in ASL turbulent transport, implicitly assumes that large-scale structures are inactive, despite their large energy content. Widely adopted mixing-length models also rest on this assumption in the ASL. The difficulty of characterizing low-wavenumber turbulent motions with field measurements motivates the use of high-resolution Direct Numerical Simulation (DNS), which is free from subgrid-scale parametrizations and a d h o c assumptions near the boundary. Despite the low Reynolds number and idealized geometry of the DNS, DNS-estimated MOST functions are consistent with ASL field experiments, as are low-frequency features of the spectra. Parsimonious spectral models for MO stability correction functions for momentum ( φ m) and heat ( φ h) are derived, based on idealized vertical velocity variance and buoyancy variance spectra fit to the corresponding DNS spectra. For φ m, a spectral model, based only on local length-scales, matches DNS and field measurements well. In contrast, for φ h, the model is substantially biased unless contributions from larger length-scales are also included. These results are supported by sensitivity analyses based on field measurements that are independent of the DNS. They show that ASL heat transport is not MO-similar, even under mild stratification, and in the absence of entrainment, non-stationarity and canopy effects. It further suggests that the breakdown of the Reynolds analogy is at least partially caused by the influence of large eddies on turbulent heat transport. [ABSTRACT FROM AUTHOR]