Vegetation provides a critical nexus between the subsurface, biosphere, and atmosphere through the mediation of the exchange of water, carbon, and energy. Plants respond dynamically to local microclimates at both short and long timescales via mechanisms ranging from physiological behaviors, such as stomatal closure, to acclimation and adaptation. These responses influence land-atmosphere fluxes directly and are therefore crucial to understanding and predicting Earth system responses to a changing climate. As our community progresses towards increasingly physically-realistic models of vegetation responses to the environment, we face several new challenges such as understanding how whole-plant hydraulic strategies can best be represented using limited parameters, connecting non-linearly related observations such as water content and water potential within different organs, and representing responses to simultaneous stressor such as high temperatures and low water supply or high salinities and high evaporative demands. We use vignettes from two long-term tree and ecosystem level studies to demonstrate both progress and pitfalls towards overcoming each of these hurdles in terms of observational understanding of plant function and individual tree-based simulations of new observational data and the accompanying uncertainties. As we progress towards further incorporation of such vegetation ecohydrology modules within climate and weather models, it is becoming increasingly critical to represent well, and with limited parameters the manners in which plants manifest stress responses across time scales in order to better predict the complex feedbacks to carbon, water, and energy fluxes that subsequently develop.

How to cite: Matheny, A. M., Restrepo Acevedo, A. M., Ulatowski, M., Cabraal, S., and Missik, J.: Sensing and modeling plant ecohydrology for understanding tree and ecosystem responses to water stress, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11776, https://doi.org/10.5194/egusphere-egu24-11776, 2024.

The impact of plants on runoff under high atmospheric CO₂ is a major uncertainty for the future of global water resources. An emerging consensus based on theory and Earth System Models (ESMs) suggests that stricter plant stomatal regulation under high CO₂ will reduce transpiration, potentially boosting runoff. Yet, across a 12-member ensemble of idealized ESM simulations that isolate plant responses to CO₂, we show that lower transpiration robustly enhances runoff over only 5% of global land area. Instead, we find that precipitation changes are five times more important than transpiration changes in driving runoff responses when only plants respond to CO₂, and are a significant signal of CO₂ physiological forcing over31-57% of land areas across models. Crucially, the models largely disagree on where physiologically forced precipitation changes occur, but agree that plant responses in most locations are as likely to reduce runoff as increase it, absent any effects from radiative warming. These results imply that large model uncertainties in precipitation responses, rather than transpiration responses, explain why ESMs disagree on plant physiologically driven runoff changes over most of the globe. Together, our findings implicate land-atmosphere rather than land-hydrologic responses as the key mechanistic source of uncertainty in runoff responses under CO₂ physiological forcing. They further emphasize that any interpretation of plant-driven runoff responses must consider how precipitation itself will respond to CO₂ physiological forcing.

How to cite: Lesk, C., Winter, J., and Mankin, J.: Projected runoff declines from plant physiological effects on precipitation, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6128, https://doi.org/10.5194/egusphere-egu24-6128, 2024.

Plant water potential is a dynamic and fundamental driver of carbon and water fluxes, yet observations over time remain sparse. In a Sonoran Desert grassland, we utilized a precipitation manipulation experiment to a) evaluate the temporal trajectory of plant water potential following different precipitation pulse sizes and b) relate plant water potential to remote sensing proxies at the leaf scale. Beginning in 2020, natural summer rainfall was excluded and replaced with consistent irrigation divided among three watering treatments, which received identical seasonal totals delivered in return intervals of 3.5, 7, and 21 days (P3.5, P7, and P21, respectively), with correspondingly varied event magnitudes, between July and September. In 2023, we measured predawn and midday leaf water potential (Ψ_PD and Ψ_MD) as well as leaf-level hyperspectral on Digitaria californica, a native perennial bunchgrass, characterizing pulse events on Aug 14 (all treatments) and Aug 21 (second pulses for P3.5 & P7 only). Two spectra per leaf were measured, corrected for known breakpoints, and averaged prior to calculating NDVI, NDWI, and PRI.

Prior to the Aug 14 pulse irrigation, Ψ_PD was above -2 MPa in P7 while P3.5 and P21 both exhibited Ψ_PD around -2.5 MPa. While Ψ_PD peaked on day 1 following irrigation in all treatments, the amount of time spent in the well-watered range differed greatly. Ψ_PD dropped after day 1 in P3.5, after day 2 in P7, and after day 12 in P21, consistent with the varying pulse magnitudes. Uniquely in P21, pulse irrigation increased soil water content at 25 cm, indicating the availability of deeper soil moisture to D. californica with fewer/larger precipitation events. When comparing the replicated pulse events in the frequent/smaller treatments, the water potential response to Aug 14 and Aug 21 pulses differed greatly in P3.5 but not in P7. While soil water contents were similar across pulses, Ψ_PD in P3.5 started above -1 MPa during the Aug 21 pulse and did not exhibit a peaked response, coinciding with lower cumulative VPD. Reduced atmospheric demand may significantly moderate water potential responses to small precipitation events. Finally, across treatments and time-of-day, leaf water potential was most closely correlated with greenness indices of NDVI and PRI (R² = 0.166 and 0.183, respectively), while only loosely correlated with the water content index NDWI (R² = 0.018). Next, we intend to develop our own indices that can better capture the temporal response of leaf water potential to precipitation dynamics. 

How to cite: Guo, J., Smith, W., Scott, R., and Biederman, J.: Water potential dynamics in a precipitation pulse experiment: comparing direct and remote sensing metrics at the leaf scale, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10498, https://doi.org/10.5194/egusphere-egu24-10498, 2024.

In recent decades, the frequency and intensity of hotter droughts have increased, posing a serious threat to our forests. During hot droughts, increasing evapotranspiration depletes soil moisture reserves and thus exposes the trees’ xylem to critical water potentials. Several of Central Europe’s major timber species have been found to be especially susceptible to repeated summer droughts. Therefore, the forestry sector is increasingly considering the establishment of mixed stands and the inclusion of putatively more drought-resistant non-native tree species. However, silvicultural decisions about increasing the cultivation of non-native species and planting them in mixture requires empirical data on species-specific water consumption in pure and mixed culture in order to assess climate risks and to avoid potential negative competition effects.

To address these questions, we installed 32 dual-method-approach type sap-flow sensors capable of measuring the entire range of sap flow rates in pure European beech and Douglas fir stands as well as in a nearby mixed beech-Douglas fir stand on deep sandy soil in northern Germany. Additionally, heat-field-deformation type sap-flow sensors were used for measuring the radial sap flow profile in the xylem of each individual tree. The trees equipped with sensors covered a broad DBH range which allowed extrapolating to stand-level water consumption. Sap flow, soil moisture, soil matric potential and weather conditions were monitored over the wet year 2021 and the dry year 2022. We further applied time-dependent probe misalignment correction to account for measurement errors related to sensor installation and to changes in stem water content over the growing season.

Sapwood depth increased with the increasing DBH of a tree and ranged for beech from 6.5 to 15.5 cm and for Douglas fir from 7.5 cm to 14.5 cm. In general, both beech and Douglas fir in mixture had deeper sap flow profiles compared to their pure stands. Stand-level water consumption was higher in the pure beech than in the Douglas fir stand; the mixed stand consumed even more water than the two pure stands. Further, tree-level water use was related to tree size and the radial sap flow profile. Total tree water consumption was markedly higher in the dry year 2022 than in the moist year 2021 due to a higher evaporative demand.

The findings of this study are crucial for supporting foresters in silvicultural decision making and for better understanding the water cycle dynamics in forest ecosystems in the face of climate change.

How to cite: Paligi, S., Coners, H., Hackmann, C., and Leuschner, C.: Water consumption of mature European beech and Douglas fir trees growing in pure and mixed stands, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16457, https://doi.org/10.5194/egusphere-egu24-16457, 2024.

Warming temperatures, high vapor pressure deficit (VPD), and excess light during the middle of the day can reduce CO₂ assimilation and cause stomatal closure, a phenomenon known as midday depression of photosynthesis. However, the role of light, temperature and VPD in driving the diurnal cycle of photosynthesis remain poorly studied in tropical biomes. Here we use quantum efficiency of photosystem II in the light (ϕPSII) as indicator of photosynthetic efficiency for top-of-canopy leaves for six tree species with distinct leaf morphology, across eight sampling campaigns over two years. We find midday decreases in ϕPSII when temperature, solar radiation and VPD were higher than normal. Interestingly, the difference between leaf temperature and air temperature is the most important factor driving changes in ϕPSII, while light is less prominent. We also estimated canopy temperature using outgoing longwave irradiance and found that canopy temperature deviates from air temperature at air temperatures of around 27-28 °C, likely indicating a thermal threshold for photochemistry at the canopy level. These measurements can be combined with state-of-the-art satellite remote sensing (e.g. solar-induced chlorophyll fluorescence and land surface temperatures) to better understand temperature thresholds to photosynthesis and transpiration across scales.

How to cite: Rodriguez-Caton, M., Seibt, U., Stutz, J., Parazoo, N., Wong, C. Y., Dierick, D., Rao, M. P., Filella, I., Bigwood, J., Cooperdock, S., Penuelas, J., and Magney, T.: Increased leaf temperature reduces photosynthetic capacity of top-of-canopy leaves in the wet tropical forest of Costa Rica, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12656, https://doi.org/10.5194/egusphere-egu24-12656, 2024.

The biological processes of carbon (C) uptake via plant photosynthesis (gross primary productivity, GPP) and carbon loss by autotrophic and heterotrophic respiration (ecosystem respiration, R_eco) each constitute a C flux of approx. 130 Gt C per year, equal to 1/7 of the atmospheric C pool. Still, because the biological processes driving GPP and R_eco are both active during daytime, they are intrinsically difficult to measure directly. The eddy covariance technique, which is effectively the gold standard for measuring net ecosystem exchange (NEE), relies on partitioning models of NEE to estimate GPP and R_eco, but these methods remain debated because other processes, such as inhibition of leaf-level respiration during daytime, are not accounted for.  

In ecosystems with short-stature vegetation like grasslands, shrublands, tundra, and many agricultural systems, light and dark closed chamber measurements at the ecosystem scale enable direct daytime measurements of NEE (under light conditions) and R_eco (under dark conditions) while GPP can be directly estimated as NEE - R_eco. Long-term data series of automated light and dark chamber measurements are, however, very rare.

Here, we present data of > 50,000 measurements over six years from a novel, automated light and dark gas exchange measurement chamber that was tested in heathland, wetland, and agricultural vegetation types. In the heathland, we applied standard eddy covariance gap-filling methods to estimate annual NEE across the six years of observations. The results show annual NEE rates ranging from -96 (net uptake) to 21 (net release) g C m^-2y^-1 over the different years. We further applied standard eddy covariance nighttime and daytime methods to partition the observed NEE measurements into GPP and R_eco. Using the nighttime method, GPP ranged from 966 to 1355 g C m^-2y^-1 while R_eco ranged from 867 to 1372 g C m^-2y^-1. On average, this was only 0-4% higher than observed rates from the chamber measurements. In comparison, the daytime method yielded GPP and R_eco rates that were approximately 11-30% higher than observed rates. The slightly to moderately lower direct measurements with the automatic light and dark chamber could indicate that the chamber observations are able to account at least partially for the daytime leaf-level inhibition of respiration and thus may provide a sound method for measuring the actual rates of GPP and R_eco. While potential biases cannot be ruled out and will be discussed, our results indicate that automated light and dark chambers may provide an additional and highly useful tool for estimating rates of GPP and R_eco in short-stature vegetation and may further serve to help constrain methods for partitioning NEE fluxes observed with other techniques, such as the eddy covariance methodology.

How to cite: Larsen, K. S., Pullens, J. W. M., Christiansen, J. R., Tariq, A., Bruun, S., Lærke, P. E., Larsen, P., and Jørgensen, P.: Using large, automated, light and dark chamber systems to directly measure rates of ecosystem gross primary productivity (GPP) and respiration (Reco), EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17732, https://doi.org/10.5194/egusphere-egu24-17732, 2024.

The Sahel is a semi-arid savanna region located as a transition zone between the dry Sahara Desert in the north and the humid Sudanian savanna in the south. It is one of the poorest and most understudied regions in the world and highly affected by climate change. Remote sensing studies found that the majority of Sahel is greening in the 21^st century, with some areas experiencing browning, which is closely linked to the annual rainfall. Yet, there is a scarcity of in-situ data of the responses of ecosystem to the ongoing changes, which makes it hard to validate Earth Observation findings. In this study, we have quantified Net Ecosystem Exchange (NEE) and its components - Gross Primary Production (GPP) and Ecosystem Respiration (Reco) using 13-year long time series of Eddy Covariance data from Dahra, Senegal. We have found decreasing trends in the carbon sink over the period 2010-2022 and a link to the decreasing water availability. 

How to cite: Wieckowski, A., Tagesson, T., Ardö, J., Vestin, P., and Diatta, O.: Decreased water availability reduces the CO2 sink of a semi-arid savanna in Sahel based on a thirteen-year eddy covariance measurement  , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1761, https://doi.org/10.5194/egusphere-egu24-1761, 2024.

Hurdal (NO-Hur) is a recently labelled ICOS class 2 station in Southeast Norway. It represents a typical southern boreal forest of medium productivity, dominated by old Norway spruce (average tree height: 25 m, ages: up to 100 years) with some pine and broadleaved trees. The eddy covariance technique is used to measure CO2 fluxes on a 42 m tower since 2021 . The measurements have an average footprint area of approximately 63 ha.

In 2023, the region experienced an unusual dry spring and then an extraordinary flood in August. Both events showed significant impact on the Net Ecosystem Exchange (NEE) and heat fluxes. The station is also equipped with automatic dendrometers and sap flow devices on the dominant spruce trees, allowing us to investigate the impact of these events at the individual tree scale. We will present tree growth and transpiration flux at different temporal scales (from sub-daily to seasonal), and relate these single tree observations with environmental variables, ecosystem-level NEE and evapotranspiration using phase synchronization analysis. These observational data will yield insights into carbon and water processes of a boreal forest at different scales in response to multiple disturbances.

How to cite: Lange, H., Zhao, J., and Meissner, H.: Response of carbon, water and energy fluxes to drought and flood at a forest ICOS station in Norway, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3237, https://doi.org/10.5194/egusphere-egu24-3237, 2024.

Vegetation plays a fundamental role in shaping Earth's climate by controlling the energy, water, and carbon cycles across terrestrial landscapes. It exerts influence by altering surface roughness, consuming significant water resources through transpiration and interception, regulating atmospheric CO₂ concentration, and controlling net radiation and its partitioning. This influence propagates through the atmosphere, from microclimate scales to the atmospheric boundary layer, subsequently impacting large-scale circulation and the global transport of heat and moisture. Understanding the feedbacks between vegetation and atmosphere across multiple scales is crucial for predicting the influence of land use and cover changes and for accurately representing these processes in climate models. 

This presentation aims to review the mechanisms through which vegetation modulates climate across scales. Particularly, I will evaluate the vegetation impact on circulation patterns, precipitation and temperature during extreme events, such as droughts and heatwaves. Key questions regarding the influence of vegetation feedbacks during these events will be explored: What is the impact of extreme meteorological conditions on ecosystem transpiration? How does vegetation regulate the atmospheric boundary layer and affect the potential intensification and propagation of droughts and heatwaves? Furthermore, I will review the climatic consequences of land use/cover changes, with specific emphasis on extreme events. The goal of this presentation is not to provide a convincing answer to these questions, but rather to highlight the state of science and review recent studies that may help advance our collective understanding of vegetation feedbacks and the role they play in climate.

How to cite: Miralles, D. G.: Vegetation and climate: Exploring feedbacks across scales, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17864, https://doi.org/10.5194/egusphere-egu24-17864, 2024.

Mapping in-situ eddy covariance measurements of terrestrial carbon and water fluxes to the globe is a key method for diagnosing the Earth system from a data-driven perspective. We describe the first global products (called X-BASE) from a newly implemented up-scaling framework, FLUXCOM-X. The X-BASE products comprise of estimates of CO₂ net ecosystem exchange (NEE), gross primary productivity (GPP) as well as evapotranspiration (ET) and, for the first time, a novel fully data-driven global transpiration product (ET_T), at high spatial (0.05°) and temporal (hourly) resolution for the period 2001-2020.

One key improvement of the new products is the much more realistic estimates of global carbon uptake (NEE) at  -5.75 PgC yr^-1, which is a marked improvement compared to previous FLUXCOM versions as well as reconciles the bottom-up global eddy-covariance-based NEE and estimates from top-down atmospheric inversions. The improvement of global NEE was likely only possible thanks to the international effort to improve the precision and consistency of eddy covariance collection and processing pipelines, as well as to the extension of the measurements to more site-years resulting in a wider coverage of bio-climatic conditions. However, X-BASE global net ecosystem exchange shows a very low inter-annual variability, which is common to state-of-the-art data-driven flux products and remains a scientific challenge.

With 125 PgC yr^-1, X-BASE GPP is slightly higher than previous FLUXCOM estimates, mostly in temperate and boreal areas and shows a good agreement with TROPOMI based SIF. X-BASE evapotranspiration amounts to 74.7x10³ km³ yr^-1 globally, but exceeds precipitation in many dry areas likely indicating overestimation in these regions. On average 57% of evapotranspiration are estimated to be transpiration, in good agreement with isotope-based approaches, but higher than estimates from many land surface models.

Despite considerable improvements to the previous up-scaling products, many further opportunities for development exist. Pathways of exploration include methodological choices in the selection and processing of eddy-covariance and satellite observations, their ingestion into the framework, and the configuration of machine learning methods. Here we will outline how the new FLUXCOM-X framework provides the necessary flexibility to experiment, diagnose, and converge to more accurate global flux estimates.

How to cite: Nelson, J. A., Walther, S., Kraft, B., Gans, F., Duveiller, G., Weber, U., Hamdi, Z. M., Zhang, W., and Jung, M. and the FLUXCOM Contributors: Terrestrial carbon and water flux products from an extended data-driven scaling framework, FLUXCOM-X, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16573, https://doi.org/10.5194/egusphere-egu24-16573, 2024.

Photosynthesis rate is the key element in the carbon cycle process. Accurate photosynthesis rate estimate hinges on the maximum carboxylation rate (V²⁵_cmax). The high uncertainty in deriving V²⁵_cmax has long hampered efforts toward the performance of the photosynthesis models from leaf to global scales. Recently studies suggest a strong relationship between spectral reflectance and V25 cmax,0. We proposed the spectrum-driven V²⁵_cmax simulator using deep learning methods and built the hybrid modelling framework for photosynthesis rate estimation by integrating the data-driven V²⁵_cmax simulator in the process-based model. The performance of hybrid photosynthesis models was evaluated at leaf, field and global scales. At the leaf scale, we developed a novel deep learning architecture, which incorporated spatial attention and prior knowledge of spectral indices calculation modules, to extract the V²⁵_cmax from leaf hyperspectral images. At field scale, we combined the high-resolution unmanned aerial vehicle (UAV) multispectral imagery and convolutional neural networks (CNN) to estimate V²⁵_cmax at the paddy field. At a global scale, we utilized a fully connected deep neural network (DNN) to construct the satellite multispectral-driven V²⁵_cmax model based on the FLUXNET2015 dataset. Our result showed that spectrum information can accurately estimate V²⁵_cmax. The hybrid framework fully extracts the information of all available spectral bands using deep learning to reduce parameter uncertainty while maintains the description of the photosynthetic process to ensure its physical reasonability. We also highlighted the significance of spatial heterogeneity for V²⁵_cmax estimation. This study provides new insight into monitoring photosynthesis rate across different spatial scales.

How to cite: Hu, X., Shi, L., and Deng, X.: Hybrid modelling framework for photosynthesis rate estimation from leaf to global scales: integrating the spectrum-based V25cmax simulator into the process-based model, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4285, https://doi.org/10.5194/egusphere-egu24-4285, 2024.

The data-driven VPRM model is a simple light-use-efficiency model, driven by satellite-derived indices of Enhanced Vegetation Index (EVI) and Land Surface Water Index (LSWI) to extract information at high spatial resolution. High temporal resolution is provided through meteorological driving data, namely 2-m temperature and shortwave radiation at the surface. Two parameters per vegetation type are fit using flux tower measurements from the region for previous years. This well-established model has been widely used to model carbon exchange fluxes (gross primary productivity and respiration) between the land biosphere and the atmosphere. A common application is as a background (prior) model for estimating carbon fluxes through inversion techniques at regional scales, given the high temporal and spatial resolution of the fluxes compared to complex process models.

Historically, the VPRM preprocessor relied on data from the 500-m-resolution MODIS satellite and a static 1-km land cover classification map. As MODIS approaches discontinuation, this presentation introduces an updated VPRM software framework - pyVPRM - capable of handling satellite data from MODIS, VIIRS, and Sentinel-2, as well as high-resolution land cover products from ESA WorldCover and the Copernicus Global Land Service. The extremely high spatial resolution of the Sentinel-2 reflectances and updated land cover maps now allow vegetated area within cities to be resolved. In addition, the framework naturally provides an interface to generate VPRM inputs for use in online mesoscale models, such as the greenhouse gas module of the Weather Research and Forecasting Model (part of the WRF-Chem distribution). In our presentation we provide an overview of the model, present fit parameters for all cases using data from European eddy-covariance towers, and present exemplary applications ranging from city to continental scales.

How to cite: Glauch, T. and Marshall, J.: A Vegetation Photosynthesis and Respiration Model (VPRM) for the post-MODIS era, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16946, https://doi.org/10.5194/egusphere-egu24-16946, 2024.

This research tackles the constraints inherent in current global carbon flux datasets and introduces a groundbreaking new dataset, the Global Carbon Fluxes Dataset (GCFD), which integrates cutting-edge deep learning methodologies alongside in situ measurements. GCFD delivers unprecedented high-resolution spatial and temporal data on Gross Primary Productivity (GPP), Terrestrial Ecosystem Respiration (RECO), and Net Ecosystem Exchange (NEE). The Convolutional Neural Network (CNN) model employed in this study surpasses conventional machine learning techniques, demonstrating robust performance in modeling GPP, RECO, and NEE.

The precision and spatial granularity of GCFD outshine those of alternative global carbon flux datasets, like FLUXCOM, and it exhibits strong coherence with remote sensing vegetation condition data. Serving as a reliable reference for both meteorological and ecological investigations, GCFD is particularly valuable when high-resolution carbon flux mapping is essential. Its reliability has been rigorously tested by comparative analysis against existing data products, revealing insightful details about the global spatial and temporal patterns of carbon fluxes, especially within tropical and dry climate zones where notable trends have emerged.

This study significantly advances our comprehension of worldwide carbon flux dynamics and underscores the untapped potential of deep learning technologies to enhance the quality of carbon flux datasets. Accessible at https://dx.doi.org/DOI:10.11888/Terre.tpdc.300009, GCFD offers data resolutions ranging from 1 km to 9 km.

How to cite: Shangguan, W., Xiong, Z., and Huang, F.: Enhancing Carbon Cycle Understanding through Deep Learning: Development and Validation of the Global Carbon Fluxes Dataset (GCFD), EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2662, https://doi.org/10.5194/egusphere-egu24-2662, 2024.

Tropical forests store about half of the world’s above ground carbon and act as critical climate regulators as they absorbed one third of the global CO2 emissions over the past decades. These estimates of the present-day (and future) land carbon sinks are primarily obtained from land surface models (LSM) which are mechanistic tools that simulate the processes occurring at the interface between the atmosphere, the biosphere and the pedosphere. LSM are hence critical tools for understanding and predicting the dynamics of the land surface, its role in a changing Earth, and the impact of future climate and disturbances on its functioning. However, LSM have become increasingly complex and slow machinery that require heavy expert knowledge and computational tools to run. In this study, we tested whether data-driven (black box) models could efficiently reproduce process-based (mechanistic) models. To do so, we trained machine learning algorithms (gradient-boosted decision trees) with the model outputs of TrENDYv11 that were initially generated to estimate the global land carbon sink. Data-driven models performed extremely well in reproducing the long-term trends and the seasonality of the carbon sink over the Tropics, with an average accuracy of 91% and could further be used to make predictions, including near real-time forecasting of the carbon cycle of forests. We illustrate the latter by quantifying the impacts of last-year El-Niño on tropical ecosystem productivity, with a specific focus on the severe drought in the Amazon. While the simulations of the process-based models will only emerge in a year or so when the different teams will have run their own models, our tool could simulate in near real-time that the 2023 drought was for the largest reduction of Amazon GPP in recent history.

How to cite: Meunier, F., Sitch, S., Dietze, M., Boeckx, P., and Verbeeck, H.: An effective machine learning approach for predicting near real-time ecosystem carbon cycle, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9200, https://doi.org/10.5194/egusphere-egu24-9200, 2024.

Remote sensing of sun-induced fluorescence (SIF) provides a valuable tool to assess vegetation state and productivity at large spatial scales and in an unintrusive way. SIF is the absorbed light re-emitted by chlorophyll pigments in the red to infrared range (650-850 nm). It represents one of the three main mechanisms that plants use to dissipate the absorbed light energy, the other two being photochemistry and thermal dissipation (non-photochemical quenching, NPQ). Because both photosynthesis and SIF emission occur at the chloroplast level and share the same excitation energy, SIF can be related to photosynthesis. SIF has been successfully used to predict GPP and the absorbed photosynthetic active radiation (APAR), and in recent years, it has been implemented in state-of-the-art radiative transfer models and several TBMs. Still, the development of SIF-based methods for the prediction of GPP is hindered by the lack of data, especially in regard to coupled GPP-SIF-NPQ estimates. The NPQ mechanism has proven to be the dominant energy dissipation pathway, especially during extreme heat events, and it is the key missing element to correctly relate SIF to GPP. However, so far, SIF has mostly been linked to GPP in a simplistic way, without properly considering the effect of NPQ and with no explicit calculation of the allocation of excitation energy.

The present work aims at presenting a novel dataset from the AustroSIF project. In this project, we collected time series of ground-based active and passive chlorophyll fluorescence and hyperspectral reflectance from 7 eddy-covariance flux tower sites in Europe. The dataset includes sites from Austria, Italy, Poland, Germany and Spain. Key variables present in the dataset include GPP from eddy-covariance, SIF and reflectance-based indices from tower-mounted hyperspectral spectrometers, as well as NPQ, photochemical quenching (PQ), and electron transport rate (ETR) from a continuous pulse amplitude modulation (PAM) instrument. These data have been obtained for periods varying from 3 to 9 months per site between 2018-2022. In this contribution, we will present the dataset and highlight potential applications for model development and improved mechanistic understanding of the SIF-GPP-NPQ interplay. Prospective applications include improved NPQ characterizations in models such as SCOPE (a radiative transfer and energy balance model) and ORCHIDEE (a terrestrial biosphere model capable of ingesting SIF).

How to cite: Martini, D., Sakowska, K., Migliavacca, M., Julitta, T., Hammerle, A., Schwarz, M., Scholz, K., Galvagno, M., Duveiller, G., Pacheco-Labrador, J., and Wohlfahrt, G.: Introducing AustroSIF: A compilation of combined passive and active fluorescence data at flux tower sites across Europe; dataset overview and potential applications, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-18884, https://doi.org/10.5194/egusphere-egu24-18884, 2024.

During the 2022 CloudRoots-Amazonia field campaign high-frequency measurements of state variables and atmospheric composition were taken at the Amazon Tall Tower Observatory (ATTO) in Brazil. Specifically, we measured wind fields, radiation, H₂O and CO₂ mole fractions, and H₂O and CO₂ isotopic compositions at 4Hz or faster at 57m height. A main objective was to use these high-frequency measurements to investigate how the coherent canopy-atmosphere turbulent structures are influenced by the non-stationary passage of clouds. Novel in our investigation is the use of quadrant analysis in combination with high frequency isotopic composition measurements, as well as our approach to finding lag between cloud passages and turbulent variables.

Using quadrant analysis to distinguish sweeping and ejection motions, we find that the passage of clouds influences the transport of scalars and energy. Preceding the passage of well-developed cumulus clouds, we see that the gust front forcefully ejects this subcanopy air into the atmospheric mixed layer. It seems that this is an effective upward transport mechanism for CO₂ and other scalars emitted by the soil, plant roots, and understory. The understory separation was based on the q_CO2’ > 0, q_H2O’ > 0 quadrant. Keeling plots of this quadrant, made using the dD isotope of H₂O, indicate strong midday depletion of understory water vapour (-20 ‰). This effect can only be explained by a major downwards moisture flux from the atmosphere into the soils through the process of condensation, even - or especially – when the air temperatures are highest. Finally, we show what the responses of the major fluxes (H, LE, F_CO2) and their respective isotopologues are to the passage of clouds, including their lag times. Our study contributes to an improved quantification and understanding of the canopy-atmosphere fluxes influenced by the perpetual presence of clouds.

How to cite: Moonen, R., Adnew, G., Hartogensis, O., Vilà-Guerau de Arellano, J., Bonell Fontas, D., and Röckmann, T.: High frequency isotopic composition measurements to classify cloud induced turbulent patterns above the amazon rain forest, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17878, https://doi.org/10.5194/egusphere-egu24-17878, 2024.

Land exchanges carbon with the atmosphere through numerous processes, such as photosynthesis and vegetation respiration. As land carbon uptake is greater than land carbon emissions, land surface represents nowadays a carbon sink, absorbing around a third of total anthropogenic carbon emissions every year. Land surface models, used in global climate models, model these processes and provide estimates of net carbon fluxes between the atmosphere and land surfaces. However, simulations of these models still contain significant uncertainties. Other methods, known as Atmospheric CO2 Inversion, exist to estimate these fluxes, and are not consistent with estimates given by land surface models which usually provide a stronger carbon sink in the tropics than Atmospheric CO2 Inversions. These methods cannot provide simulations of the future that are needed for future projections of Climate Models. For that reason, it is important to understand and improve key processes controlling ecosystem carbon budgets, and to embed this understanding in predictive models. Land surface models typically use many free parameters to describe vegetation, which need to be rigorously calibrated or tuned. Here, we aim at calibrating ORCHIDEE, the land surface model used by the IPSL Earth system model, on the atmospheric inversion fluxes (taken as data-driven constraints) in order to study ORCHIDEE's ability to reconcile with Atmospheric CO2 Inversion by finding physically acceptable parameter sets, or detect models’ inability to recover the same spatio-temporal distribution of carbon fluxes. Calibration usually requires many model simulations, which are very costly. Emulators, and especially Gaussian processes, can replace the computationally time-consuming model and help us to run a large number of simulations to fill the parameter space and rule out parameter subspace that give inconsistent simulation. This method, called History Matching, is emerging in the climate community and has shown many advantages. We show the capacity of History Matching to calibrate ORCHIDEE on global simulations using different targets: Known, using twin experiments, which leads to a very rich source of information on parameter sensitivity, uncertainty, equifinality and global and specific knowledge of the model. Unknown, using fluxes from atmospheric CO2 inversion which could also be combined with vegetation activity data (i.e, such as vegetation fluorescence) to add a physical constraint to parameter calibration. This calibration can provide parameter sets that reconcile to a certain extent bottom-up and top-down approaches, or key information on missing processes in ORCHIDEE that need to be added or modified. In both cases, this is highly instructive and leads to a better understanding of the model and processes being modeled and highlights the potential of current land surface models to simulate carbon flux distribution compatible with existing atmospheric CO2 observation (in situ or from satellite).

How to cite: Beylat, S., Raoult, N., Bastrikov, V., Hourdin, F., Chevallier, F., Bacour, C., Ottlé, C., and Peylin, P.: Calibration of a Land Surface Model to adjust the carbon balance using History Matching., EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6394, https://doi.org/10.5194/egusphere-egu24-6394, 2024.

Increasing wetland CH₄ emissions are projected to have significant implications for keeping global warming below 2 °C. However, to better understand the future dynamics of wetland CH₄ emissions, improved deployment of observational systems (e.g., aircraft and flux towers) is required. One avenue to allow targeted field observations is to improve subseasonal-to-seasonal (S2S) CH₄ forecasting. Here, we present a workflow that enables low-latency (<4-week lag) ecosystem model carbon cycle products associated with the United States of America Interagency Greenhouse Gas (GHG) Center. The workflow increases accessibility to the LPJwsl v2.0 dynamic global vegetation model by providing near-real-time CH₄, CO₂, and other carbon cycle products to science users, allowing interaction with the codebase from a user-friendly website front-end integrated with Amazon Web Services computing resources. In the immediate future, with this increase in accessibility and decreased latency in carbon cycle products, analyses related to the onset of the 2023 El Niño mode of ENSO can be rapidly implemented to improve our understanding of carbon cycle dynamics. To demonstrate the utility of the near-real-time carbon cycle products, we couple 9-month GMAO GEOS climate forecast data with MERRA2 S2S reanalysis data to forecast LPJ carbon products until the end of the 2024 calendar year. LPJ S2S carbon forecasts are verified against historic ENSO anomalies for skillfulness. We highlight regions and periods forecasted to have anomalously higher or lower CH₄ emissions during the late 2023 and early 2024 strong El Niño cycle. S2S CH₄ forecasts enable the pre-positioning of in situ measurement networks to improve the coverage of CH₄ observations. By migrating institutional, high-performance computing processes to a cloud-ecosystem framework, we provide increased access to carbon cycle products on a near-real-time basis.

How to cite: Quinn, C., Colligan, T., Calle, L., and Poulter, B.: Low-latency forecasting framework for assessing ENSO impacts on terrestrial carbon cycle, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13079, https://doi.org/10.5194/egusphere-egu24-13079, 2024.