BG3.18

Increasing water-use efficiency (WUE), the ratio of carbon gain to water loss, is a key mechanism that enhances carbon uptake by terrestrial vegetation under rising atmospheric CO₂ (c_a). Existing theory and empirical evidence suggest a proportional increase of WUE in response to rising c_a as plants maintain a relatively constant ratio between the leaf internal (c_i) and ambient (c_a) partial CO₂ pressure (c_i/c_a). This has been hypothesized as the main driver of the strengthening of the terrestrial carbon sink over the recent decades. However, proportionality may not characterize CO₂ effects on WUE on longer time-scales and the role of climate in modulating these effects is uncertain. We evaluated the long-term WUE responses to c_a and climate from 1901-2012 CE by reconstructing intrinsic WUE (iWUE, the ratio of photosynthesis to stomatal conductance) using carbon isotopes in tree rings across temperate forests in the northeastern USA. We further replicated iWUE reconstructions at eight additional sites for the 1992-2012 period-overlapping with the common period of the longest flux-tower record at Harvard Forest to evaluate the spatial coherence of recent iWUE variation across the region. Finally, we compared tree-ring based and modelled c_i/c_a over the 1901-2012 period to examine whether temporal patterns of c_i/c_a reconstructions are consistent with predictions based on the optimality principle of balancing the costs of water loss and carbon gain.

We found that iWUE increased steadily from 1901 to 1975 CE but remained constant thereafter despite continuously rising c_a. This finding is consistent with a passive physiological response to c_a and coincides with a shift to significantly wetter conditions across the region. Tree physiology was driven by summer moisture at multi-decadal time-scales and did not maintain a constant c_i/c_a in response to rising c_a indicating that a point was reached where rising CO₂ had a diminishing effect on tree iWUE.  The c_i/c_a derived from tree-ring d¹³C and the predicted values based on the optimality theory model had similar median values over the 1901-2012 CE period, though with a modest agreement (R²_adj = 0.22, p < 0.001). The reconstructed and predicted c_i/c_a trends were not statistically different from 0 when estimated over the 1901-2012 CE period; however, isotope-based reconstruction of the c_i/c_a trendshowed distinct multidecadal variation while the predicted c_i/c_a remained nearly constant. Our results challenge the mechanism, magnitude, and persistence of CO₂’s effect on iWUE with significant implications for projections of terrestrial productivity under a changing climate.

How to cite: Belmecheri, S., Maxwell, R. S., Taylor, A. H., Davis, K. J., Guerrieri, R., Moore, D. J. P., and Rayback, S. A.: Predicted and observed multidecadal variations of tree physiological responses to climate and rising  CO2: insights from tree-ring carbon isotopes in temperate forests., EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10443, https://doi.org/10.5194/egusphere-egu21-10443, 2021.

Plant photosynthetic physiology is a crucial process reflecting plant growth and productivity. The maximum rate of Rubisco carboxylation (V_c,max) and the maximum rate of electron transport (J_max) of plant leaves are the main limiting factors of photosynthetic capacity and indispensable parameters in ecosystem mechanism models. Accurate simulation of V_c,max and J_max is vital to improve the prediction precision of vegetation dynamics under the background of climate changes. However, using traditional CO₂ response curves to obtain V_c,max and J_max was time-consuming (about 30 to 60 minutes for each CO₂ response curve) and labor-intensive in the field. The rapid photosynthesis-intercellular CO₂ concentration (A-Ci) response technique (RACiR) provided a potential convenient way to obtain A-Ci curve in an open gas exchange system, which would greatly improve the measurement efficiency. Nevertheless, whether the RACiR detecting method verified by limited conifers and deciduous species (especially poplar trees) in previous studies could be generally used for other plant functional groups remains unclear.

Therefore, here we selected Viburnum Odoratissimum as the target and used Li-cor 6800 to test the applicability of the rapid RACiR detecting method on evergreen species. As the changes of CO₂ ranges and rates are the most important parameters in the method, we set 10 different change ranges of reference [CO₂] (i.e., 400-1500 ppm, 400-200-800 ppm, 420-20-620 ppm, 420-20-820 ppm, 420-20-1020 ppm, 420-20-1220 ppm, 420-20-1520 ppm, 420-20-1820 ppm, 450-50-650 ppm, 650-50-650 ppm) to verify the accuracy of traditional CO₂ response curves and RACiR and to explore suitable [CO₂] change ranges for evergreen species.

Finally, our results showed that V_c,max and J_max calculated by 10 rapid A-Ci response curves except J_max calculated by 650-50-650 ppm [CO₂] were not significantly different from those calculated by traditional A-Ci response curves. Moreover, 400-200-800 ppm [CO₂] compared with the other [CO₂] ranges was suitable for V. Odoratissimum. Our results indicated the advantage of RACiR method for evergreen species and implied that preliminary experiments should be carried out according to specific tree species to determine the most appropriate change range of [CO₂] when using RACiR to calculate V_c,max and J_max.

How to cite: Lin, Q., Zhao, C., Liu, Z., and Tian, D.: A test of the rapid measurement of leaf photosynthesis-intercellular CO2 concentration response curve of an evergreen shrub Viburnum Odoratissimum, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-15242, https://doi.org/10.5194/egusphere-egu21-15242, 2021.

Forests play an important role in climate regulation due to carbon sequestration. However, a deeper understanding of forest carbon flux dynamics are often missing due to a lack of information about forest structure and species composition, especially for non-even-aged and mixed forests. In this study, we combined field inventory data of a mixed deciduous forest in Germany with an individual-based forest gap model to investigate daily carbon fluxes and to examine the role of tree size and species composition for the overall stand productivity. Simulation results show that the forest model is capable to reproduce daily eddy covariance measurements (R² of 0.73 for gross primary productivity and of 0.65 for ecosystem respiration). The simulation results showed that the forest act as a carbon sink with a net uptake of 3.2 t_C ha^-1 yr^-1  (net ecosystem productivity) and an overall gross primary productivity of 18.2  t_C ha^-1 yr^-1. At the study site, medium sized trees (30-60cm) account for the largest share (66%) of the total productivity. Small (0-30cm) and large trees (>60cm) contribute less with 8.5% and 25.5% respectively. Simulation experiments showed, that species composition showed less effect on forest productivity. Stand productivity therefore is highly depended on vertical stand structure and light climate. Hence, it is important to incorporate small scale information’s about forest stand structure into modelling studies to decrease uncertainties of carbon dynamic predictions. Experiments with such a modelling approach might help to investigate large scale mitigation strategies for climate change that takes local forest stand characteristics into account.

How to cite: Holtmann, A., Huth, A., Pohl, F., Rebmann, C., and Fischer, R.: Carbon sequestration in mixed deciduous forests: The importance of mid-storey trees for forest productivity, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-7228, https://doi.org/10.5194/egusphere-egu21-7228, 2021.

The vegetation’s response to climate change is a major source of uncertainty in terrestrial biosphere model (TBM) projections. Constraining carbon cycle feedbacks to climate change requires improving our understanding of both the direct plant physiological responses to global change, as well as the role of legacy effects (e.g. reductions in plant growth, damage to the plant’s hydraulic transport system), that drive multi-timescale feedbacks. In particular, the role of these legacy effects - both the timescale and strength of the memory effect - have been largely overlooked in the development of model hypotheses. This is despite the knowledge that plant responses to climatic drivers occur across multiple time scales (seconds to decades), with the impact of climate extremes (e.g. drought) resonating for many years. Using data from 13 eddy covariance sites, covering two rainfall gradients in Australia, in combination with a hierarchical Bayesian model, we characterised the timescales of influence of antecedent drivers on fluxes of net carbon exchange and evapotranspiration. Using our data assimilation approach we were able to partition the influence of ecological memory into both biological and environmental components. Overall, we found that the importance of ecological memory to antecedent conditions increased as water availability declines. Our results therefore underline the importance of capturing legacy effects in TBMs used to project responses in water limited ecosystems.

How to cite: Page, J., De Kauwe, M., and Abramowitz, G.: Exploring the role of lags and legacies in the productivity of Australian ecosystems, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3684, https://doi.org/10.5194/egusphere-egu21-3684, 2021.

Chlorophyll fluorescence (ChlF) takes place in green leaves of the plants during photosynthesis. It has therefore been proposed that ChlF can be used to track the photosynthetic activity of plants and the current possibility to observe sun-induced chlorophyll fluorescence (SIF) via remote sensing provides an unprecedented tool to monitor terrestrial photosynthesis at global scale. However, the relationship between photosynthesis and ChlF is not linear at all scales and is partly controlled by the non-photochemical quenching - which dissipates excess energy as heat. The relationship between the photochemical and fluorescence yields changes when the photochemical quenching is dominating at low irradiance conditions or at high stress conditions. Interpretation of observed SIF is complicated by its dependence on incoming absorbed radiation, observation geometry and radiative transfer of SIF photons within the canopy. To fully exploit remotely sensed SIF to estimate photosynthesis at ecosystem and global scales, it is important to account for these aspects through modelling that include ecosystem processes.

In this work we have implemented a ChlF model into a state-of-the-art land surface model QUantifying Interactions between terrestrial Nutrient CYcles and the climate system (QUINCY) simulating the terrestrial energy, water and biogeochemical cycles of carbon, nitrogen and phosphorus. The simulation of radiative transfer is highly influential for the simulated SIF signal, but the complex solutions of radiative transfer are computationally too heavy, making them impractical approaches at global scale. Therefore, we have investigated different radiative transfer techniques for the SIF signal of varying complexity at site scale in Niwot Ridge, U.S. The most complex solution is based on the mSCOPE and Fluspects model, that explicitly calculates signal transfer. The intermediate solution is based on a two-stream flux approach and the most simple is using a simple fraction for the escape ratio of SIF. Our aim is to assess which solution is most suitable for simulating the SIF signal at different scales and also test different formulations for modelling of non-photochemical quenching.

How to cite: Thum, T., Pacheco-Labrador, J., Magney, T., Migliavacca, M., Quaife, T., and Zaehle, S.: Using different solutions for radiative transfer of solar-induced fluorescence in a land surface model, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8316, https://doi.org/10.5194/egusphere-egu21-8316, 2021.

In a boreal region, terrestrial vegetation carbon balance is controlled by vegetation phenology, which steers photosynthesis, respiration, and biomass turnover processes. In the absence of a full mechanistic understanding of environmental processes controlling vegetation phenology (i.e. senescence, dormancy, chilling), empirically-based models are often applied. These models are typically based on greenness proxies (obtained from satellite data) with fixed amplitude thresholds (e.g. of 15%, 50%, 90%) to determine timings of different phenological events. Yet, it is not known how well those percentiles correspond with the timings of events such as the green-up and the senescence. Especially in the boreal region, estimating timings of different phenological events across large spatial scales remains challenging due to the lack of sufficient ground validation data representative of both forest tree canopy and forest understory species compositions, which are both observed by a satellite sensor. From the land surface modeling perspective, there is a need to develop methods to improve the mapping of phenological events for prudent prediction of the land vegetation-atmosphere interactions under different future climates. In this study, we developed a new approach for calibrating boreal forest greenness amplitude thresholds which indicate timings of different phenological transitions in satellite data. The new approach to calibrate satellite-based greenness thresholds was demonstrated using boreal Finland as a case study area (60-70 N°). Using the approach, we computed satellite-based phenological events and compared them to ground reference data on temperature from a network of meteorological stations across Finland. We also investigated the effects of using different phenological events or ground reference temperature data on estimated growing season length. Results showed that while the standard greenness amplitude threshold values corresponded fairly well with the growing season start, the autumn phenology was not well captured by the standard greenness amplitude threshold values, which has direct impact on growing season length. Based on our data, boreal conifer forest senescence (default is 90%) corresponds with the timing of greenness amplitude of ~45%, while boreal conifer forest dormancy (default is 15%) corresponds with the timing of reaching greenness amplitude of ~0%. The approach allows flexible application across spatial scales (i.e. point or grid) and different satellite sensors, and may be combined with any land cover product, and it provides a meaningful linking between surface temperature data and seasonal reflectance measured by satellite sensors.

How to cite: Majasalmi, T. and Rautiainen, M.: A new approach for improving representation of boreal forest phenology in land surface models, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-210, https://doi.org/10.5194/egusphere-egu21-210, 2021.

Rapid warming in northern high latitudes during the past two decades may have profound impacts on the structures and functioning of ecosystems. Understanding how ecosystems respond to climatic change is crucial for the prediction of climate-induced changes in plant phenology and productivity. Here we investigate spatial patterns of polynomial trends in ecosystem productivity for northern (> 30 °N) biomes and their relationships with climatic drivers during 2000–2018. Based on a moderate resolution (0.05°) of satellite data and climate observations, we quantify polynomial trend types and change rates of ecosystem productivities using plant phenology index (PPI), a proxy of gross primary productivity (GPP), and a polynomial trend identification scheme (Polytrend). We find the yearly-integrated PPI (PPI_INT) shows a high degree of agreement with an OCO-2-based solar‐induced chlorophyll fluorescence GPP product (GOSIF-GPP) for distinct spatial patterns of trend types of ecosystem productivities. The averaged slope for linear trends of GPP is found positive across all the biomes, among which deciduous broadleaved and evergreen needle-leaved forests show the highest and lowest rates respectively. The evergreen needle-leaved forests, low shrub, and permanent wetland show linear trends in PPI_INT over more than 50% of the covered area and permanent wetland also shows a large fraction of the area with the quadratic and cubic trends. Spatial patterns of linear trends for growing season sum of temperature, precipitation, and photosynthetic active radiation have been quantified. Based on the partial correlations between PPI_INT and climate drivers, we found that there is a consistent shift of dominant drivers from temperature or radiation to precipitation across all the biomes except the permeant wetland when the trend type of ecosystem productivity changes from linear to non-linear. This may imply precipitation changes in recent years may determine the linear or non-linear responses of ecosystem productivity to climate change. Our results highlight the importance of understanding how changes in climatic drivers may affect the overall responses of ecosystems productivity. Our findings will facilitate the sustainable management of ecosystems accounting for the resilience of ecosystem productivity and phenology to future climate change.

How to cite: Zhang, W., Jin, H., Jamali, S., Duan, Z., Wu, M., Sun, H., and Ran, Y.: Elucidating climatic controls of polynomial trends in interannual variations of northern ecosystem productivities, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12919, https://doi.org/10.5194/egusphere-egu21-12919, 2021.

It is very important to obtain regional crop growth conditions efficiently and accurately in the agricultural field. The data assimilation between crop growth model and remote sensing data is a widely used method for obtaining vegetation growth information. This study aims to present a parallel method based on graphic processing unit (GPU) to improve the efficiency of the assimilation between RS data and crop growth model to estimate rice growth parameters. Remote sensing data, Landsat and HJ-1 images were collected and the World Food Studies (WOFOST) crop growth model which has a strong flexibility was employed. To acquire continuous regional crop parameters in temporal-spatial scale, particle swarm optimization (PSO) data assimilation method was used to combine remote sensing images and WOFOST and this process is accompanied by a parallel method based on the Compute Unified Device Architecture (CUDA) platform of NVIDIA GPU. With these methods, we obtained daily rice growth parameters of Zhuzhou City, Hunan, China and compared the efficiency and precision of parallel method and non-parallel method. Results showed that the parallel program has a remarkable speedup (reaching 240 times) compared with the non-parallel program with a similar accuracy. This study indicated that the parallel implementation based on GPU was successful in improving the efficiency of the assimilation between RS data and the WOFOST model and was conducive to obtaining regional crop growth conditions efficiently and accurately.

How to cite: Zhao, B., Liu, M., Wu, J., Liu, X., Liu, M., and Wu, L.: An efficient and accurate method for obtaining regional scale rice growth conditions based on WOFOST model and satellite images, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10688, https://doi.org/10.5194/egusphere-egu21-10688, 2021.

Biophysical parameters such as the Leaf Area Index (LAI), Fraction Absorbed Photosynthetically Active Radiation (FAPAR) and Canopy Water Content (CWC) are key inputs for ecological, meteorological and agricultural applications and models. Moreover, LAI and FAPAR are considered Essential Climate Variables (ECVs) which are feasible for global climate observation. Within this context, there are two main issues to achieve a reliable biophysical variable retrieval: cloud contamination and oversimplified uncertainties from the operational products. We propose a methodology based on a hybrid method which inverts a radiative transfer model (PROSAIL) with artificial neural networks (ANN) to produce 30m resolution continuous time series of biophysical variables (FAPAR, LAI, FVC, CWC, CCC) over large areas. To obtain gap free input reflectance data, we used a cloud optimized fusion algorithm (HISTARFM) combining MODIS and Landsat information. In addition, HISTARFM provides realistic uncertainty estimates along with the fused reflectances. This valuable information allows us to carry out an exhaustive uncertainty analysis considering the aleatoric uncertainty (data error) that needs to be propagated through the ANN, and the epistemic uncertainty (model error). We validate our biophysical retrieval with operational MODIS and Copernicus products. This study is performed over the contiguous US (CONUS) area with Google Earth Engine (GEE). The proposed retrieval methodology combined with the unprecedented GEE computational power allows to obtain high spatial resolution biophysical products and realistic uncertainty estimates to capture the needed spatial detail and adequately monitor croplands and heterogeneous vegetated landscapes at very broad scales.

How to cite: Martínez-Ferrer, L., Moreno-Martínez, Á., Muñoz-Marí, J., Izquierdo-Verdiguier, E., Campos-Taberner, M., García-Haro, J., Maneta, M., Robinson, N., Clinton, N., Kimball, J., Running, S. W., and Camps-Valls, G.: Epistemic and aleatoric uncertainty maps in high resolution biophysical parameter retrieval, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-15196, https://doi.org/10.5194/egusphere-egu21-15196, 2021.

Machine learning (ML) provides a powerful set of tools that can improve the accuracy of distribution models by automatically representing underlying ecological relationships empirically captured from large sets of data. However, more recent methodological advances in ML that have been less frequently applied, e.g. in the field of deep learning, may yield additional potential for Distribution Modeling. In this project, we use two ML algorithms, Random Forest (RF) and multi-layer feed-forward artificial neural networks (ANN), to predict the occurrences of Vegetation Types (VT) across Norway. Accurate predictions may support environmental management or the validation of earth system models. The VT data (derived from the AR18x18 data set; n=31 classes) covers the entire spatial scope in 0.9 ha plots on a systematically sampled 18 km grid (n = 1,081 plots, n = 22,154 observations). It was obtained through a field-based survey by a group of trained experts between 2004 and 2014. We use the cloud-based platform "Google Earth Engine" to generate a set of remotely sensed predictor variables based on SENTINEL-2 satellite imagery (i.e. surface reflectance from 12 spectral bands and six vegetation indices). These are then combined with ancillary environmental rasters used previously to model Norwegian VT distribution, e.g. representing climate, land cover, or geological properties (n=55, before one-hot encoding). Preliminary results suggest that in both modeling approaches, the generated SENTINEL-2 variables, particularly the Normalized Difference Vegetation Index (NDVI), have the highest predictive power as measured by permutation importance. The mean overall accuracy using 5-fold cross-validation shows only minor differences between the two methods (approx. 0.45 for ANN vs. 0.44 for RF; the respective F1-scores are 0.35 for ANN and 0.34 for RF; most frequent class baseline accuracy = 0.136). The modeling challenges we currently face include a class imbalance in the VT data set, reconciling the different spatial resolutions of the environmental predictors, and discrepancies in the timing of data acquisition. The next steps in the project will be to incorporate spatial cross-validation into the workflow and to analyze the differences between the ML methods in detail (e.g. regarding the ability to model rare VTs or differences in variable importance). Moreover, we will evaluate the possibility to include additional satellite data sources. This work is a contribution to the Strategic Research Initiative ‘Land Atmosphere Interaction in Cold Environments’ (LATICE) of the University of Oslo.

How to cite: Keetz, L. T., Bryn, A., Horvath, P., Skarpaas, O., Tallaksen, L. M., and Žliobaitė, I.: Using machine learning to model the distribution of Vegetation Types across Norway, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-15211, https://doi.org/10.5194/egusphere-egu21-15211, 2021.

The ALtimetry for BIOMass (ALBIOM) project is an ESA-funded Permanent Open Call Project that aims to retrieve forest biomass using Copernicus Sentinel-3 (S3) altimeter data. The overall goal of ALBIOM is to estimate biomass with sufficient accuracy to be able to increase existing satellite data for biomass retrieval, as well as to improve the global mapping and monitoring of this fundamental variable.  

The project core tasks consist of 1) an analysis of the sensitivity of altimetry backscatter data on land parameters; 2) the development and validation of a Sentinel-3 altimeter backscatter simulator, including the effect of both topography and vegetation  and 3) the development and validation of a machine-learning biomass estimation algorithm.

Here we present a summary of the results obtained from the project. The sensitivity analysis reveals that both the altimetric waveforms and the corresponding Normalised Radar Cross Sections (NRCSs) can be sensitive to the presence of biomass in the order of 100-400 tons/ha, but they are also influenced by topography and water bodies. Different sensitivities with respect to the different frequencies and resolution modes are observed, highlighting non-linear behaviours of the NRCSs. The use of differential NRCSs, defined as the difference among those calculated over two different bandwidths, was demonstrated to be not necessarily more sensitive to vegetation, as it was instead highlighted by previous studies like [Papa et al., 2003].

The tracking window often appears partly or completely misplaced, when the tracking mode is in open-loop mode prescribing a predetermined range, and its size is often not long enough when collecting data over land, especially over regions with complex topography. The length and correct positioning of the tracking window over land represent therefore critical aspects for a study like ALBIOM.

The modelling work has been focused on the development of a merged model approach to simulate altimeter waveforms over vegetated areas. The merging is obtained via the simultaneous use of the modifiedTor Vergata Scattering Model (TOVSM) [Ferrazzoli and Guerriero, 1995, 1996] to simulate the waveform of a flat surface covered by forest vegetation, and the use of the Soil And Vegetation Reflection Simulator (SAVERS) [Pierdicca et al., 2014], originally conceived for GNSS-Reflectometry, and here adapted to the Altimetry system. The simulator developed within ALBIOM shows promising ability to reproduce the general characteristics of the S3 waveforms. The simulations related to forested surfaces present at least two peaks, due to the top of canopy and to the ground, but the presence of topography may introduce other peaks in the waveforms, making the identification of vegetation and topographic effects challenging.

Initial results on the algorithm development using Artificial Neural Networks (ANN) highlight some promising biomass estimates over specific areas (e.g. Central Africa) but also differences in algorithm performances among different regions. The corrected “ice” backscatter coefficient showed the highest sensitivity to biomass, but its values are often invalid over land, which limits the number of meaningful retrievals. The different altimeter tracking mode of Sentinel-3 over different areas of the globe (i.e., open loop and closed loop) could also be responsible for the differences in results.

How to cite: Clarizia, M. P., Pascual, D., Guerriero, L., De Felice-Proia, G., Vittucci, C., Comite, D., Pierdicca, N., Restano, M., and Benveniste, J.: Recent Results from the ALBIOM Project on Biomass Estimates from Sentinel-3 Altimetry Data, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12180, https://doi.org/10.5194/egusphere-egu21-12180, 2021.

Plant roots have less water available when soils have low moisture content and, consequently, limit their root-to-leaf water potential gradient to protect their xylem, which reduces H₂O and CO₂ exchanges with the atmosphere. In vegetation, hydrological and land-surface models, plant responses to reduced available water in the soil have been implemented in various ways depending on data availability, type of ecosystem, and modelling assumptions. Most models use soil water stress functions – commonly known as beta functions – to reduce transpiration and carbon assimilation, by applying a factor that reflects the soil water availability for plants. These functions usually produce reasonably satisfactory results, but rely on the information on soil properties (e.g. wilting point and field capacity) that are not widely available. On a global level, soil information is mediocre, and data uncertainty is compensated by tuning parameters that rarely represent a physiological process. We propose instead the use of a beta function derived from a mass-balance approach focused on the root zone water capacity. This method quantifies the root zone water storage by calculating the accumulated water deficit based on the balance between water influxes and effluxes, and it does not require land-cover or soil information. We assessed how our approach performs compared to those other soil water stress functions. We used global datasets, including WDFE5 and PMLv2, to extract precipitation and evapotranspiration and compute water deficit. For most vegetation types and climates our approach yielded promising results. Worst results were found for some (semi-)arid sites due to the overestimation of the water deficit. We aim to deliver an approach that can be easily applied on global scales.

How to cite: Nóbrega, R. and Prentice, I. C.: Developing a climate-driven root zone water stress function for different climates and ecosystems, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3480, https://doi.org/10.5194/egusphere-egu21-3480, 2021.