Intact Amazon forests hit a record low gross primary productivity level in 2023-2024

Intact Amazon forests hit a record low gross primary productivity level in 2023-2024 | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Biological Sciences - Article Intact Amazon forests hit a record low gross primary productivity level in 2023-2024 Felicien Meunier, Pascal Boeckx, Santiago Botía, Marijn Bauters, and 9 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4161696/v2 This work is licensed under a CC BY 4.0 License Status: Posted Version 2 posted You are reading this latest preprint version Show more versions Abstract In the Amazon, the dry season of 2023 as well as the beginning of the wet season in 2024 were marked by unprecedented high temperatures and large precipitation deficits. While the tropical forests in the Amazon play a crucial role in the global carbon cycle and are a biodiversity hotspot, they were also shown to suffer from El-Niño related droughts in the past, leading to legitimate concerns about the ecological consequences of the recent climate conditions. To this day, while there is a growing effort to make remote sensing products available close to real-time, land surface models that are critical tools to understand the interactions between the biosphere and the environment have lagged behind the present due to the complexity to run and process large model ensembles. In this study, we employed advanced machine learning models trained on state-of-the-art remote sensing and dynamic global vegetation model estimates of gross primary productivity (GPP). The models provide near real-time GPP estimates, revealing significant productivity reductions during the 2023/2024 drought. Negative GPP anomalies were more widespread across the Amazon than during any other recent major drought event. The Climate-GPP relationships that emerged from the models suggest that future temperature increases and changes in precipitation will severely challenge Amazon forest resilience. Biological sciences/Ecology/Forest ecology Earth and environmental sciences/Environmental sciences/Environmental impact Earth and environmental sciences/Climate sciences/Climate change/Climate-change impacts/Environmental health Biological sciences/Plant sciences/Photosynthesis Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction The Amazon hosts the World’s largest contiguous tropical forest 1,2 , and the largest expanse of intact tropical forests (i.e. forests unaffected by disturbances caused by human activities in the recent past) 3 . Amazon tropical forests play a crucial role in global biogeochemical cycles 4 , e.g. being responsible for 25% of the global land carbon sink and storing approximately 120 petagrams of carbon 2,5 , a large fraction of which is in aboveground tree biomass 6 . Intact forests’ critical contribution to climate change mitigation has been frequently acknowledged as they are among the most carbon dense 7 and productive 8 ecosystems on Earth. They are also critical biodiversity hotspots 9 . However, Amazon intact forests are under pressure 10 and have experienced a persistent decline in their sink capacity in the past decades 11 , potentially attributable to drought-induced increases in tree mortality 12,13 . Climate change, including the increasing intensity and severity of extreme events, may play an important role in this progressive shift of intact Amazon forests from a carbon sink to a carbon source, as extreme productivity and mortality anomalies notably occur during or directly after extremely hot and dry events such as the large-scale drought of 2015/2016 14,15 . In 2023, the Amazon was struck by an unprecedented drought event and extreme heat wave which started in the dry season 16 , of which the ecological consequences on tropical forests are yet to be determined. Multiple methods exist to monitor the productivity of tropical forests 17 . Plot (re-)censusing is the most direct way to coincidently measure forest woody productivity and mortality 17 but requires extensive networks of large plots to be accurately representative of a vast biome like the Amazon rainforests, which makes this technique particularly labour-intensive and time-consuming. In addition, plot recensusing has typically a low periodicity (e.g. 5 years) which prevents observations of inter-annual variability 14 . Eddy-covariance flux towers directly measure in situ the net CO 2 exchange between the atmosphere and the ecosystem, but their spatial and temporal coverage in the Tropics is limited due to the cost and the technical difficulties of their implementation and maintenance. Remote sensing 18 (RS), inverse methods 19,20 , land data assimilation 21 , and dynamic global vegetation models 20 (DGVM) can provide large-scale estimates of land carbon fluxes. While more and more remote sensing products are made available close to real time, significant delays remain for new model output release due to the lengthy process of new data acquisition, processing, computing, and formatting even when no new developments are required. This delay is further exacerbated when multiple RS products or DGVMs are combined in ensembles to estimate uncertainties and produce an ensemble mean likely closer to the ground truth 22 . Those technical workflows typically translate to months or years of lag between the occurrence of a climate event like the 2023 drought and the quantification of its ecological impact by DGVMs. This hinders near real time assessment of extreme events such as the extreme climatic conditions experienced in 2023/2024, which is necessary for early-warning systems to inform policy-makers, accelerate research and make it more relevant to society 23 . Here, we used a new near-real time approach to achieve a very fast evaluation of large-scale impacts of climate events. We trained advanced gradient-boosted trees (GBT) models 24 to emulate multiple RS products and DGVM estimates of gross primary productivity (GPP) using climatic variables from state-of-the-art reanalyses (ERA5, JRA-55) as explanatory features. We focused our analysis on intact forests as defined by Potapov et al. (2008) 25 in the Amazon basin 26 and used recent DGVM simulations made for the global carbon assessment (TRENDY 27 ) as training data. The DGVM outputs used for training were only available until 2021 but the GBT models allowed us to extend the time series until near real-time (May 2024), revealing the ecological consequences of the 2023 El-Niño year for the Amazon rainforests in near real-time. This opens new avenues for a broad range of vegetation ecology applications, including early warning signalling systems. Main The Amazon drought of 2023/2024 for intact Amazon forests According to ERA5 re-analyses, the period between July 2023 and April 2024 was the hottest in at least the past 30 years over intact Amazon forests (see methods for definition and mask of intact forests of the Amazon) with an average air temperature of 26.7°C (Figure 1), 1.4°C above the the past 30 years average (25.3°C) and 1.3°C above the average value for the July-May window of the last three decades (25.4°C). July 2023 - April 2024 was on average 0.5°C and 0.1°C above the El Niño related drought events of September 1997 - April 1998 (26.2°C) and August 2015 - March 2016 (26.6°C). Over the 10-month time period between July 2023 and April 2024, the precipitation was also below average with a monthly mean of 142 mm month -1 , which is 47 mm month -1 below its average for that period of the year (Figure 1a and c, supplementary Figure S1). The precipitation anomaly was larger than during previous drought events, which reached -30 mm month -1 between August 2015 and March 2016 and -29 mm month -1 between September 1997 and April 1998 (Figure 1a and c). The hot and dry episode of 2023-2024 was also characterised by exceptionally high monthly minimum (23.1°C) and maximum (31.9°C) temperature (0.9°C and 1.8°C, respectively, above the 1994-2023 averages). In October 2023, the monthly air temperature reached its peak with an average 27.9°C, 1.8°C above the monthly mean of the last three decades (Figure 1b and d). This 1.8°C anomaly corresponded to 3.4 standard deviations of the monthly mean temperature (supplementary Figure S1). In September 2023, the precipitation dropped to an at least 30-years time low with 78 mm, 91 mm below (or 3.2 monthly standard deviations) its monthly average. Both temperature and precipitation anomalies were widespread throughout the Amazon biome (supplementary Figure S2). The impact of drought on intact Amazon forest productivity GBT models trained on RS products and TRENDY simulations could successfully reproduce the seasonality, and the long-term trends, both at the regional level and at the local scale, for each and every RS product and DVGM (supplementary Figures S3). GBT models could also accurately predict the impacts of an unprecedented drought like 2015-2016 on Gross Primary Productivity (GPP), when trained with pre-2015 data only. While GPP has steadily increased over the last three decades according to both RS-trained and TRENDY-trained GBT models (p-value < 0.001 for the slope of the linear models in Figure 2a), it showed noticeable negative anomalies during the past (1997-1998, 2015-2016) and recent (2023-2024) dry and hot episodes (Figure 2b). Since July 2023, GBT models forecasted consistently lower GPP compared to the average of the past three decades (Figure 2a and b). Between July 2023 and April 2024, GBT models projected a GPP reduction of 1.2 Mg C ha -1 year -1 (RS-trained GBT models) and 1.5 Mg C ha -1 year -1 (TRENDY-trained GBT models) compared to the seasonal average for all intact forests in the Amazon, which corresponded to an average decline of 2.6 and 2.2 monthly standard deviations (Figure 2b, supplementary Figure S5). The productivity of 2023-2024 reached its lowest level in October 2023 with RS and TRENDY trained GBT models predicting a respective mean GPP value of 30.7 and 27.9 Mg C ha -1 year -1 , respectively 2.1 and 3.4 Mg C ha -1 year -1 lower than their expected value in the absence of drought. It was the least productive October of the past 30 years according to both RS and TRENDY models. This sharp decrease corresponds to a reduction equivalent to 3.9 and 3.4 monthly standard deviations, a three-decades minimum only surpassed during the month of January 2016 for TRENDY-trained GBT models (3.6 monthly standard deviations then, but for a smaller absolute GPP reduction: 2.2 Mg C ha -1 year -1 in January 2016). Negative anomalies were distributed all across the Amazon during the 2023-2024 dry and hot episode: 98.2% (RS models) or 98.6% (TRENDY models) of all intact Amazon forest grid cells experienced negative GPP anomalies, on average, over the July 2023 - April2024 period (Figure 3). It was the most spatially widespread drought anomaly, surpassing the 1997-1998 drought according to both RS- and TRENDY-trained GBT models (respectively 97.7% and 96.3% of grid cells had negative anomalies then). During the month of October 2023, 78.8%/62.0% (RS-trained GBT models) or 49.3%/42.3% (TRENDY-trained GBT models) of all intact forest grid cells in the Amazon had a negative GPP anomaly corresponding to at least two/three monthly standard deviations. ERA5 mean temperatures and precipitation, as well as their anomalies, showed strong correlations with those of JRA-55 which also marked October 2023 exceptionally dry and hot (supplementary Figure S3). Hence, repeating the analyses with JRA-55 as climate drivers of the GBT models led to very similar conclusions about both the magnitude of the 2023 dry and hot episode and its spatial distribution (supplementary Figures S6 and S7). Climate-GPP relationships of the intact Amazon forests Both RS- and TRENDY-trained GBT models exhibited significant, nonlinear relationships between monthly GPP and mean temperature anomalies (p-value < 0.001, Figure 4), as well as with VPD and precipitation anomalies (supplementary Figure S8). In the next 30 years (2024-2053), air temperature is expected to increase almost linearly over the study area, with warming rates varying between 0.19°C decade -1 (SSP1-2.6) and 0.44°C decade -1 (SSP5-8.5), according to a weighted average of CMIP6 model simulation made to reproduce ERA5 temperatures (supplementary Figures S9). In the next three decades, temperature anomalies similar to the 2023-2024 dry and hot episodes are expected to become common for all climatic scenarios (Figure 4c). By the end of this century, such temperature anomalies are expected to remain the norm for SSP1-2.6 and become cold outliers for all other scenarios (Supplementary Figure S10). These future high temperatures will generate correspondingly high VPD while precipitation is expected to decrease by the end of this century (Supplementary Figure S10), which should further aggravate GPP anomalies. Discussion According to the novel model projection tool that we developed, the 2023-2024 dry and hot atmospheric conditions will lead to exceptional ecological consequences for the intact forests of the Amazon. Driven by abnormally high temperatures and precipitation deficits, intact forests’ GPP has likely hit a record low anomaly in the month of October 2023. This event could reverberate in the upcoming months and years as drought has been unequivocally linked to tree mortality 13,28–31 . TRENDY (Supplementary Figure S11) and other models 32 , in line with field inventory data 14,33 , predict significant reductions of the intact Amazon forest carbon sink for previous large-scale drought events, which is likely to repeat given the 2023-2024 drought magnitude. Regardless of the future emission scenario, temperatures are expected to increase in the next three decades and Amazon intact forest GPP will likely be affected by this temperature increase, the related increase in VPD, and the possible reduction in precipitation. Hence, the CO 2 fertilisation 34 , also observed here in the RS- and TRENDY-trained GBT models, might be outpaced in the near future by the future increase in frequency and severity of droughts, even in the absence of CO 2 saturation for photosynthesis 35 . This applies to GPP and potentially to net ecosystem productivity (NEP), which has shown no increasing trends (p-value 0.19) and strong drought negative responses, according to both TRENDY models (Supplementary Figure S12) and field data 12 . The new method presented in this article is highly flexible, robust (supplementary Figure S3), fast, and easy to use. The models were trained on a specific region but could easily be extended to include more (human-disturbed) biomes, RS products or DGVMs, features (e.g., including sub-monthly variability such as climate extremes as explanatory variables or more features), or even novel types of data (e.g., eddy covariance fluxes, field inventories, or soil moisture observations that were shown to critically impact land carbon fluxes 36 ). After training, the GBT models require no expert knowledge and only need the location (latitudes and longitudes), time (years and months), and climate of the region of interest to simulate GPP. It took less than one second on a regular 8-cores laptop with 16 GB of RAM to run any of the RS or TRENDY surrogates for a full year for the entire intact Amazon forests, contrasting with the large (but very model-specific) number of CPU-hours required to run a typical TRENDY model . Despite these advantages, it is unclear whether the models developed in this study can accurately predict land fluxes under unprecedented climatic circumstances. On the one hand, the emulators were able to reproduce the extreme 2015-2016 dry and hot episode once trained on earlier GPP estimates (supplementary Figure S4). On the other hand, Amazon rainforest tipping points have been postulated 10,37–39 if drought intensity and/or frequency increases beyond some threshold, but so far such threshold responses have not been observed at large scales and hence are not part of the training data. As new RS products or model outputs are released, our GBT model projections can be confirmed or if needed, those newly available data, as well as other covariates, can feed the GBT models as additional training data to improve their robustness. Hybrid AI 40 , as well as DGVM future projection 41 under hotter and drier conditions, will help overcome this potential issue. The methodology suggested here opens the door for a wide range of other potential applications. Because climate variables needed to force those GBT models are widely available including for the future, it makes the development of near-real time ecological forecasting 23 or early warning systems 42 for land fluxes based on RS and DGVM possible. Such a tool could become incredibly useful for decision-makers and conservation policies overall. Such approaches can also help identify (the reasons behind) discrepancies between DVGM outputs and RS products, prioritise necessary model development 43 or refine part of RS algorithms for land flux retrieval. Declarations Acknowledgements FM was funded by the FWO as a senior postdoc and is thankful to this organisation for its financial support (FWO grant no.1214723N). Michael C. Dietze was supported by funding from NSF Grant 1638577 and NASA Carbon Monitoring System (80NSSC17K0711). SDH was supported by the BELSPO project DAMOCO (B2/223/P1/DAMOCO). MP acknowledges the financial support of the European Research Council (grant No. 101117001 LEAFPACE). References 1. Hoang, N. T. & Kanemoto, K. Mapping the deforestation footprint of nations reveals growing threat to tropical forests. Nat. Ecol. Evol. 5 , 845–853 (2021). 2. Saatchi, S. S. et al. Benchmark map of forest carbon stocks in tropical regions across three continents. Proc. Natl. Acad. Sci. 108 , 9899–9904 (2011). 3. Schulze, K., Malek, Ž. & Verburg, P. H. Towards better mapping of forest management patterns: A global allocation approach. For. Ecol. Manag. 432 , 776–785 (2019). 4. Lewis, S. L., Edwards, D. P. & Galbraith, D. Increasing human dominance of tropical forests. Science 349 , 827–832 (2015). 5. Malhi, Y. et al. The regional variation of aboveground live biomass in old-growth Amazonian forests. Glob. Change Biol. 12 , 1107–1138 (2006). 6. Feldpausch, T. R. et al. Tree height integrated into pantropical forest biomass estimates. Biogeosciences 9 , 3381–3403 (2012). 7. Spawn, S. A., Sullivan, C. C., Lark, T. J. & Gibbs, H. K. Harmonized global maps of above and belowground biomass carbon density in the year 2010. Sci. Data 7 , 112 (2020). 8. Li, X. & Xiao, J. A Global, 0.05-Degree Product of Solar-Induced Chlorophyll Fluorescence Derived from OCO-2, MODIS, and Reanalysis Data. Remote Sens. 11 , 517 (2019). 9. de Lima, R. A. F. et al. The erosion of biodiversity and biomass in the Atlantic Forest biodiversity hotspot. Nat. Commun. 11 , 6347 (2020). 10. Flores, B. M. et al. Critical transitions in the Amazon forest system. Nature 626 , 555–564 (2024). 11. Brienen, R. J. W. et al. Long-term decline of the Amazon carbon sink. Nature 519 , 344–8 (2015). 12. Hubau, W. et al. Asynchronous carbon sink saturation in African and Amazonian tropical forests. Nature 579 , 80–87 (2020). 13. Tavares, J. V. et al. Basin-wide variation in tree hydraulic safety margins predicts the carbon balance of Amazon forests. Nature 617 , 111–117 (2023). 14. Bennett, A. C. et al. Sensitivity of South American tropical forests to an extreme climate anomaly. Nat. Clim. Change 13 , 967–974 (2023). 15. Jiménez-Muñoz, J. C. et al. Record-breaking warming and extreme drought in the Amazon rainforest during the course of El Niño 2015–2016. Sci. Rep. 6 , 33130 (2016). 16. Wagner, F. H. et al. Amazon’s 2023 Drought: Sentinel-1 Reveals Extreme Rio Negro River Contraction. Preprint at http://arxiv.org/abs/2401.16393 (2024). 17. ForestPlots.net et al. Taking the pulse of Earth’s tropical forests using networks of highly distributed plots. Biol. Conserv. 260 , 108849 (2021). 18. Heinrich, V. H. A. et al. Large carbon sink potential of secondary forests in the Brazilian Amazon to mitigate climate change. Nat. Commun. 12 , 1785 (2021). 19. Frankenberg, C. et al. Data Drought in the Humid Tropics: How to Overcome the Cloud Barrier in Greenhouse Gas Remote Sensing . https://essopenarchive.org/users/528524/articles/717473-data-drought-in-the-humid-tropics-how-to-overcome-the-cloud-barrier-in-greenhouse-gas-remote-sensing?commit=c1c370fe625963945ecaea0faa78b3b911e10e05 (2024) doi:10.22541/essoar.170923255.57545328/v1. 20. Rosan, T. M. et al. Synthesis of the land carbon fluxes of the Amazon region between 2010 and 2020. Commun. Earth Environ. 5 , 1–15 (2024). 21. Dokoohaki, H. et al. Development of an open-source regional data assimilation system in PEcAn v. 1.7.2: application to carbon cycle reanalysis across the contiguous US using SIPNET. Geosci. Model Dev. 15 , 3233–3252 (2022). 22. Li, L. et al. The optimization of model ensemble composition and size can enhance the robustness of crop yield projections. Commun. Earth Environ. 4 , 1–11 (2023). 23. Dietze, M. C. et al. Iterative near-term ecological forecasting: Needs, opportunities, and challenges. Proc. Natl. Acad. Sci. 115 , 1424–1432 (2018). 24. Greener, J. G., Kandathil, S. M., Moffat, L. & Jones, D. T. A guide to machine learning for biologists. Nat. Rev. Mol. Cell Biol. 23 , 40–55 (2022). 25. Potapov, P. et al. The last frontiers of wilderness: Tracking loss of intact forest landscapes from 2000 to 2013. Sci. Adv. 3 , e1600821. 26. Boulton, C. A., Lenton, T. M. & Boers, N. Pronounced loss of Amazon rainforest resilience since the early 2000s. Nat. Clim. Change 12 , 271–278 (2022). 27. Friedlingstein, P. et al. Global Carbon Budget 2023. Earth Syst. Sci. Data 15 , 5301–5369 (2023). 28. Williamson, G. B. et al. Amazonian Tree Mortality during the 1997 El Niño Drought. Conserv. Biol. 14 , 1538–1542 (2000). 29. Yao, Y. et al. Forest fluxes and mortality response to drought: model description (ORCHIDEE-CAN-NHA, r7236) and evaluation at the Caxiuanã drought experiment. Geosci. Model Dev. Discuss. 1–38 (2021) doi:10.5194/gmd-2021-362. 30. Rowland, L. et al. Death from drought in tropical forests is triggered by hydraulics not carbon starvation. Nature 528 , 119–122 (2015). 31. McDowell, N. G. Mechanisms Linking Drought, Hydraulics, Carbon Metabolism, and Vegetation Mortality. Plant Physiol. 155 , 1051–1059 (2011). 32. Tian, H. et al. Effect of interannual climate variability on carbon storage in Amazonian ecosystems. Nature 396 , 664–667 (1998). 33. Feldpausch, T. R. et al. Amazon forest response to repeated droughts. Glob. Biogeochem. Cycles 30 , 964–982 (2016). 34. Chen, C., Riley, W. J., Prentice, I. C. & Keenan, T. F. CO2 fertilization of terrestrial photosynthesis inferred from site to global scales. Proc. Natl. Acad. Sci. 119 , e2115627119 (2022). 35. Wang, S. et al. Recent global decline of CO2 fertilization effects on vegetation photosynthesis. Science 370 , 1295–1300 (2020). 36. Yang, H. et al. Climatic and biotic factors influencing regional declines and recovery of tropical forest biomass from the 2015/16 El Niño. Proc. Natl. Acad. Sci. 119 , e2101388119 (2022). 37. Cox, P. M. et al. Amazonian forest dieback under climate-carbon cycle projections for the 21st century. Theor. Appl. Climatol. 78 , 137–156 (2004). 38. Parry, I., Ritchie, P. & Cox, P. Evidence of Amazon Rainforest Dieback in CMIP6 Models . (2022). doi:10.5194/egusphere-2022-82. 39. Lovejoy, T. E. & Nobre, C. Amazon Tipping Point. Sci. Adv. 4 , eaat2340 (2018). 40. Koppa, A., Rains, D., Hulsman, P., Poyatos, R. & Miralles, D. G. A deep learning-based hybrid model of global terrestrial evaporation. Nat. Commun. 13 , 1912 (2022). 41. Frieler, K. et al. Scenario setup and forcing data for impact model evaluation and impact attribution within the third round of the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP3a). Geosci. Model Dev. 17 , 1–51 (2024). 42. Boulton, C. A., Good, P. & Lenton, T. M. Early warning signals of simulated Amazon rainforest dieback. Theor. Ecol. 6 , 373–384 (2013). 43. Fisher, R. A. & Koven, C. D. Perspectives on the Future of Land Surface Models and the Challenges of Representing Complex Terrestrial Systems. J. Adv. Model. Earth Syst. 12 , e2018MS001453 (2020). 44. Harris, I., Osborn, T. J., Jones, P. & Lister, D. Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. Sci. Data 7 , 109 (2020). 45. Harris, I., Jones, P. d., Osborn, T. j. & Lister, D. h. Updated high-resolution grids of monthly climatic observations – the CRU TS3.10 Dataset. Int. J. Climatol. 34 , 623–642 (2014). 46. Kobayashi, S. et al. The JRA-55 Reanalysis: General Specifications and Basic Characteristics. J. Meteorol. Soc. Jpn. Ser II 93 , 5–48 (2015). 47. Bi, W. et al. A global 0.05° dataset for gross primary production of sunlit and shaded vegetation canopies from 1992 to 2020. Sci. Data 9 , 213 (2022). 48. Wild, B. et al. VODCA2GPP – a new, global, long-term (1988–2020) gross primary production dataset from microwave remote sensing. Earth Syst. Sci. Data 14 , 1063–1085 (2022). 49. Wang, S., Zhang, Y., Ju, W., Qiu, B. & Zhang, Z. Tracking the seasonal and inter-annual variations of global gross primary production during last four decades using satellite near-infrared reflectance data. Sci. Total Environ. 755 , 142569 (2021). 50. Pedelty, J. et al. Generating a long-term land data record from the AVHRR and MODIS Instruments. in 2007 IEEE International Geoscience and Remote Sensing Symposium 1021–1025 (IEEE, Barcelona, Spain, 2007). doi:10.1109/IGARSS.2007.4422974. 51. Chen, T. & Guestrin, C. XGBoost: A Scalable Tree Boosting System. in Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining 785–794 (ACM, San Francisco California USA, 2016). doi:10.1145/2939672.2939785. 52. Chen, T. et al. xgboost: Extreme Gradient Boosting. (2024). 53. A Short Introduction to the caret Package. https://cran.r-project.org/web/packages/caret/vignettes/caret.html. 54. PEcAn functions used for managing climate driver data. https://pecanproject.github.io/modules/data.atmosphere/docs/index.html. 55. Hijmans, R. J. et al. raster: Geographic Data Analysis and Modeling. (2023). 56. Eyring, V. et al. Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev. 9 , 1937–1958 (2016). 57. Riahi, K. et al. The Shared Socioeconomic Pathways and their energy, land use, and greenhouse gas emissions implications: An overview. Glob. Environ. Change 42 , 153–168 (2017). 58. Vinutha, H. P., Poornima, B. & Sagar, B. M. Detection of Outliers Using Interquartile Range Technique from Intrusion Dataset. in Information and Decision Sciences (eds. Satapathy, S. C., Tavares, J. M. R. S., Bhateja, V. & Mohanty, J. R.) 511–518 (Springer, Singapore, 2018). doi:10.1007/978-981-10-7563-6_53. Methods Study area The Amazon basin ( https://github.com/gamamo/AmazonBasinLimits ) delimited our study area 26 . We further restricted the analyses to intact forests using the intact forest map of Potapov et al. (2008) 25 ( http://www.intactforests.org/data.ifl.html ) for the year 2020. TRENDY model simulations We used simulations of a set of 16 Dynamic Global Vegetation Models (DGVMs), which participated in the past global carbon budgets assessment 27 , named TRENDY-v11. Because we focus on intact forests, we extracted model outputs from TRENDY-v11 scenario S2 in which CO 2 and climate vary with time but which uses a time-invariant “pre-industrial” land use mask. In TRENDY, all models are forced with the same climate forcing, namely CRUJRA, see 44–46 . We mainly extracted GPP monthly outputs from TRENDY-v11 but we also pulled monthly averages of NPP and heterotrophic respiration to compute NEP from their difference for every model that provided those variables as well. NEP values were only used as is (i.e. did not serve for GBT model training) to investigate their correlation with GPP. TRENDY-v11 model outputs are available for the time period 1901-2021. Remote sensing estimates To benchmark TRENDY model outputs, we used four recent GPP products, hereafter referred to as Li and Xiao (2019) 8 , Bi et al. (2022) 47 , Wild et al. (2022) 48 , and Wang et al. (2021) 49 . Briefly, Li and Xiao (2019) developed a global, high temporal (8-days) and spatial (0.05°) resolutions GPP product based on Orbiting Carbon Observatory-2 SIF data, MODIS and meteorological re-analyses through a data-driven approach over the 2000-2022 period 8 . Bi et al. (2022) generated a global 0.05°, 8-day dataset for GPP (1992-2020) with a two-leaf light use efficiency model, driven by CRUJRA reanalysis, ESA-CCI land cover and leaf area index from GLOBMAP 47 . Wild et al. (2022) utilised microwave remote sensing of vegetation optical depth to derive global GPP at moderate spatial (0.25°) and high temporal (8-days) resolutions 48 for the period 1988-2020. Finally, Wang et al. (2021) correlated eddy-covariance GPP and AVHRR near-infrared reflectance from the Land Long Term Data Record 50 to generate a global long-term (1982-2018) time series at high spatial resolution (0.05°) of monthly GPP. We aggregated all four datasets at the monthly timescale for further analyses. AI models We used the extreme gradient boosting (XGBoost 51 ) approach, as implemented in the R-package xgboost 52 , to reproduce time series of GPP as estimated from DGVMs or from remote sensing. XGBoost is an advanced machine learning algorithm reputed to be highly efficient in terms of model performance and computational speed. Individual GBT models were trained for each DGVM/RS estimate of GPP using the longest possible time series and the largest possible area. We splitted all the data into training (60%), validation (20%) and test (20%) for each grid cell and year. So on average, 2.4 months per year served as validation and test for each grid cell while 7.2 months per year were used for training. The months were randomly attributed to training, validation and test dataset. We optimised/tuned GBT model (hyper)parameters on validation data with the R-package caret 53 and 8 resampling iterations, focusing on the maximum decision tree depth and the number of decision trees to grow. Training variables included monthly averages of climate data (monthly mean, minimum, and maximum temperature, monthly precipitation, VPD, total incoming short- and long-wave radiation), atmospheric CO 2 concentration (constant spatially), as well as temporal (year/month) and spatial (latitude/longitude) information. The latter variables were important to include to account for model-specific spatial variability in e.g., soil maps and plant functional type spatial distribution. We computed VPD from sub-daily air temperatures and specific humidity using the R-package Pecan.data.atmosphere 54 before taking the monthly average. Similarly, we used sub-daily air temperature values to compute the daily minimum and maximum temperatures and finally calculate the monthly minimum and maximum temperature as the average of all daily minimum and maximum temperatures of a given month. The CRUJRA meteorological reanalysis, which is used as the driver for TRENDY models, is only updated once a year (typically between April and August), which prevents its use for near-real time forecasting. Therefore we decided to use ERA5 and JRA-55 reanalysis as input data for the GBT models. Doing so we internalised the bias between reanalysis sources into the GBT models, as CRUJRA is constructed by adjusting data from JRA-55 where possible to align with the CRU TS data. We downloaded all ERA5 and JRA-55 climatic drivers available in early June 2024, i.e. until May 2024 and December 2023 respectively. We performed the same model training and validation described above, independently with ERA5 and JRA-55 reanalysis to check the influence of climatic forcings on the model predictions. We also checked how absolute values and anomalies correlated for both reanalysis sources for the region of interest. For ERA5, we used the 2-metre air temperature variable for the air temperature. Final GBT model performance was evaluated on test data only and quantified using the root mean square deviation and the square of the Pearson correlation coefficient (R²) between reference (DGVM outputs or RS estimates) and predicted values. Analyses Given the different resolutions of the individual TRENDY models and the remote sensing GPP estimates, we reprojected the GBT model features (climate forcings) on each DGVM/RS grid before training with the R-package raster 55 . To generate spatial ensemble means of GPP, we also reprojected all rasters to a common grid (ERA5) using bilinear interpolations. When reprojection was needed for the intact forest map, we used bilinear interpolations, introducing decimal numbers. We then limited the analyses to all grid cells with interpolated values larger than 50% intact forest cover. Unless otherwise stated, we limited our analyses to the past 30 years (1994-2023). For generating model ensemble means, we applied simple means to all RS products and DGVM outputs, giving equal weight to each ensemble member. We used the GBT models for extending RS time series to recent years (e.g., 2023/2024) but also to older times (from 1994 to the actual start of the product time series) to avoid discontinuities when computing ensemble means when not all products were available. We compared the recent dry and hot episode of 2023-2024 with previous major drought events, namely August 2015-March 2016, and September 1997-April 1998. We delimited those drought events of the past from their precipitation and air temperature anomalies in ERA5. To assess whether the GBT models were able to accurately predict GPP in case of unprecedented events, we verified how the density distribution of the residuals compared for the test data of the dry and hot episodes of 1997-1998 and 2015-2016 and all other climatic conditions. We also repeated the model training and validation described above using pre-2015 data only and tested the model performance specifically for the 2015-2016 dry and hot episode. In both RS estimates and DGVM simulations, GPP presents both a seasonality and an increasing long-term trend due to CO 2 fertilisation. We detrended the GPP time series by subtracting to the raw data the best linear model estimates from the 1994-2023 time period and its average seasonal variation. We used a linear model for detrending as GPP has been steadily increasing the past decades due to CO 2 fertilisation. We detrended the climate variables by subtracting the overall variable average and its seasonal mean to the raw time series. The anomalies were normalised using the monthly standard deviation of the respective variables for the overall time period (1994-2023). For spatial analyses, we applied the detrending method described above to the grid cells individually, while we first averaged the time series for biome-wide analyses. Anomalies were computed on the model ensemble mean and the average of RS products. To smooth monthly time series, we applied rolling averages with a centred 6-month time window. We tested the effect of this choice by changing the window size between 1 month (no average) and one year and verified how it changed the correlations between variables. For future trends, we downloaded monthly temperature and precipitation outputs from Earth system models (ESM) available in January 2024 for the models participating in the Coupled Model Intercomparison project 6 (CMIP6) 56 . We downloaded model outputs for the historical period and the following shared socioeconomic pathways 57 : SSP1-2.7, SSP2-4.5, SSP3-7.0, and SSP5-8.5. We restricted the analysis to the models which had model outputs available for every shared socioeconomic pathway. We then compared the regional monthly mean temperature and the mean annual precipitation time series predicted by each ESM for the recent historical period (1985-2014) with those of ERA5, and only kept the 10 with the lowest root mean square error for each variable. We then averaged those 10 models using the inverse of the squared bias as weight for both the historical period and the future simulations. Temperature and precipitation anomalies were defined using 1985-2014 or 1994-2023 as reference for the historical period or the future scenarios, respectively, since CMIP6 historical simulations stop in 2014. Temperature anomaly outliers were identified following the classical outlier 1.5 interquartile (IQR) rule 58 : any observation that was smaller than 1.5 IQR below the first quartile or larger than 1.5 IQR above the third quartile was considered as a cold or hot outlier, respectively. Additional Declarations The authors declare no competing interests. Supplementary Files SupplementaryInformation.docx Cite Share Download PDF Status: Posted Version 2 posted You are reading this latest preprint version Show more versions Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4161696","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Biological Sciences - Article","associatedPublications":[],"authors":[{"id":284068276,"identity":"bb7e27f2-27fe-47c5-869d-75b0a20e765f","order_by":0,"name":"Felicien Meunier","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABEUlEQVRIiWNgGAWjYBAC9gbmhgMMBjAukMEPohMKcGvhOcCIpkWyAaTFAL8WVBGDA1C9OLWwNzYe+FFgk8cgdviYdEGBTb7x+dWJHx4YMMjzix3AroXnYMPBHoO0YgbptDTpGQZplttuvN0sAXSY4czZCVi12EskNhxmMDic2CCdYybNY3DYwOzG2Q0gLQkGt7Fr4ZF/CNLyH6gl/xtQy38D4xlnN//Aq0WCEaTlAMgWNqCWAwYG/L3b8NvCkwjyS3Jim3SasTWPQbKBxA3ebRYJBhI4/cLDfvjwhx9/7BL7pZMf3ub5Y2fA3392880fFTby/NLYtcABG5wlAVYpgV85KuA/QIrqUTAKRsEoGAEAAFCHW7kao0T5AAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0003-2486-309X","institution":"Ghent University","correspondingAuthor":true,"prefix":"","firstName":"Felicien","middleName":"","lastName":"Meunier","suffix":""},{"id":284068277,"identity":"9489056c-ec99-4ee4-9ea6-d992492397be","order_by":1,"name":"Pascal Boeckx","email":"","orcid":"https://orcid.org/0000-0003-3998-0010","institution":"Ghent University","correspondingAuthor":false,"prefix":"","firstName":"Pascal","middleName":"","lastName":"Boeckx","suffix":""},{"id":284068278,"identity":"888036a7-6a18-4aae-b0d4-591722ea4766","order_by":2,"name":"Santiago Botía","email":"","orcid":"","institution":"Max Planck Institute for Biogeochemistry","correspondingAuthor":false,"prefix":"","firstName":"Santiago","middleName":"","lastName":"Botía","suffix":""},{"id":284068279,"identity":"77355ac2-6ed4-4bac-ab7f-d8fabfbc57b1","order_by":3,"name":"Marijn Bauters","email":"","orcid":"https://orcid.org/0000-0003-0978-6639","institution":"Ghent University","correspondingAuthor":false,"prefix":"","firstName":"Marijn","middleName":"","lastName":"Bauters","suffix":""},{"id":284068280,"identity":"de1afa97-bfd1-4a73-b720-097707f552a4","order_by":4,"name":"Wout Cherlet","email":"","orcid":"https://orcid.org/0000-0003-3505-6971","institution":"Ghent University","correspondingAuthor":false,"prefix":"","firstName":"Wout","middleName":"","lastName":"Cherlet","suffix":""},{"id":284068281,"identity":"bfdc5818-d058-44aa-bf01-aa28ff4a5084","order_by":5,"name":"Philippe Ciais","email":"","orcid":"https://orcid.org/0000-0001-8560-4943","institution":"Laboratoire des Sciences du Climat et de l'Environnement","correspondingAuthor":false,"prefix":"","firstName":"Philippe","middleName":"","lastName":"Ciais","suffix":""},{"id":284068282,"identity":"cec03b43-0f3f-44e8-984b-4fca85750336","order_by":6,"name":"Steven De Hertog","email":"","orcid":"","institution":"Ghent University","correspondingAuthor":false,"prefix":"","firstName":"Steven","middleName":"","lastName":"De Hertog","suffix":""},{"id":284068283,"identity":"3c614b78-8e90-40d5-bb45-c86b6f0ae68b","order_by":7,"name":"Michael Dietze","email":"","orcid":"","institution":"Boston University","correspondingAuthor":false,"prefix":"","firstName":"Michael","middleName":"","lastName":"Dietze","suffix":""},{"id":284068284,"identity":"f806d557-be1d-48a0-a61d-246a38cab47d","order_by":8,"name":"Marc Peaucelle","email":"","orcid":"https://orcid.org/0000-0003-0324-4628","institution":"INRAE, Université de Bordeaux","correspondingAuthor":false,"prefix":"","firstName":"Marc","middleName":"","lastName":"Peaucelle","suffix":""},{"id":284068285,"identity":"bbfbfdea-0cc4-4944-be0b-901004155f8b","order_by":9,"name":"Thomas Sibret","email":"","orcid":"https://orcid.org/0000-0002-1496-3444","institution":"Ghent University","correspondingAuthor":false,"prefix":"","firstName":"Thomas","middleName":"","lastName":"Sibret","suffix":""},{"id":284068286,"identity":"ce1a904a-1d3d-4905-bd63-e69a1a85b05e","order_by":10,"name":"Stephen Sitch","email":"","orcid":"https://orcid.org/0000-0003-1821-8561","institution":"University of Exeter","correspondingAuthor":false,"prefix":"","firstName":"Stephen","middleName":"","lastName":"Sitch","suffix":""},{"id":284068287,"identity":"1f23c52a-014b-4b0d-8ba0-d2ed2d9ecd7b","order_by":11,"name":"Wei Li","email":"","orcid":"https://orcid.org/0000-0003-2543-2558","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Li","suffix":""},{"id":284068288,"identity":"a98b293c-187c-4e8b-9653-297d2de5ee19","order_by":12,"name":"Hans Verbeeck","email":"","orcid":"https://orcid.org/0000-0003-1490-0168","institution":"Ghent University","correspondingAuthor":false,"prefix":"","firstName":"Hans","middleName":"","lastName":"Verbeeck","suffix":""}],"badges":[],"createdAt":"2024-03-25 08:16:49","currentVersionCode":2,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-4161696/v2","doiUrl":"https://doi.org/10.21203/rs.3.rs-4161696/v2","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":59948887,"identity":"49c7855a-6b0f-49ac-8a33-3ede7896f728","added_by":"auto","created_at":"2024-07-09 16:48:41","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1993178,"visible":true,"origin":"","legend":"\u003cp\u003eSeasonal cycle of the precipitation and the average air temperature (subplots a-b) and their anomalies (subplots c-d) over the intact Amazon forests according to ERA5 reanalyses. The individual years of the 1994-2023 time period are represented by the thin grey lines (subplots a-d) and their averages, by the thick black line (subplots a-b). Major dry and hot periods (1997-1998, 2015-2016, 2023-2024) are highlighted with the coloured lines and are discontinuous because they overlap two calendar years.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-4161696/v2/bbf9bfeb027fbacee318fdb0.png"},{"id":59948886,"identity":"0b22a26f-23d6-4839-9f37-2f34b4d26b9d","added_by":"auto","created_at":"2024-07-09 16:48:41","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1649829,"visible":true,"origin":"","legend":"\u003cp\u003eTime series of the monthly GPP (subplot a) and its normalised anomalies (subplot b) over the intact Amazon forests for the 1994-2024 time period, as generated by GBT models trained on TRENDY simulations (yellow, TRENDY) or remote sensing products (green, RS) with ERA5 re-analyses as model features. The anomalies were normalised by the standard deviations of the respective months and sources. Each data point represents the average of multiple vegetation models or multiple remote sensing products. The coloured dots are the monthly values while the curves are rolling averages over a 6-month time window. In subplot a, the dashed lines represent the respective linear increases used to detrend GPP. The insets (subplots c and d) show the correlations between the TRENDY model ensemble mean and the average of the remote sensing products. In subplot a, the R² of the linear models are 0.20 and 0.28 for RS and TRENDY models, respectively (p-value \u0026lt; 0.001). The coloured rectangles above the figure represent the time coverage of the training data (until the end of 2021 for all TRENDY models, varying between 1992-2001 and 2018-2022 for the different RS products, see methods).\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-4161696/v2/de874a8fe6501f415f900a0b.png"},{"id":59948888,"identity":"62d27fd1-7a01-40f6-831c-a27aaaff4dc6","added_by":"auto","created_at":"2024-07-09 16:48:41","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1302717,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial distribution of the average normalised GPP anomaly for the 2023-2024 dry and hot period as predicted from GBT models trained on remote sensing products (subplot a), or TRENDY models (subplot b). In subplots a and b, the black line delineates our study area and only intact forests were included in the analysis (white grid cells within the Amazon basin are mostly not intact forests). Subplots c shows the marginal density distribution of the anomalies in both cases.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-4161696/v2/2dd6f8598bc1d0e038056072.png"},{"id":59949103,"identity":"2e055242-3de9-474b-b418-caedda58dbb7","added_by":"auto","created_at":"2024-07-09 16:56:41","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":923093,"visible":true,"origin":"","legend":"\u003cp\u003eRelationships between normalised monthly GPP anomalies and monthly mean air temperature anomalies for the intact Amazon forests, based on GBT models trained on remote sensing (subplot a) or TRENDY models estimates of GPP (subplot b). Each data point represents the monthly average of multiple vegetation models (subplot a) or multiple remote sensing products (subplot b). In subplots a-b, temperature and GPP anomalies were smoothed using a moving average over a 6-month time period and the coloured dots and zones correspond to monthly values of recent major dry and hot episodes. The monthly temperature anomalies predicted from CMIP6 simulations are shown in subplot c for the historical period (1985-2014) or the next 30 years (2024-2053) according to the different shared socioeconomic pathways. In subplot c, the temperature anomalies of the 2023-2024 dry and hot episode are highlighted for the ERA5 reanalysis. P-values of the second-order polynomial models are \u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-4161696/v2/0a2e96d8093d25effed1d808.png"},{"id":59949394,"identity":"134dc1e3-bbda-4622-99a1-09516b7b83f2","added_by":"auto","created_at":"2024-07-09 17:04:48","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5870515,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4161696/v2/f2776b20-5c32-4758-8e66-1464ad34b582.pdf"},{"id":59948890,"identity":"d4c537a1-159e-4d9f-984e-f201a96e181d","added_by":"auto","created_at":"2024-07-09 16:48:41","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":7791365,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-4161696/v2/ca43299586881fcb14e023e1.docx"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"Intact Amazon forests hit a record low gross primary productivity level in 2023-2024","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe Amazon hosts the World\u0026rsquo;s largest contiguous tropical forest\u003ca href=\"https://www.zotero.org/google-docs/?IBXUQy\"\u003e\u003csup\u003e1,2\u003c/sup\u003e\u003c/a\u003e, and the largest expanse of intact tropical forests (i.e. forests unaffected by disturbances caused by human activities in the recent past)\u003ca href=\"https://www.zotero.org/google-docs/?fIb5f3\"\u003e\u003csup\u003e3\u003c/sup\u003e\u003c/a\u003e. Amazon tropical forests play a crucial role in global biogeochemical cycles\u003ca href=\"https://www.zotero.org/google-docs/?AwJMyB\"\u003e\u003csup\u003e4\u003c/sup\u003e\u003c/a\u003e, e.g. being responsible for 25% of the global land carbon sink and storing approximately 120 petagrams of carbon\u003ca href=\"https://www.zotero.org/google-docs/?avRug3\"\u003e\u003csup\u003e2,5\u003c/sup\u003e\u003c/a\u003e, a large fraction of which is in aboveground tree biomass\u003ca href=\"https://www.zotero.org/google-docs/?EYp575\"\u003e\u003csup\u003e6\u003c/sup\u003e\u003c/a\u003e. Intact forests\u0026rsquo; critical contribution to climate change mitigation has been frequently acknowledged as they are among the most carbon dense\u003ca href=\"https://www.zotero.org/google-docs/?4DlQEz\"\u003e\u003csup\u003e7\u003c/sup\u003e\u003c/a\u003e and productive\u003ca href=\"https://www.zotero.org/google-docs/?GeXMAK\"\u003e\u003csup\u003e8\u003c/sup\u003e\u003c/a\u003e ecosystems on Earth. They are also critical biodiversity hotspots\u003ca href=\"https://www.zotero.org/google-docs/?049sjd\"\u003e\u003csup\u003e9\u003c/sup\u003e\u003c/a\u003e.\u003c/p\u003e\n\u003cp\u003eHowever, Amazon intact forests are under pressure\u003ca href=\"https://www.zotero.org/google-docs/?GDXsVk\"\u003e\u003csup\u003e10\u003c/sup\u003e\u003c/a\u003e and have experienced a persistent decline in their sink capacity in the past decades\u003ca href=\"https://www.zotero.org/google-docs/?wRd2QW\"\u003e\u003csup\u003e11\u003c/sup\u003e\u003c/a\u003e, potentially attributable to drought-induced increases in tree mortality\u003ca href=\"https://www.zotero.org/google-docs/?0xogaW\"\u003e\u003csup\u003e12,13\u003c/sup\u003e\u003c/a\u003e. Climate change, including the increasing intensity and severity of extreme events, may play an important role in this progressive shift of intact Amazon forests from a carbon sink to a carbon source, as extreme productivity and mortality anomalies notably occur during or directly after extremely hot and dry events such as the large-scale drought of 2015/2016\u003ca href=\"https://www.zotero.org/google-docs/?HLp6es\"\u003e\u003csup\u003e14,15\u003c/sup\u003e\u003c/a\u003e. In 2023, the Amazon was struck by an unprecedented drought event and extreme heat wave which started in the dry season\u003ca href=\"https://www.zotero.org/google-docs/?mojlJL\"\u003e\u003csup\u003e16\u003c/sup\u003e\u003c/a\u003e, of which the ecological consequences on tropical forests are yet to be determined.\u003c/p\u003e\n\u003cp\u003eMultiple methods exist to monitor the productivity of tropical forests\u003ca href=\"https://www.zotero.org/google-docs/?TH27vN\"\u003e\u003csup\u003e17\u003c/sup\u003e\u003c/a\u003e. Plot (re-)censusing is the most direct way to coincidently measure forest woody productivity and mortality\u003ca href=\"https://www.zotero.org/google-docs/?zbWUVA\"\u003e\u003csup\u003e17\u003c/sup\u003e\u003c/a\u003e but requires extensive networks of large plots to be accurately representative of a vast biome like the Amazon rainforests, which makes this technique particularly labour-intensive and time-consuming. In addition, plot recensusing has typically a low periodicity (e.g. 5 years) which prevents observations of inter-annual variability\u003ca href=\"https://www.zotero.org/google-docs/?Oh4wLN\"\u003e\u003csup\u003e14\u003c/sup\u003e\u003c/a\u003e. Eddy-covariance flux towers directly measure in situ the net CO\u003csub\u003e2\u003c/sub\u003e exchange between the atmosphere and the ecosystem, but their spatial and temporal coverage in the Tropics is limited due to the cost and the technical difficulties of their implementation and maintenance. Remote sensing\u003ca href=\"https://www.zotero.org/google-docs/?7iOCtY\"\u003e\u003csup\u003e18\u003c/sup\u003e\u003c/a\u003e (RS), inverse methods\u003ca href=\"https://www.zotero.org/google-docs/?zoXAWZ\"\u003e\u003csup\u003e19,20\u003c/sup\u003e\u003c/a\u003e, land data assimilation\u003ca href=\"https://www.zotero.org/google-docs/?wUQ5eB\"\u003e\u003csup\u003e21\u003c/sup\u003e\u003c/a\u003e, and dynamic global vegetation models\u003ca href=\"https://www.zotero.org/google-docs/?RGMRNB\"\u003e\u003csup\u003e20\u003c/sup\u003e\u003c/a\u003e (DGVM) can provide large-scale estimates of land carbon fluxes.\u003c/p\u003e\n\u003cp\u003eWhile more and more remote sensing products are made available close to real time, significant delays remain for new model output release due to the lengthy process of new data acquisition, processing, computing, and formatting even when no new developments are required. This delay is further exacerbated when multiple RS products or DGVMs are combined in ensembles to estimate uncertainties and produce an ensemble mean likely closer to the ground truth\u003ca href=\"https://www.zotero.org/google-docs/?IIQeEq\"\u003e\u003csup\u003e22\u003c/sup\u003e\u003c/a\u003e. Those technical workflows typically translate to months or years of lag between the occurrence of a climate event like the 2023 drought and the quantification of its ecological impact by DGVMs. This hinders near real time assessment of extreme events such as the extreme climatic conditions experienced in 2023/2024, which is necessary for early-warning systems to inform policy-makers, accelerate research and make it more relevant to society\u003ca href=\"https://www.zotero.org/google-docs/?KqavzP\"\u003e\u003csup\u003e23\u003c/sup\u003e\u003c/a\u003e.\u003c/p\u003e\n\u003cp\u003eHere, we used a new near-real time approach to achieve a very fast evaluation of large-scale impacts of climate events. We trained advanced gradient-boosted trees \u0026nbsp; (GBT) models\u003ca href=\"https://www.zotero.org/google-docs/?qdCAmU\"\u003e\u003csup\u003e24\u003c/sup\u003e\u003c/a\u003e to emulate multiple RS products and DGVM estimates of gross primary productivity (GPP) using climatic variables from state-of-the-art reanalyses (ERA5, JRA-55) as explanatory features. We focused our analysis on intact forests as defined by Potapov et al. (2008)\u003ca href=\"https://www.zotero.org/google-docs/?rOygcQ\"\u003e\u003csup\u003e25\u003c/sup\u003e\u003c/a\u003e in the Amazon basin\u003ca href=\"https://www.zotero.org/google-docs/?TMQ93Y\"\u003e\u003csup\u003e26\u003c/sup\u003e\u003c/a\u003e and used recent DGVM simulations made for the global carbon assessment (TRENDY\u003ca href=\"https://www.zotero.org/google-docs/?gIVj4N\"\u003e\u003csup\u003e27\u003c/sup\u003e\u003c/a\u003e) as training data. The DGVM outputs used for training were only available until 2021 but the GBT models allowed us to extend the time series until near real-time (May 2024), revealing the ecological consequences of the 2023 El-Ni\u0026ntilde;o year for the Amazon rainforests in near real-time. This opens new avenues for a broad range of vegetation ecology applications, including early warning signalling systems.\u003c/p\u003e"},{"header":"Main","content":"\u003ch2\u003eThe Amazon drought of 2023/2024 for intact Amazon forests\u003c/h2\u003e\n\u003cp\u003eAccording to ERA5 re-analyses, the period between July 2023 and April 2024 was the hottest in at least the past 30 years over intact Amazon forests (see methods for \u0026nbsp; definition and mask of intact forests of the Amazon) with an average air temperature of 26.7\u0026deg;C (Figure 1), 1.4\u0026deg;C above the the past 30 years average (25.3\u0026deg;C) and 1.3\u0026deg;C above the average value for the July-May window of the last three decades (25.4\u0026deg;C). July 2023 - April 2024 was on average 0.5\u0026deg;C and 0.1\u0026deg;C above the El Ni\u0026ntilde;o related drought events of September 1997 - April 1998 (26.2\u0026deg;C) and August 2015 - March 2016 (26.6\u0026deg;C). Over the 10-month time period between July 2023 and April 2024, the precipitation was also below average with a monthly mean of 142 mm month\u003csup\u003e-1\u003c/sup\u003e, which is 47 mm month\u003csup\u003e-1\u003c/sup\u003e below its average for that period of the year (Figure 1a and c, supplementary Figure S1). The precipitation anomaly was larger than during previous drought events, which reached -30 mm month\u003csup\u003e-1\u003c/sup\u003e between August 2015 and March 2016 and -29 mm month\u003csup\u003e-1\u003c/sup\u003e between September 1997 and April 1998 (Figure 1a and c). The hot and dry episode of 2023-2024 was also characterised by exceptionally high monthly minimum (23.1\u0026deg;C) and maximum (31.9\u0026deg;C) temperature (0.9\u0026deg;C and 1.8\u0026deg;C, respectively, above the 1994-2023 averages).\u003c/p\u003e\n\u003cp\u003eIn October 2023, the monthly air temperature reached its peak with an average 27.9\u0026deg;C, 1.8\u0026deg;C above the monthly mean of the last three decades (Figure 1b and d). This 1.8\u0026deg;C anomaly corresponded to 3.4 standard deviations of the monthly mean temperature (supplementary Figure S1). In September 2023, the precipitation dropped to an at least 30-years time low with 78 mm, 91 mm below (or 3.2 monthly standard deviations) its monthly average. Both temperature and precipitation anomalies were widespread throughout the Amazon biome (supplementary Figure S2).\u003c/p\u003e\n\u003ch2\u003eThe impact of drought on intact Amazon forest productivity\u003c/h2\u003e\n\u003cp\u003eGBT models trained on RS products and TRENDY simulations could successfully reproduce the seasonality, and the long-term trends, both at the regional level and at the local scale, for each and every RS product and DVGM (supplementary Figures S3). GBT models could also accurately predict the impacts of an unprecedented drought like 2015-2016 on Gross Primary Productivity (GPP), when trained with pre-2015 data only.\u003c/p\u003e\n\u003cp\u003eWhile GPP has steadily increased over the last three decades according to both RS-trained and TRENDY-trained GBT models (p-value \u0026lt; 0.001 for the slope of the linear models in Figure 2a), it showed noticeable negative anomalies during the past (1997-1998, 2015-2016) and recent (2023-2024) dry and hot episodes (Figure 2b). Since July 2023, GBT models forecasted consistently lower GPP compared to the average of the past three decades (Figure 2a and b). Between July 2023 and April 2024, GBT models projected a GPP reduction of 1.2 Mg C ha\u003csup\u003e-1\u003c/sup\u003e year\u003csup\u003e-1\u003c/sup\u003e (RS-trained GBT models) and 1.5 Mg C ha\u003csup\u003e-1\u003c/sup\u003e year\u003csup\u003e-1\u003c/sup\u003e (TRENDY-trained GBT models) compared to the seasonal average for all intact forests in the Amazon, which corresponded to an average decline of 2.6 and 2.2 monthly standard deviations (Figure 2b, supplementary Figure S5). The productivity of 2023-2024 reached its lowest level in October 2023 with RS and TRENDY trained GBT models predicting a respective mean GPP value of 30.7 and 27.9 Mg C ha\u003csup\u003e-1\u003c/sup\u003e year\u003csup\u003e-1\u003c/sup\u003e, respectively 2.1 and 3.4 Mg C ha\u003csup\u003e-1\u003c/sup\u003e year\u003csup\u003e-1\u003c/sup\u003e lower than their expected value in the absence of drought. It was the least productive October of the past 30 years according to both RS and TRENDY models. This sharp decrease corresponds to a reduction equivalent to 3.9 and 3.4 monthly standard deviations, a three-decades minimum only surpassed during the month of January 2016 for TRENDY-trained GBT models (3.6 monthly standard deviations then, but for a smaller absolute GPP reduction: 2.2 Mg C ha\u003csup\u003e-1\u003c/sup\u003e year\u003csup\u003e-1\u003c/sup\u003e in January 2016).\u003c/p\u003e\n\u003cp\u003eNegative anomalies were distributed all across the Amazon during the 2023-2024 dry and hot episode: 98.2% (RS models) or 98.6% (TRENDY models) of all intact Amazon forest grid cells experienced negative GPP anomalies, on average, over the July 2023 - April2024 period (Figure 3). It was the most spatially widespread drought anomaly, surpassing the 1997-1998 drought according to both RS- and TRENDY-trained GBT models (respectively 97.7% and 96.3% of grid cells had negative anomalies then). During the month of October 2023, 78.8%/62.0% (RS-trained GBT models) or 49.3%/42.3% (TRENDY-trained GBT models) of all intact forest grid cells in the Amazon had a negative GPP anomaly corresponding to at least two/three monthly standard deviations.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eERA5 mean temperatures and precipitation, as well as their anomalies, showed strong correlations with those of JRA-55 which also marked October 2023 exceptionally dry and hot (supplementary Figure S3). Hence, repeating the analyses with JRA-55 as climate drivers of the GBT models led to very similar conclusions about both the magnitude of the 2023 dry and hot episode and its spatial distribution (supplementary Figures S6 and S7).\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eClimate-GPP relationships of the intact Amazon forests\u003c/h2\u003e\n\u003cp\u003eBoth RS- and TRENDY-trained GBT models exhibited significant, nonlinear relationships between monthly GPP and mean temperature anomalies (p-value \u0026lt; 0.001, Figure 4), as well as with VPD and precipitation anomalies (supplementary Figure S8). In the next 30 years (2024-2053), air temperature is expected to increase almost linearly over the study area, with warming rates varying between 0.19\u0026deg;C decade\u003csup\u003e-1\u003c/sup\u003e (SSP1-2.6) and 0.44\u0026deg;C decade\u003csup\u003e-1\u003c/sup\u003e (SSP5-8.5), according to a weighted average of CMIP6 model simulation made to reproduce ERA5 temperatures (supplementary Figures S9). In the next three decades, temperature anomalies similar to the 2023-2024 dry and hot episodes are expected to become common for all climatic scenarios (Figure 4c). By the end of this century, such temperature anomalies are expected to remain the norm for SSP1-2.6 and become cold outliers for all other scenarios (Supplementary Figure S10). These future high temperatures will generate correspondingly high VPD while precipitation is expected to decrease by the end of this century (Supplementary Figure S10), which should further aggravate GPP anomalies.\u0026nbsp;\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eAccording to the novel model projection tool that we developed, the 2023-2024 dry and hot atmospheric conditions will lead to exceptional ecological consequences for the intact forests of the Amazon. Driven by abnormally high temperatures and precipitation deficits, intact forests\u0026rsquo; GPP has likely hit a record low anomaly in the month of October 2023. This event could reverberate in the upcoming months and years as drought has been unequivocally linked to tree mortality\u003ca href=\"https://www.zotero.org/google-docs/?1dgMLZ\"\u003e\u003csup\u003e13,28\u0026ndash;31\u003c/sup\u003e\u003c/a\u003e. TRENDY (Supplementary Figure S11) and other models\u003ca href=\"https://www.zotero.org/google-docs/?Zs88BY\"\u003e\u003csup\u003e32\u003c/sup\u003e\u003c/a\u003e, in line with field inventory data\u003ca href=\"https://www.zotero.org/google-docs/?oBrUHm\"\u003e\u003csup\u003e14,33\u003c/sup\u003e\u003c/a\u003e, predict significant reductions of the intact Amazon forest carbon sink for previous large-scale drought events, which is likely to repeat given the 2023-2024 drought magnitude.\u003c/p\u003e\n\u003cp\u003eRegardless of the future emission scenario, temperatures are expected to increase in the next three decades and Amazon intact forest GPP will likely be affected by this temperature increase, the related increase in VPD, and the possible reduction in precipitation. Hence, the CO\u003csub\u003e2\u003c/sub\u003e fertilisation\u003ca href=\"https://www.zotero.org/google-docs/?Tr21Qy\"\u003e\u003csup\u003e34\u003c/sup\u003e\u003c/a\u003e, also observed here in the RS- and TRENDY-trained GBT models, might be outpaced in the near future by the future increase in frequency and severity of droughts, even in the absence of CO\u003csub\u003e2\u003c/sub\u003e saturation for photosynthesis\u003ca href=\"https://www.zotero.org/google-docs/?rcP03x\"\u003e\u003csup\u003e35\u003c/sup\u003e\u003c/a\u003e. This applies to GPP and potentially to net ecosystem productivity (NEP), which has shown no increasing trends (p-value 0.19) and strong drought negative responses, according to both TRENDY models (Supplementary Figure S12) and field data\u003ca href=\"https://www.zotero.org/google-docs/?tBEtb2\"\u003e\u003csup\u003e12\u003c/sup\u003e\u003c/a\u003e.\u003c/p\u003e\n\u003cp\u003eThe new method presented in this article is highly flexible, robust (supplementary Figure S3), fast, and easy to use. The models were trained on a specific region but could easily be extended to include more (human-disturbed) biomes, RS products or DGVMs, features (e.g., including sub-monthly variability such as climate extremes as explanatory variables or more features), or even novel types of data (e.g., eddy covariance fluxes, field inventories, or soil moisture observations that were shown to critically impact land carbon fluxes\u003ca href=\"https://www.zotero.org/google-docs/?d7myrx\"\u003e\u003csup\u003e36\u003c/sup\u003e\u003c/a\u003e). After training, the GBT models require no expert knowledge and only need the location (latitudes and longitudes), time (years and months), and climate of the region of interest to simulate GPP. It took less than one second on a regular 8-cores laptop with 16 GB of RAM to run any of the RS or TRENDY surrogates for a full year for the entire intact Amazon forests, contrasting with the large (but very model-specific) number of CPU-hours required to run a typical TRENDY model .\u003c/p\u003e\n\u003cp\u003eDespite these advantages, it is unclear whether the models developed in this study can accurately predict land fluxes under unprecedented climatic circumstances. On the one hand, the emulators were able to reproduce the extreme 2015-2016 dry and hot episode once trained on earlier GPP estimates (supplementary Figure S4). On the other hand, Amazon rainforest tipping points have been postulated\u003ca href=\"https://www.zotero.org/google-docs/?elNw8H\"\u003e\u003csup\u003e10,37\u0026ndash;39\u003c/sup\u003e\u003c/a\u003e if drought intensity and/or frequency increases beyond some threshold, but so far such threshold responses have not been observed at large scales and hence are not part of the training data. As new RS products or model outputs are released, our GBT model projections can be confirmed or if needed, those newly available data, as well as other covariates, can feed the GBT models as additional training data to improve their robustness. Hybrid AI\u003ca href=\"https://www.zotero.org/google-docs/?HzWbGu\"\u003e\u003csup\u003e40\u003c/sup\u003e\u003c/a\u003e, as well as DGVM future projection\u003ca href=\"https://www.zotero.org/google-docs/?IxQnyY\"\u003e\u003csup\u003e41\u003c/sup\u003e\u003c/a\u003e under hotter and drier conditions, will help overcome this potential issue.\u003c/p\u003e\n\u003cp\u003eThe methodology suggested here opens the door for a wide range of other potential applications. Because climate variables needed to force those GBT models are widely available including for the future, it makes the development of near-real time ecological forecasting\u003ca href=\"https://www.zotero.org/google-docs/?qsi7fK\"\u003e\u003csup\u003e23\u003c/sup\u003e\u003c/a\u003e or early warning systems\u003ca href=\"https://www.zotero.org/google-docs/?CaSFu4\"\u003e\u003csup\u003e42\u003c/sup\u003e\u003c/a\u003e for land fluxes based on RS and DGVM possible. Such a tool could become incredibly useful for decision-makers and conservation policies overall. Such approaches can also help identify (the reasons behind) discrepancies between DVGM outputs and RS products, prioritise necessary model development\u003ca href=\"https://www.zotero.org/google-docs/?xNggPT\"\u003e\u003csup\u003e43\u003c/sup\u003e\u003c/a\u003e or refine part of RS algorithms for land flux retrieval.\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch1\u003eAcknowledgements\u003c/h1\u003e\n\u003cp\u003eFM was funded by the FWO as a senior postdoc and is thankful to this organisation for its financial support (FWO grant no.1214723N). Michael C. Dietze was supported by funding from NSF Grant 1638577 and NASA Carbon Monitoring System (80NSSC17K0711). SDH was supported by the BELSPO project DAMOCO (B2/223/P1/DAMOCO). MP acknowledges the financial support of the European Research Council (grant No. 101117001 LEAFPACE).\u003c/p\u003e"},{"header":"References","content":"\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e1.\u0026nbsp; \u0026nbsp;\u0026nbsp;Hoang, N. T. \u0026amp; Kanemoto, K. Mapping the deforestation footprint of nations reveals growing threat to tropical forests.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNat. Ecol. Evol.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e5\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 845\u0026ndash;853 (2021).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e2.\u0026nbsp; \u0026nbsp;\u0026nbsp;Saatchi, S. S.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Benchmark map of forest carbon stocks in tropical regions across three continents.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eProc. Natl. Acad. Sci.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e108\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 9899\u0026ndash;9904 (2011).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e3.\u0026nbsp; \u0026nbsp;\u0026nbsp;Schulze, K., Malek, Ž. \u0026amp; Verburg, P. H. Towards better mapping of forest management patterns: A global allocation approach.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eFor. Ecol. Manag.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e432\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 776\u0026ndash;785 (2019).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e4.\u0026nbsp; \u0026nbsp;\u0026nbsp;Lewis, S. L., Edwards, D. P. \u0026amp; Galbraith, D. Increasing human dominance of tropical forests.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eScience\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e349\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 827\u0026ndash;832 (2015).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e5.\u0026nbsp; \u0026nbsp;\u0026nbsp;Malhi, Y.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;The regional variation of aboveground live biomass in old-growth Amazonian forests.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eGlob. Change Biol.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e12\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 1107\u0026ndash;1138 (2006).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e6.\u0026nbsp; \u0026nbsp;\u0026nbsp;Feldpausch, T. R.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Tree height integrated into pantropical forest biomass estimates.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eBiogeosciences\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e9\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 3381\u0026ndash;3403 (2012).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e7.\u0026nbsp; \u0026nbsp;\u0026nbsp;Spawn, S. A., Sullivan, C. C., Lark, T. J. \u0026amp; Gibbs, H. K. Harmonized global maps of above and belowground biomass carbon density in the year 2010.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eSci. Data\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e7\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 112 (2020).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e8.\u0026nbsp; \u0026nbsp;\u0026nbsp;Li, X. \u0026amp; Xiao, J. A Global, 0.05-Degree Product of Solar-Induced Chlorophyll Fluorescence Derived from OCO-2, MODIS, and Reanalysis Data.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eRemote Sens.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e11\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 517 (2019).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e9.\u0026nbsp; \u0026nbsp;\u0026nbsp;de Lima, R. A. F.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;The erosion of biodiversity and biomass in the Atlantic Forest biodiversity hotspot.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNat. Commun.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e11\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 6347 (2020).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e10.\u0026nbsp;\u0026nbsp;Flores, B. M.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Critical transitions in the Amazon forest system.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNature\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e626\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 555\u0026ndash;564 (2024).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e11.\u0026nbsp;\u0026nbsp;Brienen, R. J. W.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Long-term decline of the Amazon carbon sink.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNature\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e519\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 344\u0026ndash;8 (2015).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e12.\u0026nbsp;\u0026nbsp;Hubau, W.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Asynchronous carbon sink saturation in African and Amazonian tropical forests.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNature\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e579\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 80\u0026ndash;87 (2020).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e13.\u0026nbsp;\u0026nbsp;Tavares, J. V.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Basin-wide variation in tree hydraulic safety margins predicts the carbon balance of Amazon forests.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNature\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e617\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 111\u0026ndash;117 (2023).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e14.\u0026nbsp;\u0026nbsp;Bennett, A. C.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Sensitivity of South American tropical forests to an extreme climate anomaly.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNat. Clim. Change\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e13\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 967\u0026ndash;974 (2023).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e15.\u0026nbsp;\u0026nbsp;Jim\u0026eacute;nez-Mu\u0026ntilde;oz, J. C.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Record-breaking warming and extreme drought in the Amazon rainforest during the course of El Ni\u0026ntilde;o 2015\u0026ndash;2016.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eSci. Rep.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e6\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 33130 (2016).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e16.\u0026nbsp;\u0026nbsp;Wagner, F. H.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Amazon\u0026rsquo;s 2023 Drought: Sentinel-1 Reveals Extreme Rio Negro River Contraction. Preprint at http://arxiv.org/abs/2401.16393 (2024).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e17.\u0026nbsp;\u0026nbsp;ForestPlots.net\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Taking the pulse of Earth\u0026rsquo;s tropical forests using networks of highly distributed plots.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eBiol. Conserv.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e260\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 108849 (2021).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e18.\u0026nbsp;\u0026nbsp;Heinrich, V. H. A.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Large carbon sink potential of secondary forests in the Brazilian Amazon to mitigate climate change.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNat. Commun.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e12\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 1785 (2021).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e19.\u0026nbsp;\u0026nbsp;Frankenberg, C.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al. Data Drought in the Humid Tropics: How to Overcome the Cloud Barrier in Greenhouse Gas Remote Sensing\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e. https://essopenarchive.org/users/528524/articles/717473-data-drought-in-the-humid-tropics-how-to-overcome-the-cloud-barrier-in-greenhouse-gas-remote-sensing?commit=c1c370fe625963945ecaea0faa78b3b911e10e05 (2024) doi:10.22541/essoar.170923255.57545328/v1.\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e20.\u0026nbsp;\u0026nbsp;Rosan, T. M.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Synthesis of the land carbon fluxes of the Amazon region between 2010 and 2020.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eCommun. Earth Environ.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e5\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 1\u0026ndash;15 (2024).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e21.\u0026nbsp;\u0026nbsp;Dokoohaki, H.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Development of an open-source regional data assimilation system in PEcAn v. 1.7.2: application to carbon cycle reanalysis across the contiguous US using SIPNET.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eGeosci. Model Dev.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e15\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 3233\u0026ndash;3252 (2022).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e22.\u0026nbsp;\u0026nbsp;Li, L.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;The optimization of model ensemble composition and size can enhance the robustness of crop yield projections.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eCommun. Earth Environ.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e4\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 1\u0026ndash;11 (2023).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e23.\u0026nbsp;\u0026nbsp;Dietze, M. C.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Iterative near-term ecological forecasting: Needs, opportunities, and challenges.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eProc. Natl. Acad. Sci.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e115\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 1424\u0026ndash;1432 (2018).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e24.\u0026nbsp;\u0026nbsp;Greener, J. G., Kandathil, S. M., Moffat, L. \u0026amp; Jones, D. T. A guide to machine learning for biologists.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNat. Rev. Mol. Cell Biol.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e23\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 40\u0026ndash;55 (2022).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e25.\u0026nbsp;\u0026nbsp;Potapov, P.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;The last frontiers of wilderness: Tracking loss of intact forest landscapes from 2000 to 2013.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eSci. Adv.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, e1600821.\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e26.\u0026nbsp;\u0026nbsp;Boulton, C. A., Lenton, T. M. \u0026amp; Boers, N. Pronounced loss of Amazon rainforest resilience since the early 2000s.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNat. Clim. Change\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e12\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 271\u0026ndash;278 (2022).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e27.\u0026nbsp;\u0026nbsp;Friedlingstein, P.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Global Carbon Budget 2023.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eEarth Syst. Sci. Data\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e15\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 5301\u0026ndash;5369 (2023).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e28.\u0026nbsp;\u0026nbsp;Williamson, G. B.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Amazonian Tree Mortality during the 1997 El Ni\u0026ntilde;o Drought.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eConserv. Biol.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e14\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 1538\u0026ndash;1542 (2000).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e29.\u0026nbsp;\u0026nbsp;Yao, Y.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Forest fluxes and mortality response to drought: model description (ORCHIDEE-CAN-NHA, r7236) and evaluation at the Caxiuan\u0026amp;atilde; drought experiment.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eGeosci. Model Dev. Discuss.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;1\u0026ndash;38 (2021) doi:10.5194/gmd-2021-362.\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e30.\u0026nbsp;\u0026nbsp;Rowland, L.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Death from drought in tropical forests is triggered by hydraulics not carbon starvation.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNature\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e528\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 119\u0026ndash;122 (2015).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e31.\u0026nbsp;\u0026nbsp;McDowell, N. G. Mechanisms Linking Drought, Hydraulics, Carbon Metabolism, and Vegetation Mortality.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003ePlant Physiol.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e155\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 1051\u0026ndash;1059 (2011).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e32.\u0026nbsp;\u0026nbsp;Tian, H.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Effect of interannual climate variability on carbon storage in Amazonian ecosystems.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNature\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e396\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 664\u0026ndash;667 (1998).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e33.\u0026nbsp;\u0026nbsp;Feldpausch, T. R.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Amazon forest response to repeated droughts.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eGlob. Biogeochem. Cycles\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e30\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 964\u0026ndash;982 (2016).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e34.\u0026nbsp;\u0026nbsp;Chen, C., Riley, W. J., Prentice, I. C. \u0026amp; Keenan, T. F. CO2 fertilization of terrestrial photosynthesis inferred from site to global scales.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eProc. Natl. Acad. Sci.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e119\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, e2115627119 (2022).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e35.\u0026nbsp;\u0026nbsp;Wang, S.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Recent global decline of CO2 fertilization effects on vegetation photosynthesis.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eScience\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e370\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 1295\u0026ndash;1300 (2020).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e36.\u0026nbsp;\u0026nbsp;Yang, H.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Climatic and biotic factors influencing regional declines and recovery of tropical forest biomass from the 2015/16 El Ni\u0026ntilde;o.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eProc. Natl. Acad. Sci.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e119\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, e2101388119 (2022).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e37.\u0026nbsp;\u0026nbsp;Cox, P. M.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Amazonian forest dieback under climate-carbon cycle projections for the 21st century.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eTheor. Appl. Climatol.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e78\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 137\u0026ndash;156 (2004).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e38.\u0026nbsp;\u0026nbsp;Parry, I., Ritchie, P. \u0026amp; Cox, P.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eEvidence of Amazon Rainforest Dieback in CMIP6 Models\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e. (2022). doi:10.5194/egusphere-2022-82.\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e39.\u0026nbsp;\u0026nbsp;Lovejoy, T. E. \u0026amp; Nobre, C. Amazon Tipping Point.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eSci. Adv.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e4\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, eaat2340 (2018).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e40.\u0026nbsp;\u0026nbsp;Koppa, A., Rains, D., Hulsman, P., Poyatos, R. \u0026amp; Miralles, D. G. A deep learning-based hybrid model of global terrestrial evaporation.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eNat. Commun.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e13\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 1912 (2022).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e41.\u0026nbsp;\u0026nbsp;Frieler, K.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Scenario setup and forcing data for impact model evaluation and impact attribution within the third round of the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP3a).\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eGeosci. Model Dev.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e17\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 1\u0026ndash;51 (2024).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e42.\u0026nbsp;\u0026nbsp;Boulton, C. A., Good, P. \u0026amp; Lenton, T. M. Early warning signals of simulated Amazon rainforest dieback.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eTheor. Ecol.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e6\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 373\u0026ndash;384 (2013).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e43.\u0026nbsp;\u0026nbsp;Fisher, R. A. \u0026amp; Koven, C. D. Perspectives on the Future of Land Surface Models and the Challenges of Representing Complex Terrestrial Systems.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eJ. Adv. Model. Earth Syst.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e12\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, e2018MS001453 (2020).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e44.\u0026nbsp;\u0026nbsp;Harris, I., Osborn, T. J., Jones, P. \u0026amp; Lister, D. Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eSci. Data\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e7\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 109 (2020).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e45.\u0026nbsp;\u0026nbsp;Harris, I., Jones, P. d., Osborn, T. j. \u0026amp; Lister, D. h. Updated high-resolution grids of monthly climatic observations \u0026ndash; the CRU TS3.10 Dataset.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eInt. J. Climatol.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e34\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 623\u0026ndash;642 (2014).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e46.\u0026nbsp;\u0026nbsp;Kobayashi, S.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;The JRA-55 Reanalysis: General Specifications and Basic Characteristics.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eJ. Meteorol. Soc. Jpn. Ser II\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e93\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 5\u0026ndash;48 (2015).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e47.\u0026nbsp;\u0026nbsp;Bi, W.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;A global 0.05\u0026deg; dataset for gross primary production of sunlit and shaded vegetation canopies from 1992 to 2020.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eSci. Data\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e9\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 213 (2022).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e48.\u0026nbsp;\u0026nbsp;Wild, B.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;VODCA2GPP \u0026ndash; a new, global, long-term (1988\u0026ndash;2020) gross primary production dataset from microwave remote sensing.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eEarth Syst. Sci. Data\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e14\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 1063\u0026ndash;1085 (2022).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e49.\u0026nbsp;\u0026nbsp;Wang, S., Zhang, Y., Ju, W., Qiu, B. \u0026amp; Zhang, Z. Tracking the seasonal and inter-annual variations of global gross primary production during last four decades using satellite near-infrared reflectance data.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eSci. Total Environ.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e755\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 142569 (2021).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e50.\u0026nbsp;\u0026nbsp;Pedelty, J.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Generating a long-term land data record from the AVHRR and MODIS Instruments. in\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003e2007 IEEE International Geoscience and Remote Sensing Symposium\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;1021\u0026ndash;1025 (IEEE, Barcelona, Spain, 2007). doi:10.1109/IGARSS.2007.4422974.\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e51.\u0026nbsp;\u0026nbsp;Chen, T. \u0026amp; Guestrin, C. XGBoost: A Scalable Tree Boosting System. in\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eProceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;785\u0026ndash;794 (ACM, San Francisco California USA, 2016). doi:10.1145/2939672.2939785.\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e52.\u0026nbsp;\u0026nbsp;Chen, T.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;xgboost: Extreme Gradient Boosting. (2024).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e53.\u0026nbsp;\u0026nbsp;A Short Introduction to the caret Package. https://cran.r-project.org/web/packages/caret/vignettes/caret.html.\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e54.\u0026nbsp;\u0026nbsp;PEcAn functions used for managing climate driver data. https://pecanproject.github.io/modules/data.atmosphere/docs/index.html.\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e55.\u0026nbsp;\u0026nbsp;Hijmans, R. J.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;raster: Geographic Data Analysis and Modeling. (2023).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e56.\u0026nbsp;\u0026nbsp;Eyring, V.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eGeosci. Model Dev.\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e9\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 1937\u0026ndash;1958 (2016).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e57.\u0026nbsp;\u0026nbsp;Riahi, K.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eet al.\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;The Shared Socioeconomic Pathways and their energy, land use, and greenhouse gas emissions implications: An overview.\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eGlob. Environ. Change\u0026nbsp;\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cstrong\u003e42\u003c/strong\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e, 153\u0026ndash;168 (2017).\u003c/a\u003e\u003c/p\u003e\n\u003cp\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e58.\u0026nbsp;\u0026nbsp;Vinutha, H. P., Poornima, B. \u0026amp; Sagar, B. M. Detection of Outliers Using Interquartile Range Technique from Intrusion Dataset. in\u0026nbsp;\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u003cem\u003eInformation and Decision Sciences\u003c/em\u003e\u003c/a\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WXQjmi\"\u003e\u0026nbsp;(eds. Satapathy, S. C., Tavares, J. M. R. S., Bhateja, V. \u0026amp; Mohanty, J. R.) 511\u0026ndash;518 (Springer, Singapore, 2018). doi:10.1007/978-981-10-7563-6_53.\u003c/a\u003e\u003c/p\u003e"},{"header":"Methods","content":"\u003ch2\u003eStudy area\u003c/h2\u003e\n\u003cp\u003eThe Amazon basin (\u003ca href=\"https://github.com/gamamo/AmazonBasinLimits\"\u003ehttps://github.com/gamamo/AmazonBasinLimits\u003c/a\u003e) delimited our study area\u003ca href=\"https://www.zotero.org/google-docs/?U5jtZz\"\u003e\u003csup\u003e26\u003c/sup\u003e\u003c/a\u003e. We further restricted the analyses to intact forests using the intact forest map of Potapov et al. (2008)\u003ca href=\"https://www.zotero.org/google-docs/?SsHqi2\"\u003e\u003csup\u003e25\u003c/sup\u003e\u003c/a\u003e (\u003ca href=\"http://www.intactforests.org/data.ifl.html\"\u003ehttp://www.intactforests.org/data.ifl.html\u003c/a\u003e) for the year 2020.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eTRENDY model simulations\u003c/h2\u003e\n\u003cp\u003eWe used simulations of a set of 16 Dynamic Global Vegetation Models (DGVMs), which participated in the past global carbon budgets assessment\u003ca href=\"https://www.zotero.org/google-docs/?kD3tXm\"\u003e\u003csup\u003e27\u003c/sup\u003e\u003c/a\u003e, named TRENDY-v11. Because we focus on intact forests, we extracted model outputs from TRENDY-v11 scenario S2 in which CO\u003csub\u003e2\u003c/sub\u003e and climate vary with time but which uses a time-invariant \u0026ldquo;pre-industrial\u0026rdquo; land use mask. In TRENDY, all models are forced with the same climate forcing, namely CRUJRA, see\u003ca href=\"https://www.zotero.org/google-docs/?cdRnXV\"\u003e\u003csup\u003e44\u0026ndash;46\u003c/sup\u003e\u003c/a\u003e. We mainly extracted GPP monthly outputs from TRENDY-v11 but we also pulled monthly averages of NPP and heterotrophic respiration to compute NEP from their difference for every model that provided those variables as well. NEP values were only used as is (i.e. did not serve for GBT model training) to investigate their correlation with GPP. TRENDY-v11 model outputs are available for the time period 1901-2021.\u003c/p\u003e\n\u003ch2\u003eRemote sensing estimates\u003c/h2\u003e\n\u003cp\u003eTo benchmark TRENDY model outputs, we used four recent GPP products, hereafter referred to as Li and Xiao (2019)\u003ca href=\"https://www.zotero.org/google-docs/?UHfnIA\"\u003e\u003csup\u003e8\u003c/sup\u003e\u003c/a\u003e, Bi et al. (2022)\u003ca href=\"https://www.zotero.org/google-docs/?WqYjZL\"\u003e\u003csup\u003e47\u003c/sup\u003e\u003c/a\u003e, Wild et al. (2022)\u003ca href=\"https://www.zotero.org/google-docs/?7gIrcz\"\u003e\u003csup\u003e48\u003c/sup\u003e\u003c/a\u003e, and Wang et al. (2021)\u003ca href=\"https://www.zotero.org/google-docs/?w31rhC\"\u003e\u003csup\u003e49\u003c/sup\u003e\u003c/a\u003e. Briefly, Li and Xiao (2019) developed a global, high temporal (8-days) and spatial (0.05\u0026deg;) resolutions GPP product based on Orbiting Carbon Observatory-2 SIF data, MODIS and meteorological re-analyses through a data-driven approach over the 2000-2022 period\u003ca href=\"https://www.zotero.org/google-docs/?uelJJH\"\u003e\u003csup\u003e8\u003c/sup\u003e\u003c/a\u003e. Bi et al. (2022) generated a global 0.05\u0026deg;, 8-day dataset for GPP (1992-2020) with a two-leaf light use efficiency model, driven by CRUJRA reanalysis, ESA-CCI land cover and leaf area index from GLOBMAP\u003ca href=\"https://www.zotero.org/google-docs/?UrALRp\"\u003e\u003csup\u003e47\u003c/sup\u003e\u003c/a\u003e. Wild et al. (2022) utilised microwave remote sensing of vegetation optical depth to derive global GPP at moderate spatial (0.25\u0026deg;) and high temporal (8-days) resolutions\u003ca href=\"https://www.zotero.org/google-docs/?7GpZfW\"\u003e\u003csup\u003e48\u003c/sup\u003e\u003c/a\u003e for the period 1988-2020. Finally, Wang et al. (2021) correlated eddy-covariance GPP and AVHRR near-infrared reflectance from the Land Long Term Data Record\u003ca href=\"https://www.zotero.org/google-docs/?bAia8D\"\u003e\u003csup\u003e50\u003c/sup\u003e\u003c/a\u003e to generate a global long-term (1982-2018) time series at high spatial resolution (0.05\u0026deg;) of monthly GPP. We aggregated all four datasets at the monthly timescale for further analyses.\u003c/p\u003e\n\u003ch2\u003eAI models\u003c/h2\u003e\n\u003cp\u003eWe used the extreme gradient boosting (XGBoost\u003ca href=\"https://www.zotero.org/google-docs/?3p29g0\"\u003e\u003csup\u003e51\u003c/sup\u003e\u003c/a\u003e) approach, as implemented in the R-package \u003cem\u003exgboost\u003c/em\u003e\u003ca href=\"https://www.zotero.org/google-docs/?FIljbp\"\u003e\u003csup\u003e52\u003c/sup\u003e\u003c/a\u003e, to reproduce time series of GPP as estimated from DGVMs or from remote sensing. XGBoost is an advanced machine learning algorithm reputed to be highly efficient in terms of model performance and computational speed. Individual GBT models were trained for each DGVM/RS estimate of GPP using the longest possible time series and the largest possible area. We splitted all the data into training (60%), validation (20%) and test (20%) for each grid cell and year. So on average, 2.4 months per year served as validation and test for each grid cell while 7.2 months per year were used for training. The months were randomly attributed to training, validation and test dataset. We optimised/tuned GBT model (hyper)parameters on validation data with the R-package \u003cem\u003ecaret\u003c/em\u003e\u003ca href=\"https://www.zotero.org/google-docs/?RDyPBR\"\u003e\u003csup\u003e53\u003c/sup\u003e\u003c/a\u003e and 8 resampling iterations, focusing on the maximum decision tree depth and the number of decision trees to grow.\u003c/p\u003e\n\u003cp\u003eTraining variables included monthly averages of climate data (monthly mean, minimum, and maximum temperature, monthly precipitation, VPD, total incoming short- and long-wave radiation), atmospheric CO\u003csub\u003e2\u003c/sub\u003e concentration (constant spatially), as well as temporal (year/month) and spatial (latitude/longitude) information. The latter variables were important to include to account for model-specific spatial variability in e.g., soil maps and plant functional type spatial distribution. We computed VPD from sub-daily air temperatures and specific humidity using the R-package \u003cem\u003ePecan.data.atmosphere\u003c/em\u003e\u003ca href=\"https://www.zotero.org/google-docs/?tp1xZY\"\u003e\u003csup\u003e54\u003c/sup\u003e\u003c/a\u003e before taking the monthly average. Similarly, we used sub-daily air temperature values to compute the daily minimum and maximum temperatures and finally calculate the monthly minimum and maximum temperature as the average of all daily minimum and maximum temperatures of a given month.\u003c/p\u003e\n\u003cp\u003eThe CRUJRA meteorological reanalysis, which is used as the driver for TRENDY models, is only updated once a year (typically between April and August), which prevents its use for near-real time forecasting. Therefore we decided to use ERA5 and JRA-55 reanalysis as input data for the GBT models. Doing so we internalised the bias between reanalysis sources into the GBT models, as CRUJRA is constructed by adjusting data from JRA-55 where possible to align with the CRU TS data. We downloaded all ERA5 and JRA-55 climatic drivers available in early June 2024, i.e. until May 2024 and December 2023 respectively.\u003c/p\u003e\n\u003cp\u003eWe performed the same model training and validation described above, independently with ERA5 and JRA-55 reanalysis to check the influence of climatic forcings on the model predictions. We also checked how absolute values and anomalies correlated for both reanalysis sources for the region of interest. For ERA5, we used the 2-metre air temperature variable for the air temperature.\u003c/p\u003e\n\u003cp\u003eFinal GBT model performance was evaluated on test data only and quantified using the root mean square deviation and the square of the Pearson correlation coefficient (R\u0026sup2;) between reference (DGVM outputs or RS estimates) and predicted values.\u003c/p\u003e\n\u003ch2\u003eAnalyses\u003c/h2\u003e\n\u003cp\u003eGiven the different resolutions of the individual TRENDY models and the remote sensing GPP estimates, we reprojected the GBT model features (climate forcings) on each DGVM/RS grid before training with the R-package \u003cem\u003eraster\u003c/em\u003e\u003ca href=\"https://www.zotero.org/google-docs/?WYN3MK\"\u003e\u003csup\u003e55\u003c/sup\u003e\u003c/a\u003e. To generate spatial ensemble means of GPP, we also reprojected all rasters to a common grid (ERA5) using bilinear interpolations.\u003c/p\u003e\n\u003cp\u003eWhen reprojection was needed for the intact forest map, we used bilinear interpolations, introducing decimal numbers. We then limited the analyses to all grid cells with interpolated values larger than 50% intact forest cover. Unless otherwise stated, we limited our analyses to the past 30 years (1994-2023).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor generating model ensemble means, we applied simple means to all RS products and DGVM outputs, giving equal weight to each ensemble member. We used the GBT models for extending RS time series to recent years (e.g., 2023/2024) but also to older times (from 1994 to the actual start of the product time series) to avoid discontinuities when computing ensemble means when not all products were available.\u003c/p\u003e\n\u003cp\u003eWe compared the recent dry and hot episode of 2023-2024 with previous major drought events, namely August 2015-March 2016, and September 1997-April 1998. We delimited those drought events of the past from their precipitation and air temperature anomalies in ERA5.\u003c/p\u003e\n\u003cp\u003eTo assess whether the GBT models were able to accurately predict GPP in case of unprecedented events, we verified how the density distribution of the residuals compared for the test data of the dry and hot episodes of 1997-1998 and 2015-2016 and all other climatic conditions. We also repeated the model training and validation described above using pre-2015 data only and tested the model performance specifically for the 2015-2016 dry and hot episode.\u003c/p\u003e\n\u003cp\u003eIn both RS estimates and DGVM simulations, GPP presents both a seasonality and an increasing long-term trend due to CO\u003csub\u003e2\u003c/sub\u003e fertilisation. We detrended the GPP time series by subtracting to the raw data the best linear model estimates from the 1994-2023 time period and its average seasonal variation. We used a linear model for detrending as GPP has been steadily increasing the past decades due to CO\u003csub\u003e2\u003c/sub\u003e fertilisation. We detrended the climate variables by subtracting the overall variable average and its seasonal mean to the raw time series. The anomalies were normalised using the monthly standard deviation of the respective variables for the overall time period (1994-2023). For spatial analyses, we applied the detrending method described above to the grid cells individually, while we first averaged the time series for biome-wide analyses. Anomalies were computed on the model ensemble mean and the average of RS products.\u003c/p\u003e\n\u003cp\u003eTo smooth monthly time series, we applied rolling averages with a centred 6-month time window. We tested the effect of this choice by changing the window size between 1 month (no average) and one year and verified how it changed the correlations between variables.\u003c/p\u003e\n\u003cp\u003eFor future trends, we downloaded monthly temperature and precipitation outputs from Earth system models (ESM) available in January 2024 for the models participating in the Coupled Model Intercomparison project 6 (CMIP6)\u003ca href=\"https://www.zotero.org/google-docs/?TP0Cif\"\u003e\u003csup\u003e56\u003c/sup\u003e\u003c/a\u003e. We downloaded model outputs for the historical period and the following shared socioeconomic pathways\u003ca href=\"https://www.zotero.org/google-docs/?nnDDNP\"\u003e\u003csup\u003e57\u003c/sup\u003e\u003c/a\u003e: SSP1-2.7, SSP2-4.5, SSP3-7.0, and SSP5-8.5. We restricted the analysis to the models which had model outputs available for every shared socioeconomic pathway. We then compared the regional monthly mean temperature and the mean annual precipitation time series predicted by each ESM for the recent historical period (1985-2014) with those of ERA5, and only kept the 10 with the lowest root mean square error for each variable. We then averaged those 10 models using the inverse of the squared bias as weight for both the historical period and the future simulations. Temperature and precipitation anomalies were defined using 1985-2014 or 1994-2023 as reference for the historical period or the future scenarios, respectively, since CMIP6 historical simulations stop in 2014.\u003c/p\u003e\n\u003cp\u003eTemperature anomaly outliers were identified following the classical outlier 1.5 interquartile (IQR) rule\u003ca href=\"https://www.zotero.org/google-docs/?8xHMuj\"\u003e\u003csup\u003e58\u003c/sup\u003e\u003c/a\u003e: any observation that was smaller than 1.5 IQR below the first quartile or larger than 1.5 IQR above the third quartile was considered as a cold or hot outlier, respectively.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4161696/v2","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4161696/v2","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn the Amazon, the dry season of 2023 as well as the beginning of the wet season in 2024 were marked by unprecedented high temperatures and large precipitation deficits. While the tropical forests in the Amazon play a crucial role in the global carbon cycle and are a biodiversity hotspot, they were also shown to suffer from El-Niño related droughts in the past, leading to legitimate concerns about the ecological consequences of the recent climate conditions. To this day, while there is a growing effort to make remote sensing products available close to real-time, land surface models that are critical tools to understand the interactions between the biosphere and the environment have lagged behind the present due to the complexity to run and process large model ensembles. In this study, we employed advanced machine learning models trained on state-of-the-art remote sensing and dynamic global vegetation model estimates of gross primary productivity (GPP). The models provide near real-time GPP estimates, revealing significant productivity reductions during the 2023/2024 drought. Negative GPP anomalies were more widespread across the Amazon than during any other recent major drought event. The Climate-GPP relationships that emerged from the models suggest that future temperature increases and changes in precipitation will severely challenge Amazon forest resilience.\u003c/p\u003e","manuscriptTitle":"Intact Amazon forests hit a record low gross primary productivity level in 2023-2024","msid":"","msnumber":"","nonDraftVersions":[{"code":2,"date":"2024-07-09 16:48:36","doi":"10.21203/rs.3.rs-4161696/v2","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}},{"code":1,"date":"2024-03-26 12:00:40","doi":"10.21203/rs.3.rs-4161696/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"896c1ee1-0bde-4f21-bcc5-6b5622ecd169","owner":[],"postedDate":"July 9th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":29903359,"name":"Biological sciences/Ecology/Forest ecology"},{"id":29903360,"name":"Earth and environmental sciences/Environmental sciences/Environmental impact"},{"id":29903361,"name":"Earth and environmental sciences/Climate sciences/Climate change/Climate-change impacts/Environmental health"},{"id":29903362,"name":"Biological sciences/Plant sciences/Photosynthesis"}],"tags":[],"updatedAt":"2024-06-27T16:21:07+00:00","versionOfRecord":[],"versionCreatedAt":"2024-07-09 16:48:36","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v2","identity":"rs-4161696","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4161696","identity":"rs-4161696","version":["v2"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}