Large-scale deforestation poses a significant threat to ecosystem stability and climate, leading to increased carbon dioxide emissions, which exacerbates global warming and ecological imbalance. To achieve sustainable forest management, precise large-scale monthly deforestation mapping has become increasingly important. The open access to Sentinel-1 data provides unprecedented opportunities for monthly deforestation mapping. However, previous monthly mapping based on Sentinel-1 and deep learning still needs improvement in accuracy, and the best strategies for large-scale model transfer have not been fully explored. This study proposes a new approach for monthly deforestation mapping based on Sentinel-1 data and an adapted Segment Anything Model (SAM), combined with active learning and transfer learning strategies for large-scale model transfer. The model was tested and evaluated in four different study sites: Rondônia in Brazil, Guangxi in China, California in the USA, and Hainan in China. The results showed the superior performance of our proposed adapted SAM method, with F1 scores ranging from 0.74 to 0.88 and IoU from 0.58 to 0.78. The combined model for the four regions achieved an F1 score of 0.81 and an IoU of 0.68, outperforming the baseline U-net model (combined F1 score of 0.78 and IoU of 0.64). When applied to new sites, the fine tune-based transfer learning significantly improved the model’s spatial generalization capability with the addition of a small number of target domain samples. Moreover, compared with random sampling approach, the active learning technique help reduce the required number of training samples to achieve the same level of accuracy. This study provides a comprehensive workflow for improved monthly deforestation mapping, emphasizing the advantages of combining Sentinel-1 SAR data with advanced models and strategies. Our method offers a reliable and efficient solution for large-scale deforestation monitoring, aiding in the timely detection of deforestation activities and supporting sustainable forest management strategies.

How to cite: Zhao, F.: Monthly mapping of deforestation in the Amazon using Sentinel-1 data and a vision foundation model, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19757, https://doi.org/10.5194/egusphere-egu25-19757, 2025.

White sand ecosystems (WSE)- known locally as Campinarana, Campinarana florestada (in Brazil) or Caatinga Amazonica (Venezuela) are typically thin-stemmed, nutrient scarce, low canopy ecosystems located on sandy soils (podzols) distributed across the Amazon basin. It has been previously documented that WSEs can form histosol layers capable of storing significant carbon, but existing studies are limited in geographic scope and quantity of data points. Notably a new study measured up to 2m of peat in Colombian WSEs, but we lack a wider understanding of the distribution and dynamics of peat forming WSEs across Amazonia.

Here we undertake a simple spatial analysis, overlapping a recently published Amazon peat map with previously published WSE distributions. We combine this with recent carbon density data and insights gained from an in-depth study of Colombian white sand peatlands (WSP) by Winton et al. (in review).

We estimate a total white sand peatland area of 78,832 (40,403 – 117,133) km² across the Amazon basin, corresponding to 26% of all white sand ecosystems. The greatest concentration is in the Rio Negro basin in Brazil. We predict that 39%, 26%, 15% and 6% of Amazon basin WSE forests are underlain by peat in Venezuela, Brazil, Colombia and Peru respectively. We in turn estimate a total carbon stock of 3.86 (0.64–7.48) Pg C in the WSPs of the Amazon basin, comparable to that of the largest known peatland region in the South America- the Pastaza-Maranon Foreland Basin.

We conclude that WSPs are critically understudied ecosystems and represent a fundamental gap in our understanding of the Amazon basin carbon cycle. Crucially, no existing studies appear to be located in the most concentrated areas of peat, with Colombia being the only substantial WSP region to be densely sampled. Our results can inform future research priorities in WSPs.

How to cite: Hastie, A., Fernández, R. H., Honorio Coronado, E. N., and Winton, R. S.: Global importance of Amazonian white-sand peat carbon, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11765, https://doi.org/10.5194/egusphere-egu25-11765, 2025.

The Amazon biome is one of the largest carbon reservoirs, a relevant carbon sink in the world. The large extension and diversity of the Amazon biome hampers the assessment of regional-scale carbon budget based solely on local observations. Land surface models can provide carbon flux estimates, but they require proper calibration to represent the dynamics of the different ecosystems, abiotic conditions and vegetation characteristics in the Amazon Basin. One of the most important land surface model is JULES being increasingly used in tropical forests to estimate carbon fluxes. However, there is a lack of parameterization information that can be applied to the Amazon biome. Thus, this study presents an optimization of JULES main sensitivities parameters for different sites of the Amazon biome. For this attempt, we selected four Eddy-covariance flux towers as a reference based on different regions of the Amazon biome: K34 (Manaus, 2.614S/60.12W); K67 (Santarem, 2.85S/54.97W); RJA (Reserva Jaru, 10.08S/61.93W and ATTO (São Sebastião do Uatumã, 2.15S/59.03W). The variables analyzed to reproduce the carbon dynamics were the Net Ecosystem Exchange (NEE), Gross Primary Production (GPP) and eutrophic respiration (RESP) during one year of analysis. JULES most sensitivities parameters adjusted were related to the Upper-temperature threshold for photosynthesis (tupp_io); Scale factor for dark respiration (fd_io); The maximum ratio of internal to external CO₂ (f0_io) and Quantum efficiency (alpha_io).  The optimization was made using the Nelder-Mead method and after a leave-one-out cross-validation method was implemented to evaluate the simulation efficiency in each site. Also, the new parametrization in each site was compared with the default version of JULES and with another model Vegetation Photosynthesis and Respiration Model (VPRM). We selected the Wilmott index of agreement (d) and the Root Mean Square Error (RMSE) to analyze simulation efficiency. The Nelder-Mead optimization method reduced the error in GPP simulations in each Tower in comparison to the two models evaluated however the new parametrization of JULES was not able to improve RESP in these sites. However, the optimization procedure presented better results in NEE in each tower evaluated in the Amazon biome being the ATTO tower that demonstrated the most efficient simulations (d =0.60;  RMSE = 2.03 g C m^-2 day^-1) in comparison to the default version (d= 0.52; 3.09 g C m^-2 day^-1) and VPRM (d = 0.58; 2.29 g C m^-2 day^-1). In general, results demonstrated that the new parametrization of JULES reduced the error of simulation compared to the last version of JULES for tropical forests and better represented the seasonality compared to the VPRM model.

How to cite: Prudente Junior, A. C., Santos da Silva, F., De Paula Cordeiro, L., Botía, S., Varanda Rizzo, L., Dias Freitas, E., Ambrizzi, T., Artaxo Netto, P. E., and Toledo Machado, L. A.: Optimization of JULES model in different sites of the Amazon biome, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-3862, https://doi.org/10.5194/egusphere-egu25-3862, 2025.

The Amazon, with its vast wetlands, is a significant hotspot for greenhouse gas emissions. However, the emissions from aquatic systems remain poorly understood. The Rio Negro is one of the main tributaries of the Amazon river but, to date, there have been few measurements on GHG concentrations and fluxes, and none for the upper Rio Negro region.

We present the first continuous measurements of dissolved CO₂, CH4, N₂O and CO in the Rio Negro, between the cities of Manaus and São Gabriel de Cachoeira (~1000 km). From a moving research vessel, water was sampled continuously from a depth of 50 cm and passed through a bubble-type-equilibrator. A closed air stream was circulated through the equilibrator, and continuously measured by an in-situ FTIR analyzer. In addition, variables such as pH, air and water temperature, DOC and coliform bacteria were determined.

All measured gases were supersaturated in the water with respect to the atmosphere, indicating an outgoing flux toward the atmosphere. CH₄ concentrations showed elevated concentrations in the middle Rio Negro, contrasting with CO₂, which peaked in the upper and lower Rio Negro. Both CH₄ and CO₂ displayed distinct hotspot regions, many of which were centered around human settlements and are therefore likely of anthropogenic origin, as also confirmed by the observed bacterial communities. A few hotspots appeared to be linked to surrounding wetlands, which may release large amounts of CH₄ during the rising water phase when reconnected to the main river. N₂O showed elevated concentrations in the upper Rio Negro, possibly linked to the extensive white sand forest areas in this part of the catchment. CO concentrations showed a clear diurnal pattern, with highest concentrations coinciding with highest incoming solar radiation.

Based on our measurements, we suggest that anthropogenic influences on remote rivers such as the Rio Negro may be greater than previously assumed, potentially affecting the representativeness of both past and future field measurements.

How to cite: van Asperen, H., Warneke, T., Estefani Batista, C., Souza da Silva, J., Rizzo, L., Klemme, A., Lopes e Oliveira, R., Duvoisin Junior, S., Forsberg, B., and Trumbore, S.: Dissolved greenhouse gases in the Rio Negro (Amazonia, Brazil): the influence of humans versus wetlands, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-13543, https://doi.org/10.5194/egusphere-egu25-13543, 2025.

Multidimensional trait relationships are imperative to understanding forest functioning in the face of ongoing environmental changes. Warming and more frequent and severe water stress are expected to adversely affect tropical forests’ large carbon uptake capacity. Therefore, it is, important to evaluate how tropical trees would perform under future climate scenarios. It is challenging to analyse trait-performance relationships solely based on observational data. However, process-based models representing key plant trait trade-offs can be applied to investigate the influence of different plant hydraulic strategies on tropical tree performance.

We used a hydraulics-enabled version of the Lund-Potsdam-Jena General Ecosystem Simulator (LPJ-GUESS) to study how trait relationships influence tropical plant performance on three study sites included in the RwandaTREE (Rwanda tropical elevation gradient) experiment. We parameterised four endemic species based on observational data and ran simulations by varying selected traits within the potential ranges to evaluate how these parameters affect the simulated biomass and woody growth rate.

The results showed a variation in optimum trait values which led to realistic simulated woody growth rates, depending on successional strategies and study sites. This can be attributed to the emerging functional strategies defined by the trait relationships. 

Our results highlight that we can evaluate complex trait relationships and trade-offs that cannot feasibly be measured across large scales. This allows us to formulate new hypotheses on which hydraulic and structural trait correlations define plant performance. Increased understanding of drought-related vegetation processes can be used to decrease uncertainty in simulating tropical forest resilience and extreme weather impact on Pan-African carbon stocks.

How to cite: Pongracz, A., Pugh, T. A. M., Olin, S., Eckes-Shephard, A., Uddling, J., Wallin, G., Manzi, O. J. L., Wittemann, M., Nsabimana, D., Zibera, E., Ziegler, C., Manishimwe, A., Papastefanou, P., and Rammig, A.: Linking form to function: Simulating plant hydraulic strategies’ impact on tree drought response in a tropical montane rainforest, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-17267, https://doi.org/10.5194/egusphere-egu25-17267, 2025.

Tropical forests, spanning wet and dry forest ecosystems, are pivotal in regulating the global carbon cycle and climate through dynamic exchanges of energy, water, and carbon. These ecosystems influence regional and global climate patterns via biogeochemical feedback mechanisms. However, climate change is altering these processes, with rising temperatures intensifying evaporative demand and affecting photosynthetic activity, as indicated by changes in net ecosystem exchange (NEE). Vegetation and biomass variations further impact microclimates, feeding back into heat and water budgets. Understanding the dynamics and meteorological drivers of carbon and water fluxes is essential for comprehending land surface–atmosphere interactions.

This study compares the climatological and ecological functions of tropical wet and dry forests by examining two contrasting sites in the tropical Andes Mountains of southern Ecuador: the montane dry forest (MDF) in the Laipuna Reserve and the montane rain forest (MRF) in the Reserva Biológica San Francisco. The MDF is characterized by a deciduous forest and exhibits pronounced seasonality, with distinct dry (June–December) and wet (January–May) periods, driven by the inter-hemispheric shift of the Intertropical Convergence Zone (ITCZ). In contrast, the MRF experiences year-round rainfall, sustaining an evergreen lower montane forest type. Eddy-covariance measurements were used to monitor water and carbon fluxes under these contrasting climatic regimes. This comparison provides valuable insights into the differential roles of these ecosystems in regulating the Earth's energy and carbon budgets under changing climatic conditions. The objective of the study is (i) to quantify the magnitude and seasonality of NEE and its partitioned components, gross primary production (GPP), and ecosystem respiration (Reco) And (ii) to identify the meteorological drivers responsible for the variations in carbon exchange within each ecosystem. The results reveal significant variations in NEE in the MDF between wet and dry seasons. During the wet season, the average NEE was -3.9 μmol m⁻² day⁻¹, while in the dry season, it declined substantially to -0.8 μmol m⁻² day⁻¹. In contrast, the MRF demonstrated a consistently higher average NEE of -18 μmol m⁻² day⁻¹. These variations are driven by distinct environmental factors. In the MDF, water availability, regulated primarily by precipitation, is the dominant factor influencing carbon exchange. Conversely, the carbon dynamics in MRF are predominantly governed by energy inputs, with light playing a critical role in driving its NEE.

How to cite: Murkute, C., Sayeed, M., Pucha-Cofrep, F., Raffelsbauer, V., Ahmed, R., Scholz, S., Limberger, O., Carillo-Rojas, G., Fries, A., Bendix, J., and Trachte, K.: Carbon flux dynamics in montane tropical wet and dry forests: A comparative study in Southern Ecuador, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-14652, https://doi.org/10.5194/egusphere-egu25-14652, 2025.

In the face of ongoing global crises, such as climate change and biodiversity loss, we urgently need to understand dynamic and complex responses of global forest ecosystems. To do so, we need to develop modeling frameworks that account for multiple temporal and organizational scales, and therefore capture functional adaptations of individuals, species, and ecosystems in response to the environment.

Here we present Plant-FATE (Plant Functional Acclimation and Trait Evolution) an eco-evolutionary vegetation model that embodies functional diversity by representing plant life-history strategies, and adaptations by accounting for short-term physiological acclimation, mid-term demographic shifts, and long-term trait evolution.

Tested with data obtained from an hyperdiverse site in the Amazon Forest, our model predicts a nonlinear response of tropical forests to increasing atmospheric CO₂ due to diverse aspects of the growth-mortality tradeoff. At moderately elevated CO₂, we found that evolution towards higher wood density increases vegetation C sequestration. By contrast, under highly elevated CO₂ levels, a darkening understorey rather triggers lower wood densities, thus reversing gains from the proposed CO₂ fertilization effect.

Our results suggest that competition for resources may modulate community-level eco-evolutionary dynamics of forest ecosystems, such that competition-induced changes in wood density may render forests more vulnerable to future climatic extreme events. Our study highlights the importance of accounting for eco-evolutionary dynamics when simulating the functional response of forest ecosystems to projected climate change.

How to cite: Hofhansl, F., Singh, S., Stefaniak, E., Maxwell, T., and Joshi, J.: Simulating plant functional acclimation and trait evolution using an eco-evolutionary vegetation model (PlantFATE), EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11302, https://doi.org/10.5194/egusphere-egu25-11302, 2025.

Volatile isoprenoids take part in a wide range of forest-atmosphere processes that scale from plant cell regulation to atmospheric particle formation. Major drivers of plant leaf emissions are light and temperature - i.e., seasonality - and leaf age, suggesting leaf phenological type (i.e., evergreen or brevideciduous) may exert control over emission rates. The Amazon Forest is the greatest and most diverse source of volatile isoprenoid emissions, but the lack of leaf-level studies and the logistical challenges of measuring in such remote and highly bio-diverse sites bring high levels of uncertainty to modeled estimates. Studies indicate that hyperspectral leaf reflectance is an effective tool for estimating leaf morphological, physiological, and chemical traits, being perhaps a promising tool for remotely assessing volatile isoprenoid emissions from vegetation. Considering this, our research aimed at evaluating i) whether leaf phenological type and functional traits are determinants of the presence and magnitude of isoprene emissions and of mono- and sesquiterpene storage, and ii) whether leaf-level hyperspectral reflectance can be used to predict the presence of isoprene emissions and mono- and sesquiterpene storage in central Amazon forest trees. We found that isoprene-emitting evergreen trees were less likely to store monoterpenes and had tougher and less photosynthetically active leaves, while higher isoprene emission rates in brevideciduous trees associated with higher storage of sesquiterpene and phenolic compounds, suggesting that isoprene emissions possibly mediate a mechanical-chemical defense trade-off in evergreen and brevideciduous trees in this forest. Furthermore, we saw that dry leaf hyperspectral reflectance data and fresh leaf reflectance at selected wavelengths (616, 694, and 1155 nm) predicted the presence of isoprene emissions with accuracies of 0.67 and 0.72, respectively. Meanwhile, the presence of terpene storage was well predicted from fresh leaf reflectance data for monoterpene storage (accuracy = 0.65) and sesquiterpene storage (accuracy = 0.67). These results indicate the possibility of using spectral readings from herbarium specimens to assist in the development of more efficient sampling designs targeted at potential isoprene emitters, as well as of using fresh leaf reflectance data to calibrate multi-sensor equipment to remotely detect potential isoprene emitters or orientate sampling efforts in the field toward potential terpene-storing trees. The use of spectral tools for detecting potential volatile isoprenoid emitters and a more functional trait-based, mechanistic representation of emissions can combine to reduce modeling emission uncertainties and contribute to understanding the roles of volatile isoprenoids within forest-atmosphere interactions, atmospheric chemistry, and the carbon cycle.

How to cite: Robin, M., Durgante, F., Mallmann, C. L., Hadlich, H., Römermann, C., Niinemets, Ü., Gershenzon, J., Huang, J., Nelson, B., Taylor, T., de Souza, V., Pinho, D., Falcão, L., Lacerda, C., Duvoisin, S., Schmidt, A., Piedade, M. T. F., Schöngart, J., Wittmann, F., and Gomes Alves, E.: Leaf phenological type, functional traits, and hyperspectral reflectance to predict volatile isoprenoid emissions in central Amazon Forest trees, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-20870, https://doi.org/10.5194/egusphere-egu25-20870, 2025.

During the rainy season, black carbon (BC) particles exhibit strong variability in concentration. At the Amazon Tall Tower Observatory (ATTO), located in central Amazonia, elevated BC concentrations have been previously identified as originating from the African continent. However, BC mass concentrations approach zero during certain periods, characterizing pristine episodes. This study aims to identify the primary factors influencing BC concentration in the Amazon. The first analytical approach involved evaluating air mass back trajectories during episodes of high BC concentration (BC > 0.46 µg/m³, with the day of maximum concentration selected from neighboring days) and low BC concentration (BC < 0.08 µg/m³, with the day of minimum concentration chosen). Ensemble back trajectories, analyzed across multiple atmospheric levels, revealed minimal differences between the air trajectories associated with these two contrasting scenarios. The second approach examined accumulated rainfall at ATTO during the three days preceding the selected high- and low-concentration days. The results indicate that precipitation plays a dominant role in modulating BC concentrations. A histogram of precipitation data revealed two distinct patterns: one corresponding to high rainfall during pristine events and another to low or negligible rainfall during more polluted days. Using ERA-5 reanalysis data, this precipitation variability was observed to extend across the Intertropical Convergence Zone (ITCZ) over the Atlantic. Simulations were conducted using the ECHAM/MESSy Atmospheric Chemistry (EMAC) model to investigate this phenomenon further. The simulations demonstrated that rainfall variability influences the transport from Africa to the Amazonas of particles such as BC, dust, and gases, including CO₂ and O₃. Composite analyses of hemispheric synoptic patterns were performed by selecting days with high and low BC concentrations from January–February from 2015 to 2022. These composites revealed that the variability is driven by oscillations in the western hemisphere synoptic patterns linked to the positioning of cold fronts in both hemispheres. This variability has significant implications for transporting vital nutrients to the Amazon rainforest. Understanding the relationship between rainfall, synoptic patterns, and BC transport is crucial, particularly in the context of climate change, which could alter these patterns and profoundly impact the ecological systems of the Amazon basin.

This study was supported by FAPESP 2022/07974-0

How to cite: Toledo Machado, L. A., Gan, M. A., M. Barbosa, H. J., Holanda, B., Pozzer, A., and Pöhlker, C.: Linking ATTO Black Carbon to Rainfall Patterns in the South America-Atlantic Tropical Zone, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-1726, https://doi.org/10.5194/egusphere-egu25-1726, 2025.

Tropical rainforests play an important role in the global carbon cycle. They store massive amounts of biomass in their trees and soils, and contribute to climate mitigation by removing carbon from the atmosphere through photosynthesis. In a large-scale free-air CO₂ enrichment (FACE) experiment in a highly diverse, old-growth, tropical forest in the Brazilian Amazon, we will assess the ecosystem responses to rising atmospheric CO₂ concentrations. The main questions are (1) whether elevated atmospheric CO₂ directly and sustainably stimulates photosynthesis (the so-called CO₂-fertilization effect) and (2) will reduce stomatal conductance, leading to reduced water loss at leave-level and whether this will result in canopy-scale changes in transpiration and soil water availability, (3) how low nutrient availability (particularly phosphorus) will limit the CO₂-fertilization effect, and (4) whether elevated CO₂ concentration will alter the functional composition of vegetation. Also the role of biodiversity (through functional traits) and socio-environmental implications of CO₂ fertilization will be investigated, with a focus on impacts, adaptations and the science-policy interface. Through integrative modelling activities, the long-term goal of the project is to improve the projections of the Amazon rainforest carbon cycle and regional and global climate under increasing atmospheric CO2 concentrations. We here present the AmazonFACE science plan, give an update on the state of the experiment construction and show baseline measurements and simulation results.

How to cite: Rammig, A. and Lapola, D. and the AmazonFACE Team: The science plan for AmazonFACE, a large-scale Free Air CO2 Enrichment Experiment in the Amazon rainforest, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-10816, https://doi.org/10.5194/egusphere-egu25-10816, 2025.