Wildfires have shaped ecosystems for millions of years, with plant functional traits playing a key role in fire behaviour and severity. Morphological and physiological traits, particularly at the leaf and shoot levels, influence flammability by determining fuel composition and structure within both canopy and litter layers. These traits are hypothesized not only to affect critical fire dynamics, such as the likelihood of surface fires transitioning into crown fires, with significant consequences for fire intensity and ecosystem impacts, but also influence the evolution of fire-related traits.

This study investigates how leaf- and shoot-level morphology influences flammability in canopy and litter contexts across six dominant conifer families: Araucariaceae, Cupressaceae, Pinaceae, Podocarpaceae, Taxaceae, and Taxodiaceae. Flammability properties were assessed using fire calorimetry to measure ignitability, flame spread, and variability in the rate of energy release from combustion. Results indicate that while shed plant parts (e.g., leaves and shoots) shape fire behaviour by influencing bulk density, aeration, and flame spread rate—ultimately affecting burn sustainability and total energy release—shoot-level traits in isolation, including leaf shape and the arrangement of leaves within shoots, do not consistently predict flammability in canopy material.

Our findings highlight the dynamic interplay between plant morphology, fire regimes, and evolutionary pressures. Traits such as leaf size, shape, and arrangement contribute to fuel structure, driving patterns of fire behaviour that influence long-term plant fitness and survival. This underscores the importance of reconciling fire behaviour, plant functional traits, and the evolutionary history of fire adaptations across phylogenies.

With global change drivers intensifying fire regimes, understanding the relationship between plant flammability, fire regimes, and the acquisition of fire-related traits is increasingly critical. Non-fire-adapted species may face heightened extinction risks, threatening ecosystem stability. Quantifying the intrinsic flammability of plant traits is therefore essential for informing fire management, guiding conservation strategies, and ensuring the long-term sustainability of vegetative communities in a changing climate.

How to cite: Koll, R. and Belcher, C.: Morphological drivers of flammability in canopy and litter contexts across conifer families, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-16840, https://doi.org/10.5194/egusphere-egu25-16840, 2025.

EGU25-16840

Download Close

How to cite: Engle, B., Bratoev, I., Crowley, M. A., Zhu, Y., and Senf, C.: Identifying Ignition Drivers of Lightning-Ignited Wildfires in Boreal Forests, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-3090, https://doi.org/10.5194/egusphere-egu25-3090, 2025.

EGU25-3090

Download Close

Lightning-induced ignitions play a major role shaping the frequency, patterns and characteristics of wildfires in several regions across the globe, including extreme wildfire events (e.g., Góis wildfire in 2017 in Portugal) and fire seasons, such as 2019-20 in Australia, 2020 in California, and 2023 in Canada. The attention to lightning-ignited wildfires has been growing in recent years. Studies on LIWs frequently associate lightning and wildfire data to discern or approximate the place and moment of fire ignition. This typically requires to select the lightning strike responsible for the ignition.

Currently, several methods are applied to select the most likely lightning strike causing the ignition. However, this selection is complicated by, at least, two aspects. First, the spatial uncertainty of fire and lightning data (e.g., the location errors of detected lightning events). Second, the holdover phenomenon. Holdover time, commonly defined as the time between lightning-induced fire ignition and fire detection, can range from a few minutes to several days, and more rarely to some weeks or even months. Long holdover times are associated to the presence of a smoldering phase that hinders the detection of these lightning fires.

Here, we present a novel method that uses location accuracy information from lightning location networks, as well as expected distributions of holdover time, to assess the probabilities of lightning igniting wildfires. Our method computes a probability metric, which is the product of two independent probabilities: a spatial and a temporal probability. The spatial component assesses the probability of a cloud-to-ground lightning event striking within a given area surrounding the fire discovery point, while the temporal component evaluates the probability of a lightning-ignited wildfire undergoing a certain holdover time. The lightning event with the maximum probability metric value is then selected as the most likely ignition source. We applied this method in three study areas: Switzerland, Catalonia (Spain), and California and Nevada (USA). The results were compared with lightning selections identified by the index of proximity, one of the currently most common methods to select the most likely ignition source of lightning-induced wildfires.

The initial results indicate that the probability metric yields a different selection of lightning events, in comparison with the index of proximity, for a great proportion of wildfires, with considerable differences across the study areas. We suggest that the probability metric provides a solid alternative to current methods. The probability metric offers some advantages: (1) it simplifies some methodological decisions despite the need for additional computations; (2) it is flexible and can be adapted to different types of lightning and fire data (e.g., fire perimeters); (3) it has a more robust theoretical basis than current methods; and (4) the lightning selection can be enhanced over time due to continuous improvements in lightning and fire databases.

How to cite: Moris, J. V., Hunt, H. G. P., Álvarez-Álvarez, P., Conedera, M., Gordillo-Vázquez, F. J., Lapierre, J., Pérez-Invernón, F. J., Pineda, N., Pezzatti, G. B., Veraverbeke, S., and Ascoli, D.: A new probabilistic method to identify fire-igniting lightning events, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-3335, https://doi.org/10.5194/egusphere-egu25-3335, 2025.

EGU25-3335

Download Close

How to cite: Gincheva, A., Torres-Vázquez, M. Á., Di Giuseppe, F., Moreno Torreira, A., Jerez, S., and Turco, M.: Rising Synchronicity of Extreme Fire Weather Across Europe in a Warming Climate , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-4352, https://doi.org/10.5194/egusphere-egu25-4352, 2025.

EGU25-4352

Download Close

Rural fires are recurrent in Southern Europe due to climate conditions, land use change, or a combination of both. Wet and mild winters and dry and warm summers favour the growth of vegetation and its subsequent low moisture content, increasing fuel availability. In Portugal, between 15 and 20 September 2024, severe wildfires burned more than 145,000 hectares and caused the death of more than 9 people. In Greece a major fire, stated as the largest recorded in the EU, started near the city of Alexandroupolis on August 21, with around 80.000 hectares burnt, mainly affecting the Dadia Forest and causing the death of almost two tens of migrants. Despite the crucial role played by dry fuel conditions fostering the propagation of wildfires, favourable meteorological conditions and fuel accumulation are related to the recorded fire activity and burned area. The influence of spring meteorological conditions on fire season burned area through their effect on fuel accumulation and dryness is assessed. The link between hot temperature and water availability in spring and the increased risk of summer flammability and fire spread through their influence on vegetation gross productivity is evaluated using satellite-derived data. The important role of fuel accumulation during the early growing season in fire-prone regions is highlighted in the case of Portugal in 2024 and Greece in 2023 and reinforces the crucial importance of fuel management for the definition of effective fire prevention measures in the context of warmer and drier conditions forecasted for southern European Countries.

Acknowledgements: This work is supported by the Portuguese Fundação para a Ciência e Tecnologia, FCT, I.P./MCTES through national funds (PIDDAC): UID/50019/2025 and LA/P/0068/2020 https://doi.org/10.54499/LA/P/0068/2020 and also on behalf of DHEFEUS -2022.09185.PTDC and the project FAIR- 2022.01660.PTDC).

How to cite: Gouveia, C., Finuras, M., Russo, A., and Ermitão, T.: The role of dry and heat extremes on vegetation dynamics in the recent fire seasons in Southern Europe, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19493, https://doi.org/10.5194/egusphere-egu25-19493, 2025.

EGU25-19493

Download Close

How to cite: Liu, Y., Qian, Y., and Wang, M.: Fire–precipitation interactions amplify the quasi-biennial variability in fires over southern Mexico and Central America , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-5454, https://doi.org/10.5194/egusphere-egu25-5454, 2025.

EGU25-5454

Download Close

Wildfires pose a significant threat to ecosystems, human life, and infrastructure, particularly in South America, where diverse climatic and environmental factors contribute to their occurrence. Climate change has exacerbated extreme weather conditions such as intense heat and drought, leading to a global increase in the frequency and intensity of wildfires. Countries like Brazil have experienced significant rises in wildfire damage, highlighting the urgent need for predictive models that accurately assess future wildfire risks to mitigate their impact effectively. This thesis addresses this need by developing a wildfire risk prediction system leveraging deep learning methods and remote sensing data.

Using Earth Observation (EO) APIs, the system avoids downloading and storing vast amounts of satellite imagery, enabling efficient data acquisition and preprocessing. The study focuses on key variables that influence wildfire activity, including dynamic variables such as Land Surface Temperature (LST), Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), radiation, Leaf Area Index (LAI), evapotranspiration (ET), wind speed, and temperature, as well as static variables like land cover, Digital Elevation Model (DEM), and population density. The system is designed to predict wildfire risk for the next day and up to eight days, offering a robust tool for proactive wildfire management.

Given the stochastic and nonlinear nature of wildfire phenomena, this research employs advanced deep learning techniques, including Random Forests (RF), Long Short-Term Memory networks (LSTM), and Convolutional LSTM (ConvLSTM) models, to predict wildfire risk in near real-time. Active fire data from MODIS products, along with their burn dates, serve as the basis for training datasets. Non-fire points are generated by mapping the land cover distribution of fire points, ensuring balanced datasets for model training. Variables are extracted and classified into dynamic and static categories to capture both temporal variability and fixed geographical characteristics.

The objectives of this research are threefold: (1) to investigate existing remote sensing-based wildfire management methodologies and identify enhancements through the integration of data cubes and deep learning; (2) to develop a scalable platform for efficient data acquisition, preprocessing, and risk prediction using deep learning algorithms; and (3) to evaluate the system’s accuracy, efficiency, and scalability with real-world datasets and disaster scenarios.

Preliminary results highlight the effectiveness of integrating remote sensing data with deep learning models for wildfire risk prediction. Dynamic variables such as EVI, LST, and NDVI, along with human influence factors like Global Human Modification Index (gHM), emerged as key predictors, demonstrating the interplay of environmental and anthropogenic drivers in wildfire occurrences. Seasonal analysis from 2021 to 2024 revealed a strong correlation between fire activity, elevated temperatures, and declining vegetation indices from November to April. The Random Forest model achieved 83% accuracy, while the LSTM model showed promise with 75% accuracy, emphasizing the potential of both static and temporal data. These findings lay a robust foundation for enhancing wildfire risk management through advanced machine-learning approaches.

How to cite: Sanchez Vargas, J. A., Heisig, J., Painho, M., and Gharun, M.: Development of a Wildfire Risk Prediction System based on Deep Learning Methods and Remote Sensing, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7107, https://doi.org/10.5194/egusphere-egu25-7107, 2025.

EGU25-7107

Download Close

Recent advancements in machine learning (ML) have significantly broadened its applications, including the potential to transition from forecasting fire weather to predicting actual fire activity. In this study, we demonstrate the feasibility of this transition using an operational forecasting system. By integrating data on human and natural ignitions along with observed fire activity, data-driven models effectively address the persistent overprediction of fire danger in fuel-limited biomes. This results in fewer false alarms and more informative outputs compared to traditional methods.

A key factor driving this improvement is the availability of global datasets for fuel dynamics and fire detection, which were not accessible during the development of earlier physics-based models. We find that the enhanced predictive skill of ML models stems largely from the comprehensive characterization of fire processes provided by these datasets, rather than from the complexity of the ML methods themselves.

As enthusiasm gather around  data-driven approaches, our findings highlight the critical importance of high-quality training data in improving forecast accuracy. While the rapid advancement of ML techniques generates excitement, there is a risk of undervaluing the essential role of data acquisition and, where necessary, its creation through physical modeling. Our results underscore that investing in robust datasets is indispensable and should not be overlooked in the pursuit of  very complex algorithm.

How to cite: Di Giuseppe, F., Mc Norton, J., Wetterhall, F., and Lombardi, A.: The real drivers of the ML revolution in fire forecasting, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8637, https://doi.org/10.5194/egusphere-egu25-8637, 2025.

EGU25-8637

Download Close

Machine learning (ML) models are widely used to predict wildfire occurrence and susceptibility (Brys et al., 2025). However, while these models excel at prediction, they often fail to provide insights into their inner workings or uncover the causal pathways driving wildfires. This study addresses this limitation by extending ML models beyond prediction to explore the drivers and causal pathways underlying wildfire occurrence. Our primary aim is to identify meaningful, interpretable patterns from wildfire data.

We developed a novel multi-stage clustering methodology inspired by Cooper et al. (2021) and Cohen et al. (2024). This approach integrates feature attribution (SHAP values), dimensionality reduction (UMAP), hierarchical clustering (HDBSCAN), and causal discovery methods: PC and FCI (Spirtes et al., 2001), and DirectLiNGAM (Shimizu et al., 2011). The causal methods were enhanced with prior background knowledge to derive meaningful insights. We used datasets from Italy (Cilli et al., 2022) and the Netherlands.

A central feature of our methodology is the use of SHAP values to define subgroups and derive causal pathways. SHAP values reduce noise in the feature space while preserving critical information for clustering. By reducing multidimensional SHAP values to two dimensions with UMAP, we improved clustering performance and interpretability. The resulting clusters were described using concise, non-overlapping decision rules based on the original variables, eliminating the need for manual filtering commonly required in clustering raw feature space. The identified clusters revealed specific relationships between wildfire drivers and occurrence. For each cluster, we applied advanced causal discovery techniques to derive probable causal pathways, aligning the findings with the knowledge of stakeholders and domain experts. These actionable and interpretable explanations offer practical utility.

Findings from the case studies demonstrate that supervised clustering effectively characterizes wildfire occurrence by linking it to influencing factors. Furthermore, the approach provides valuable insights into cluster-specific causal pathways. The methodology translates complex relationships into simple causal logic, offering stakeholders and domain experts the necessary context to understand the model's behavior.

Brys, C., La Red Martínez, D.L. & Marinelli, M. Machine learning methods for wildfire risk assessment. Earth Science Informatics 18, 148 (2025). https://doi.org/10.1007/s12145-024-01690-z

Cilli, R., Elia, M., D’Este, M., Giannico, V., Amoroso, N., Sanesi, G., Lombardi, A., Pantaleo, E., Monaco, A., Tangaro, S., Bellotti, R. & Lafortezza, R. (2022). Explainable artificial intelligence (XAI) detects wildfire occurrence in the Mediterranean countries of Southern Europe. Scientific Reports 12, 16349. https://doi.org/10.1038/s41598-022-20347-9

Cohen, J., Huan, X. & Ni, J. (2024). Shapley-based explainable AI for clustering applications in fault diagnosis and prognosis. Journal of Intelligent Manufacturing, 35, 4071-4086. https://doi.org/10.1007/s10845-024-02468-2

Cooper, A., Doyle, O. & Bourke, A. (2021). Supervised clustering for subgroup discovery: An application to covid-19 symptomatology. Communications in Computer and Information Science, 1525, 408–422. https://doi.org/10.1007/978-3-030-93733-1_29

Shimizu, S., Inazumi, T., Sogawa, Y., Hyvärinen, A., Kawahara, Y., Washio, T., Hoyer, P. O., & Bollen, K. (2011). DirectLiNGAM: A direct method for learning a linear non-Gaussian structural equation model. The Journal of Machine Learning Research, 12, 1225–1248. https://doi.org/10.48550/arXiv.1101.2489

Spirtes, P., Glymour, C. & Scheines, R. (2001). Causation, Prediction, and Search. Second Edition. MIT Press. https://doi.org/10.7551/mitpress/1754.001.0001

How to cite: Korving, H. and Van Marle, M.: Decoding Wildfires - Extracting Interpretations and Causal Pathways of Catalysts for Wildfire Occurrence from Machine Learning Models, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12889, https://doi.org/10.5194/egusphere-egu25-12889, 2025.

EGU25-12889

Download Close

In this study, we report the preliminary findings from a series of sans-human wildfire simulations using a process based dynamic global to regional vegetation model (DGVM), LPJ-GUESS v 4.1, coupled with the SIMple FIRE Model (SIMFIRE) and the wildfire combustion model (BLAZE), where we investigate the performance of the DGVM to reenact a specific wildfire instance in a Mediterranean catchment. For this, we compared the simulated burned area (BA^s) to that in the actual event (BA^o) in Manavgat, Antalya, Türkiye. The DGVM spatially captured the fire instance, albeit with a much smaller BA as a result. In July 2021, the largest single wildfire incidence for this region for the last two centuries occurred. The wildfire scorched an area of 60.000 ha.s where the dominant vegetation types were fire adapted dry conifer forests (mainly Pinus brutia) and Mediterranean shrubs. Previous years’ precipitation patterns had encouraged fuel build up, and the extreme heat of the summer of 2021, coupled with the seasonal drought and strong winds provided suitable environmental conditions for the wildfire’s spread. The ignitions in this specific case were intentional, majority were targeted arsons, and a plausible reason behind the ultimate extent of the BA. Here, we show the simulation results from our sans-human model runs using ERA5-Land reanalysis dataset, and compare BA^s to BA^o for this catchment for 2021. Our ultimate aim in these series of experiments where the ignition source is non-human is initially to decipher the dynamics, and later to develop a methodology to assess the human influence in BA in Mediterranean type ecosystems in the Eastern Mediterranean Basin, under a changing climate. 

How to cite: Ekberzade, B., Avcıoğlu, A., and Görüm, T.: Determining the human signal in burned area under a changing climate, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2313, https://doi.org/10.5194/egusphere-egu25-2313, 2025.

EGU25-2313

Download Close

The temporal coverage from ˜2000 to present of global burned area satellite observations limits many aspects of fire research. As a result, global fire models are often being used to investigate past and future fire behaviour. Unfortunately, the limited temporal coverage of the observations also hinders the development and evaluation of these fire models. The current generation of global fire models are capable of simulating some characteristics of regional fire behaviour, such as mean state and seasonality, well. However, the performance of these models differs greatly from region to region, and aspects such as extreme fire behaviour are not well represented yet.

Here, we propose a new, data-driven fire model that predicts burned area from the same input parameters that are passed to global fire models. We trained LSTMs to model burned area from GFED5. We split our data according to the IPCC regions and perform a region-based cross-validation, that is, we train different LSTMs on different region-splits of the data. We then compose the predictions of these different models so that for each region the predictions are made by LSTMs that have never seen any data during training and validation from that region before. Our model outperforms all fire models on a global scale and in most IPCC regions. With our model, we can improve our understanding of past fire behavior and simulate future fire trends.

How to cite: Lampe, S., Gudmundsson, L., Kraft, B., Le Saux, B., Hantson, S., Kelley, D., Humphrey, V., Chuvieco, E., and Thiery, W.: Modelling global burned area with deep learning, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15236, https://doi.org/10.5194/egusphere-egu25-15236, 2025.

EGU25-15236

Download Close

Peatland fires pose significant environmental and societal challenges. We recently advanced the Canadian Fire Weather Index (FWI) system for northern peatlands by integrating peatland-specific hydrological data derived from assimilating Soil Moisture and Ocean Salinity (SMOS) L-band brightness temperature observations into the NASA Catchment Land Surface model with its peatland modules, ‘PEATCLSM’. This novel FWI_peat (Mortelmans et al. 2024) was evaluated using satellite-based fire presence data over boreal peatlands from 2010 through 2018, demonstrating improved estimation of peatland fire presence.

Here, we extend the use of this renewed FWI_peat system by integrating it into a machine learning framework to gain deeper insights into when, where, and why peatlands burn. We utilize an XGBoost algorithm trained on peatland burned area data from 2012-2023, incorporating a suite of predictors, including (i) peatland distribution characteristics, (ii) peatland groundwater table, (iii) lightning occurrence, (iv) meteorological data, (v) vegetation properties, and (vi) socio-economic factors. This approach enables proactive fire risk management strategies and contributes to a comprehensive assessment of peatland fire vulnerability and resilience. Preliminary results indicate the importance of peatland groundwater table and lightning occurrence in estimating peat burned area.

Mortelmans, J., Felsberg, A., De Lannoy, G. J. M., Veraverbeke, S., Field, R. D., Andela, N., and Bechtold, M.: Improving the fire weather index system for peatlands using peat-specific hydrological input data, Nat. Hazards Earth Syst. Sci., 24, 445–464, https://doi.org/10.5194/nhess-24-445-2024, 2024.

How to cite: Mortelmans, J., De Lannoy, G., Dunmire, D., Veraverbeke, S., Waddington, J., Scholten, R., and Bechtold, M.: Modeling peat burned area and understanding its drivers with machine learning, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15841, https://doi.org/10.5194/egusphere-egu25-15841, 2025.

EGU25-15841

Download Close

Within the framework of the project PolarRES  (POLAR Regions in the Earth System) we assess the impact of climate change on the ecosystems of the terrestrial northern high latitudes by making use of a range of high resolution regional climate simulations. These regional simulations were themselves driven by global climate simulations selected following the storyline approach described in Levine et al. 2024 from the set of CMIP6 SSP3-7.0 simulations, namely NorESM2-MM and CNRM-ESM2-1. These define two extremes in the climatic envelope of the CMIP6 simulations. While NorESM2-MM shows a high warming of the Barents-Kara seas but a low Arctic tropospheric warming CNRM-ESM2-1 shows the opposite. The storyline approach is a comprehensive way of defining pathways for physical outcomes of climate change that are observable in the region of interest and can directly be linked to certain consequences.

The 2^nd generation Dynamic Global Vegetation Model (DGVM) LPJ-GUESS with its wildfire model SIMFIRE-BLAZE was applied using the high-resolution meteorological forcing from the regional climate models (RCMs) to investigate the potential impacts on both vegetation and the development of wildfires as well as the role of uncertainty implied by the variability of the forcing data.

It can clearly be stated that wildfire activity will increase significantly under the given scenarios driven mainly by shifts in vegetation distribution, i.e. northward migration of both treeline as well as shrubs and grasses. These effects differ regionally, depending on both, the storyline and the RCMs.

We present the findings from an envelope of potential future climate forcings depicting the impact of climate depending on the regionally observable effects of Arctic tropospheric warming and the Barents-Kara Seas warming, making use of the storyline approach as a comprehensive indicator for regional future change.

The results of this assessment will directly influence the research conducted in the project GreenFeedBack (GREENhouse gas fluxes and earth-system FEEDBACKs), which focusses on enhancing the knowledge on GHG dynamics in the boreal high latitude terrestrial and marine ecosystems.

Levine, X. Jet al. : Storylines of summer Arctic climate change constrained by Barents–Kara seas and Arctic tropospheric warming for climate risk assessment, Earth Syst. Dynam., 15, 1161–1177, https://doi.org/10.5194/esd-15-1161-2024, 2024

How to cite: Nieradzik, L., Lee, H., Levine, X., Miller, P., Mooney, P., Mottram, R., and Wårlind, D.: Assessing the impact of climate change on boreal high latitude wildfire using a storyline approach, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-17962, https://doi.org/10.5194/egusphere-egu25-17962, 2025.

EGU25-17962

Download Close

Whilst global burned area continues to decline, recent climate warming has led to an increase in the occurrence and intensity of extreme fires. Humanity must adapt to this new reality. Two proposed management options are a) prescribed livestock grazing, and b) prescribed fire use. Both methods promise cost-efficient means to reduce fire intensity, fire-induced vegetation mortality, and carbon emissions by reducing and fragmenting fuel loads. However, at present, there has been no systematic global assessment of the efficacy of these interventions. Reasons for this include a lack of data to understand their present-day distribution and impact as well as a lack of formal model structures to represent their uptake under future scenarios.

Here, we present two applications of the newly developed global, agent-based Wildfire Human Agency Model (WHAM!)¹ to assess the potential effect of prescribed grazing and prescribed fire as adaptations to future fire regimes. Firstly, to explore the effect of prescribed livestock grazing on global fire regimes, we share a representation of livestock grazing intensity in WHAM! and its integration with the generalised linear models of Haas et al., (²). Secondly, we present work on a tight coupling of WHAM! with the JULES-INFERNO dynamic global vegetation model, focusing on parameterisation of how managed human fire use reduces fire-induced vegetation mortality.

Overall, early results suggest both management options already play a significant role in reducing global fire intensity and highlight the importance of considering dynamic human responses to a changing climate in global projections of future fire regimes.

¹Perkins, O… et al. (2024). GMD.

²Haas, O. et al., (2022). Env. Res let.

How to cite: Perkins, O., Haas, O., Kasoar, M., Kelley, D., Teixeira, J. C. M., Voulgarakis, A., and Millington, J. D. A.: Adapting to fire in a warming climate: towards global assessment of prescribed grazing and prescribed fire, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-18640, https://doi.org/10.5194/egusphere-egu25-18640, 2025.

EGU25-18640

Download Close

How to cite: Wang, J., Li, J., Zhao, J., Zhong, X., Wang, M., He, J., and Yue, C.: Spatiotemporal changes in global cropland fire activity from 2003 to 2020, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-1776, https://doi.org/10.5194/egusphere-egu25-1776, 2025.

EGU25-1776

Download Close

There are multiple feedback mechanisms between wildfires and climate, such as temperature, emissions, cloud interactions, deposition, and land cover changes. Wildfires can also have large societal and ecological impacts and are considered as an extreme climate event. Despite this, most Earth System Models have, until recently, used prescribed fire emissions and fire plume injection heights for input into their atmospheric models that were unresponsive to climate changes. Fire plume heights, in particular, have a great influence on the radiative forcing and long-range transport of pollutants. This presentation will show recent results from global modelling of interactive fires (land-atmosphere) in the Canadian Earth System Model (CanESM), with a focus on key wildfire characteristics, such as aerosol emissions and fire plume height. These model improvements introduce the capacity to more accurately simulate future projections of wildfire characteristics under different climate scenarios. The upcoming applications of these improvements include experiments for the Hemispheric Transport of Air Pollution (HTAP) Fires project, AerChemMIP2, and Aerocom. HTAP Fires is a multi-model, multi-pollutant study with the goal of improving global fire modelling and using the multi-model ensembles to provide estimates of fire-related pollution for impact studies and policy makers.

How to cite: Whaley, C., Digby, R., Arora, V., Chen, J., Makar, P., Anderson, K., Griffin, D., Keating, T., Butler, T., Kaminski, J., and Wu, R.: Modelling interactive fires: climate-fire feedbacks on fire characteristics and multi-model projects, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-13774, https://doi.org/10.5194/egusphere-egu25-13774, 2025.

EGU25-13774

Download Close

The recovery time of ecosystems following wildfire significantly influences carbon sequestration rates, land-atmosphere exchanges, and hydrological processes. Post-fire recovery has been studied at local scales but there is a lack of comprehensive global-scale analyses. We used solar-induced chlorophyll fluorescence (SIF) to quantify the recovery of photosynthetic activity after more than 10,000 fires from diverse ecosystems. We used the relaxed lasso technique to identify key determinants of the length of time required for post-fire recovery, and used these to build a linear regression model. Our results show that vegetation characteristics, fire properties, and post-fire climatic conditions all influence recovery time. Gross primary production (GPP) is the most important determinant of recovery time: ecosystems with higher GPP recover faster. Fires with greater intensity and duration, which cause more extensive vegetation damage, are associated with longer recovery times. Post-fire climate also affects recovery time: anomalously high temperatures and temperature seasonality, and increased number of dry days, cause slower recovery, while above-average precipitation accelerates recovery. Recovery times vary between different biomes, potentially reflecting variations in plant fire adaptations: ecosystems with a higher abundance of resprouting plants recover more rapidly. These findings provide a global perspective on how vegetation responds to fire disturbances, offering insights into carbon and water cycle dynamics under changing climatic conditions.

How to cite: Shen, Y., Prentice, C., and Harrison, S.: Global Drivers of Post-Fire Ecosystem Recovery: Insights from Solar-Induced Chlorophyll Fluorescence, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-3308, https://doi.org/10.5194/egusphere-egu25-3308, 2025.

EGU25-3308

Download Close

Wildfires are an increasing concern for climate change, air quality and recognized for their substantial impacts on atmospheric composition. In addition to significant emissions of carbon dioxide (CO₂) and particular matter (PM), biomass burning events are characterized by substantial non-CO₂ emissions, which encompass a wide range of species. These emissions significantly influence atmospheric chemistry at a regional to global scale. Particularly in regions with ample fuel sources and hot, dry, or windy meteorological conditions, surface fires can lead to high-intensity crown fires and frequent downwind spotting. In certain circumstances, the intense formation of crown fires triggers the development of pyrocumulonimbus (PyroCb) atop smoke columns, which ascend to the upper troposphere and lower stratosphere (UTLS) and thus promote the dispersion of the fire emissions within wide regions. On August 2, 2019, the Williams Flats Fire ignited due to lightning from early morning thunderstorms in eastern Washington, USA. The main fire activity occurred between August 2 and August 9. On August 8, the high intensity crown fires led to the formation of a PyroCb. This event was observed and probed by the joint NOAA and NASA FIREX-AQ field campaign, providing a unique observation dataset. In this study, we utilize the Weather Research and Forecasting Model (WRF) to assess the impact of the Williams Flats fires on the atmospheric composition. In particular, we couple the representation of detailed multi-phase chemistry (WRF-CHEM) with WRF’s fire spread model (WRF-FIRE), employing WRF’s Large Eddy Simulation capabilities to resolve turbulence at resolutions of 100 m. In this presentation, results from WRF-FIRE-CHEM simulations with and without aqueous-phase chemistry will be shown to isolate its effects on the long-range transport of trace gases and aerosols.

How to cite: Rosanka, S., Juliano, T., Carlton, A. M., and Barth, M.: Large eddy simulations of the Williams Flat Fire: Aqueous chemistry in pyrocumulous clouds, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12104, https://doi.org/10.5194/egusphere-egu25-12104, 2025.

EGU25-12104

Download Close

In recent years, large wildfires have spread in Arctic regions as a consequence of ongoing climate change. Arctic organic soils are comparatively shallow but may be ancient, thus thousands of years old carbon may be released in smoldering and deeply burning fires. In Greenland, a land known for its icy expanse, fires are extremely rare. However, in summer 2019, the second-largest wildfire ever recorded on the island occurred at the Kangerluarsuk Tulleq fjord in southwestern Greenland. This study aims to produce pioneering in-field data on this tundra fire, focusing on three key aspects: 1) combustion, 2) burn depth, and 3) the age of the carbon released. Understanding whether the released carbon is modern or old is crucial due to different implications for the global carbon cycle and climate. To estimate carbon losses from the Kangerluarsuk Tulleq tundra fire, we established 14 sampling plots in burned areas and at unburned control sites. The selection of sampling plots was guided by a differenced Normalized Burn Ratio (dNBR) map generated using Sentinel-2 data and field reconnaissance. Within each plot, we assessed fire severity to estimate the above-ground carbon loss. For below-ground carbon loss estimation and burn depth analysis, organic soil samples were collected at burned plots and compared with unburned ones. To explore the vegetation succession and burned vegetation type, organic soil profiles (n=10) were extracted down to the mineral ground using a soil box corer and were studied by light-microscopy. Subsamples (n=20) from burned soil horizons were selected for radiocarbon dating to determine the age of carbon released in the fire. Our preliminary results suggest that soil carbon loss was higher than previously reported at an Alaskan tundra fire site with a mean value of 6.718 ± 0.9 kg of C m^-2. The mean burn depth was 9.0 ± 1.8 cm, and soil thaw depths during the 2024 summer were approximately 24 cm deeper in the 2019 burned area compared to unburned tundra. Expected radiocarbon results will indicate the maximum age of the carbon released by the fire. Vegetation succession measurements show that post-fire surfaces were predominantly colonized by pioneering non-Sphagnum bryophytes, Cyperaceae, and Ericaceae. The acquired results are first of a kind from a Greenland tundra fire and produce essential data for global climate modeling to assess the climate impacts of increasing Arctic wildfires.

How to cite: Granqvist, S., Diaz, L., Veraverbeke, S., Pilkama, E., Väliranta, M., and Ruppel, M.: Carbon emissions of an unprecedented Greenland wildfire, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15318, https://doi.org/10.5194/egusphere-egu25-15318, 2025.

EGU25-15318

Download Close

The Copernicus Atmosphere Monitoring Service CAMS is using ECMWF's Integrated Forecasting System IFS-COMPO with fire emissions from its Global Fire Assimilations System GFAS to monitor and forecast the effect of smoke from vegetation fires, resp. biomass burning, on atmospheric composition. The simulated atmospheric composition fields are routinely validated against observations including from satellites, aircraft and ground stations.

The emissions calculation by the operational GFAS version 1.2 have recently been updated for use in the upcoming HTAP3 multi-model, multi-pollutant study of fire impacts (Whaley et al. 2024), creating the dataset GFAS4HTAP. It is based on the dry matter burnt estimates of GFASv1.2, and uses an updated spurious signal mask, ESA CCI land cover data for 2018, a global peat map (Xu et al. 2018) and emission factors from NEIVA (Shahid et al. 2024) to calculate emission fluxes for various smoke constituents for 2003-2024. An additional GFAS-based dataset has been created by calibration against GFED5beta.

Global comparisons of dry matter, resp. biomass, combustion rates of the three GFAS-based inventories with GFED4s, GFED5beta, and the two variants of FINN2.5 reveal that these inventories can be roughly classified into one group of "traditional" inventories with lower fire activity, resp. emissions, and another of "more recent" inventories with higher fire activity. The pyrogenic carbon monoxide emission estimates from an inversion of satellite observations of atmospheric composition (Zheng et al. 2019) lie between these two groups in terms of global annual values. However, at a global level, they are more consistent with the "more recent" inventories during the late boreal summer peak of the global fire activity and with the "traditional" inventories during periods of lower fire activity.

In order to gain more insight from independent validation, we here present simulations with IFS-COMPO for 2019 based on the three GFAS-based inventories and compare these with atmospheric observations of carbon monoxide, nitrogen dioxide and aerosol optical depth. We find that the best agreement of simulation and observations is achieved by different inventories for different regions, seasons and smoke constituents. However, the emissions of the GFAS4HTAP dataset appears to lead to the overall most balanced atmospheric composition simulation. This supports the group of "traditional" inventories mentioned above.

How to cite: Kaiser, J., Huijnen, V., Remy, S., Ytre-Eide, M. A., De Jong, M. C., Zheng, B., and Wiedinmyer, C.: Evaluation of fire emissions for HTAP3 with CAMS GFAS and IFS-COMPO, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8708, https://doi.org/10.5194/egusphere-egu25-8708, 2025.

EGU25-8708

Download Close

Vegetation fires are well known as an important source of aerosol particles originating from the combustion of carbonaceous material. Much less known is that these fires can also efficiently inject soil-dust particles into the atmosphere, raised by the strong fire-induced winds. These soil-dust particles and the likely co-emitted biogenic particles are potent cloud condensation nuclei (CCN) and ice nucleating particles (INPs), and can substantially alter the cloud microphysics and thus impact the Earth’s radiation budget. Fires are an integral component of the Earth system that affect different landscapes around the globe. As they are supposed to get more frequent and more severe along with the ongoing global warming, a better knowledge of these specific fire emissions is crucial to understand their impacts on weather and climate.

Therefore, this work investigates the potential of wildfires to emit soil-dust particles on a global scale as a part of the newly established Leibniz ScienceCampus “BioSmoke” (‘smoke and bioaerosols in a changing climate’). As this particular dust emission pathway is not considered by the state-of-the-art dust emission models, a parameterization describing fire-induced dust emission fluxes has been developed and implemented into the global aerosol-climate model ICON-HAM. Fire-dust emissions are modelled as a function of the fire radiative power (FRP), the ambient wind conditions, and further soil-surface properties, including the soil type and a vegetation-dependent surface roughness correction.

Multi-year ICON-HAM simulations have revealed that fire-related dust emissions can account for up to one fifth of the total global dust emissions with strong regional and seasonal variations, both as the result of a varying fire activity and the local soil-surface conditions that can foster or impede also fire-dust emission significantly. In regions where the classic wind-driven dust emissions from arid, unvegetated soil surfaces are rather low but wildfires occur frequently, e.g., in large parts of the Southern hemisphere, fire-related dust emissions can add substantially to the atmospheric aerosol load and affect the local radiation budget there.

How to cite: Wagner, R., Tegen, I., and Schepanski, K.: Vegetation fires as a source of soil-dust particles – a global model perspective, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8307, https://doi.org/10.5194/egusphere-egu25-8307, 2025.

EGU25-8307

Download Close

The global wildland fire management community faces pressing climate change and operational challenges and requires improved capabilities in existing modelling tools or the development of novel decision support tools to limit the negative impact of wildfires and to increase use of prescribed burning where appropriate. This presentation will discuss limitations in the existing approaches to incorporating fuel structure effects in different model types (empirical, semi-empirical, detailed physics-based). In particular, novel experimental data will be presented addressing previously identified limitations [1] in the description of surface fuel beds in one of the most widely-used semi-empirical models; the Rothermel model, which underpins many current operational models.

The Rothermel model [2] involves a conservation of energy approach, incorporating separate terms to describe energy release rate in the combustion zone (reaction intensity) and energy transferred to the unburnt fuel (propagating flux), and incorporates a number of empirical closure terms.  The reaction intensity is empirically based, with the underpinning experimental measurements described in Frandsen and Rothermel [3]. By measuring the mass loss rate in a section of a fuel bed, Frandsen and Rothermel were able to characterize the intensity distribution within the combustion zone. However, the interacting effects of simultaneously varying fuel loading and packing ratio were not systematically considered, complicating efforts to understand the interacting effects of fuel loading and bulk density.

This study presents a series of laboratory-based flame spread experiments (no wind) involving excelsior fuel beds of varying structural conditions (Fuel Height: 0.02 to 0.12 m, Bulk Density: 3.3 to 20 kg/m³, Fuel Loading: 0.2 to 0.4 kg). The reaction intensity was calculated via a similar procedure to that described by Frandsen & Rothermel [2] as ‘Method 2’, in which the longitudinal length of the mass measurement region is greater than or equal to the combustion zone depth.

Clear trends in the peak mass loss rate and profile with bulk density were observed with a significant reduction at lower fuel loadings (0.2 kg/m²), and the reaction time was observed to increase at higher bulk densities along with a lengthening in the reaction intensity distribution region (further behind the combustion wave front). These results, along with existing observations of the trailing, in-depth combustion region in porous fuel beds, can be used to further investigate the observed tendency for underprediction of spread rates when the Rothermel model is applied to compressed fuel bed scenarios and has practical implications for other fire behaviour modelling applications. For example, improved characterisation of the overall combustion wave may enable improved modelling of smoke generation, surface-to-crown fire transition, and fuel consumption (e.g. to evaluate prescribed fire effectiveness).

[1] Z. Campbell-Lochrie, M. Gallagher, N. Skowronski, R.M. Hadden, The effect of fuel bed structure on Rothermel model performance, Int. J. of Wildland Fire. 33 (2023).

[2] R.C. Rothermel, A Mathematical Model for Predicting Fire Spread in Wildland Fuels, Research Paper INT-115, USDA Forest Service.,1972.

[3] W.H. Frandsen, R.C. Rothermel, Measuring the energy-release rate of a spreading fire, Combust Flame 19 (1972) 17–24.

How to cite: Campbell-Lochrie, Z.: Revisting Intensity of Combustion Waves to Address Outstanding Issues in Wildfire Modelling, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-1390, https://doi.org/10.5194/egusphere-egu25-1390, 2025.

EGU25-1390

Download Close

China is a significant region for crop cultivation. For a long time, there has been a common practice of burning crop residues during the post-harvest period (from May to October). The smoke emitted from straw burning contains both types of ozone precursors, including nitrogen oxides (NO_x=NO+NO₂) and volatile organic compounds (VOCs), and can be transported over long distances. During the transport process, secondary formation or consumption of ozone precursors occurs within the smoke plumes. After the smoke plume mixes with the atmosphere in the downwind urban area, it will lead to changes in the local ozone formation sensitivity. However, due to the nonlinear relationship between ozone and its precursors, the changes in ozone levels in downwind cities are not as straightforward as expected.

Here, we explore the temporal evolution of urban ozone and its precursors on smoke-affected days using multi-source satellite-derived fire event tracking datasets, which are screened by a semi-quantitative absorbing aerosol index (AAI), tropospheric NO₂ and HCHO columns measurements from OMI, fire points from Himawari-8, and ground-level O₃ monitoring dataset. We aimed to understand the associations between urban ground-level O₃ concentrations and crop residue burning events in China. Our analysis revealed that no consistent changes were shown in urban O₃ on smoke-affected days. In addition, there was an increase in NO₂, while HCHO and O₃ decreased in cities after mixing with smoke that had taken a long transport time. Our findings suggest that the O₃ formation sensitivity within aged smoke tends to be controlled by VOC-limited regime. We hypothesize that the large amount of NO_x carried by aged smoke consumes urban VOCs and O₃, while producing NO₂ locally. When fresh smoke, which is mainly controlled by the NO_x-limited regime, enters urban environments rich in NO_x, it leads to an increase in O₃ concentration. Our analysis may contribute to an improved understanding of the influence of straw burning on urban ozone levels in China.

How to cite: Wang, W., van der A, R., Ding, J., Cheng, T., and Wang, C.: Exploring the effect of straw burning on urban ozone levels based on multi-source satellites in northern China, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-1314, https://doi.org/10.5194/egusphere-egu25-1314, 2025.

EGU25-1314

Download Close

Forest and vegetation fires are one of the major sources of air pollution and have triggered air quality issues in many regions of Pacific Asia. Here we isolate the fire-specific PM_2.5 from monitoring concentrations using an observation-driven approach in the region. The total PM_2.5 in Pacific Asia exhibited a rapid declining trend from 2014 to 2021, while fire-specific PM_2.5 decreased in early years but begun to reverse, leading to an increasing proportions of fire-specific PM_2.5 in recent years. The inconsistency between the decreasing number of fire points and the rising levels of fire-specific PM_2.5 may be attributed to a shift in dominant sources of fire emissions in Pacific Asia, moving from anthropogenic agriculture fires to wildfires. Fire-related PM_2.5 poses a significant public health threat in Pacific Asia, contributing to approximately 334,300 premature deaths each year. Our assessment highlights the disproportionate impact of fire-specific PM_2.5 on poverty populations, indicating a pressing need for more attentions and researches in these regions. Based on the positive correlation between vapor pressure deficit and fire-specific PM_2.5, this study suggests that without further regulation and policy intervention, the contributions of fire-specific PM_2.5 to air pollution in Pacific Asia are likely to continue increasing under the influence of future climate change.

How to cite: Lu, H., Xie, M., Wang, N., and Liu, B.: The contribution of fires to PM2.5 and population exposure in Pacific Asia, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2467, https://doi.org/10.5194/egusphere-egu25-2467, 2025.

EGU25-2467

Download Close

In 2024, Bolivia experienced the worst year of fires since 2002, when Aqua MODIS began collecting data. According to estimates, more than 15 million hectares were burned this year. A sunphotometer sitting in the Bolivian lowlands recorded AOD values higher than two for several continuous days indicating the degradation of the air quality in the region. A unique set of instruments located in the Bolivian Andes recorded the transport of smoke produced by this biomass burning. Very high values of atmospheric tracers like carbon monoxide, equivalent black carbon, and others have been measured as high as 5240 m asl  at the Chacaltaya GAW station (CHC, 16.35ºS, 68.13ºW, 5240 m asl) and other sites around it both in the Altiplano and adjacent high altitude valleys. Although transport to these sites was observed previously, usually the events lasted one or two days. However, in 2024 longer periods of consecutive days with smoke arriving from the lowlands were observed for a second year in a row. Similar high values were observed in CHC in October of 2023, a year with less than half of fires in the country. The conditions that led to the transport of smoke to the mountains in the Andean Cordillera will be discussed, as well as the possible effects of the associated deforestation in terms of water availability for the central Andes.

How to cite: Andrade, M., Ticona, L., Velarde, F., Guzman, D., Blacutt, L., Forno, R., Gutierrez, R., Moreno, I., Avila, F., Uzu, G., Goloub, P., Ramonet, M., Laurent, O., Wiedensohler, A., Weinhold, K., Krejci, R., Aliaga, D., Whiteman, D., and Laj, P.: Intense transport of smoke to the Central Andes: Insights from a unique set of instruments located in the Bolivian Andean Cordillera, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-14573, https://doi.org/10.5194/egusphere-egu25-14573, 2025.

EGU25-14573

Download Close

South eastern Mediterranean region of Türkiye is well known with intense industrialization, shipping activities, agriculture and livestock production in addition to urban emission sources, thus struggle with significant air pollution problems. In addition to criteria pollutants, combination of these sources also results in high ammonia (NH₃) levels in the region.

NH₃ is released into the atmosphere mainly from agriculture, including nitrogen-based fertilizer applications and livestock farming as well as from several industries and from biomass burning. Atmospheric NH₃ plays a significant role in the formation of secondary inorganic particulate matter (PM), which negatively impacts on human health and ecosystems and indirectly influences climate change by altering radiative forcing. Climate change has increased the frequency and intensity of wildfires globally, which became another significant source of NH₃ over the Eastern Mediterranean, because the region is among the most sensitive regions. Besides wildfires, agricultural residue burning, although prohibited, also contributes to overall NH₃ levels.

Biomass burning contributes to atmospheric pollutants, as the combustion process emits nitrogen and carbon compounds from organic matter. In this study, multi-satellite derived retrievals were utilized, including IASI Level-2 NH₃ and CO, TROPOMI Level-2 NO₂, CO, and HCHO along with VIIRS S-NPP Fire Radiative Power product to investigate biomass burning related NH₃ levels. Products were processed at a 1x1km² gridded resolution to analyse spatio-temporal variations from 2019 to 2023, especially focusing on intense fire time intervals. While NH₃ levels were generally high during the summer over the region, the 2021 summer stood out with exceptionally high levels, coinciding with intense wildfires recorded that year. Similarly, CO levels revealed elevated levels during the same period, further strengthening the common impact of these extreme events. Further, fires detected over some areas by the VIIRS product were associated with residue-burning practices, as the area predominantly consists of agricultural lands.

The aim of the study is to investigate the impacts of fire-related NH₃ levels and quantify NH₃ enhancements during these fire events in the region. In this context, NH₃ to other pollutant ratios will be examined and temporal variation between different biomes will be classified. Air quality and climate change impact studies over the Mediterranean are critically important, with the absence of ground-based NH₃ measurements, satellite retrievals have to be utilized more to investigate the sensitivity of the region to extreme biomass burning events with the growing impacts of climate change.

Keywords: ammonia, carbon monoxide, nitrogen dioxide, biomass burning, wildfires

Acknowledgements: IASI is a joint mission of EUMETSAT and the Centre National d'Etudes Spatiales (CNES, France). The authors acknowledge the AERIS data infrastructure for providing access to the IASI data in this study and ULB-LATMOS for the development of the retrieval algorithms. This study was supported by the Scientific and Technological Research Council of Türkiye under the grant number 123Y364.

How to cite: Saracoglu, S., Alban, A. M., Tokgoz, S., and Kaynak, B.: Investigation of the intense wildfire events and NH3 levels over the Eastern Mediterranean, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-17646, https://doi.org/10.5194/egusphere-egu25-17646, 2025.

EGU25-17646

Download Close

For thirty years, the European Research Infrastructure IAGOS (In-Service Aircraft for a Global Observing System) has been equipping commercial aircraft with instruments to measure atmospheric composition on long-haul flights around the world.  Ten aircraft are currently equipped with IAGOS instruments to measure ozone, and the precursors carbon monoxide and nitrous oxides from the surface to the upper-troposphere during landing and take-off at worldwide airports,  and at cruise altitude where we observe the long-range transport of polluted airmasses. We analyse the transport of biomass burning pollutants from the intense Canadian wildfire seasons of 2023 and 2024 which impacted air-quality in North America and in Europe, and the extreme wildfires over the Amazon in 2024 that impacted air quality in South American cities.  The significance of these events is interpreted within the context of the 30-year climatology. The events will be compared with forecasts and analyses from the Copernicus Atmosphere Monitoring Service's global and regional models (projects CAMS2_82 and CAMS2_83) and we further  highlight the role of IAGOS  in developing air-quality networks in susceptible urban areas (project RI-URBANS) and the impacts of heatwaves and wildfires on air-quality in a changing climate (project IRISCC).

How to cite: Bennouna, Y., Clark, H., Wolff, P., Thouret, V., Blot, R., Nédélec, P., and Boulanger, D.: Impact of wildfires on air quality as seen by IAGOS in-situ measurements, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-18252, https://doi.org/10.5194/egusphere-egu25-18252, 2025.

EGU25-18252

Download Close

Wildfires exert a important influence on the chemical composition of the atmosphere, thereby impacting air quality, ecosystem, and climate forcing. The substantial emission of pollutants from such fires, coupled with their long-range transport, has the potential to counteract the progress achieved in reducing anthropogenic emissions. Numerous studies show that the increase in the frequency and intensity of fires offsets the general trend towards improved air quality observed in regions influenced by wildfires. These studies also caution of an increasing risk of the population being exposed to extreme levels of aerosols and ozone. In addition to their regional impacts, the plumes from the most intense fires can be transported on a continental or even hemispheric scale, thereby imposing health constraints on regions that are not generally affected by widespread, frequent or intense fires.

The northern hemisphere is home to a group of biomes that are particularly sensitive to hydro-meteorological conditions, and therefore to the effects of climate change on burned areas. The majority of the most intense fires of the last two decades have occurred in North America and in the boreal regions of Asia.

This study assesses the impact of fires on the variability of total CO, total PAN and AOD in the Northern Hemisphere using 16 years (2008-2023) of observations from the IASI/MetOp and MODIS/Terra and Aqua satellite instruments. More specifically, the variability in the number of detected plumes of extreme values of CO, PAN and aerosol from fires is studied.

The trajectories of these plumes are estimated using only satellite observations and are used to assess the contribution of the different regions of the Northern Hemisphere to the variability of atmospheric composition. The potential impact of the long-range transport of the identified plumes on air quality is estimated using observations of the altitude of the plumes obtained from both active CALIOP observations and passive IASI observations.

The chemical composition of the identified plumes is characterised using IASI observations of ammonia (NH₃), formic acid (HCOOH), methanol (CH₃OH) and ozone (O₃).

How to cite: Ehret, A., Turquety, S., Lecomte, G., Franco, B., George, M., Clarisse, L., Van Damme, M., Clerbaux, C., and Coheur, P.: Satellite observation of long-range transport of wildfires plumes in the northern hemisphere in 2008-2023, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-18688, https://doi.org/10.5194/egusphere-egu25-18688, 2025.

EGU25-18688

Download Close

Globally, approximately half of landscape fire emissions originate from savannas and grasslands. Furthermore, our observations in South Africa indicated major secondary aerosol formation in near-fire plume ageing. However, the measurements in South Africa are affected by anthropogenic emissions from the Highveld region, except for a clean sector towards the semi-arid Karoo region. Aiming for a savanna environment with minimal anthropogenic influence we set up a new measurement site in the Okavango delta area in northern Botswana in August 2024.

For the active savanna fire season in 2024, we operated online measurements of aerosol chemical composition with an aerosol chemical speciation monitor (ACSM), an online gas chromatograph coupled to an MS detector (GC-MS) for volatile organic compounds and a single particle soot photometer (SP2) for refractive BC. Measurements of aerosol particle size distribution with a differential mobility particle sizer (DMPS), aerosol absorption with a multi angle absorption photometer (MAAP), as well as CO and CO₂ concentrations will continue for the next couple of years at least.

For fresh plumes, initial analysis shows a strong decrease in submicron aerosol emission factor (EF_PM1) with increasing modified combustion efficiency, i.e. with increasing flaming fraction. The EF_PM1 values are in good agreement with previous observations in southern African savanna and with recent laboratory experiments that we carried out in collaboration with University of Eastern Finland. Analysis of ageing effects on the fire plumes in a clean savanna environment is ongoing.

How to cite: Vakkari, V., Mmereki, B. T., Koolebogile, D., van Niekerk, C. P. E., Le, V., Letsatle, M., Jaars, K., and van Zyl, P. G.: A new measurement site in northern Botswana to observe savanna fire plumes, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19104, https://doi.org/10.5194/egusphere-egu25-19104, 2025.

EGU25-19104

Download Close