Filippe Lemos Maia Santos

The year 2024 was classified by the World Meteorological Organization (WMO) as the hottest on record, both globally and in Brazil. In the Pantanal, exceptional heat waves and intense droughts led to an early onset of the fire season in June, marked by high fire severity and large burned areas. Although the Pantanal has been increasingly affected by combined drought and heat events, the influence of these climatic extremes on fire activity and the associated atmospheric circulation patterns is not well understood. This study assesses the conditions prior to the fire and the simultaneous occurrence of heat waves and droughts based on satellite data and climate indices; 2024 was identified as the most severe drought year over the 1980–2024 period, where virtually the entire biome reached exceptional drought levels. Heat waves were accompanied by significant precipitation and soil moisture deficits, along with the absence of the flood pulse in the rivers, reinforcing the compound nature of the 2024 extreme events. Synoptic conditions revealed patterns favourable to persistent warming and intensified drought, while fire risk indices exceeded the 99th percentile. These results highlight the need for a holistic approach when analysing the climatic factors driving fires, thereby advancing understanding of compound events and supporting public policies aimed at both mitigating and adapting to the impacts of climate extremes in the Pantanal.

The 2025 fire season in Portugal was marked by large fires, underscoring the vulnerability of the forested areas to fire. The study analyzes the main meteorological conditions during a critical period of fire activity and addresses the following question: Why can the northeast (NE) weather pattern be so critical for fire danger in Portugal? Fire severity in the Arouca wildfire, the largest fire of the period, was estimated using a methodology that integrates foundation vision models with computer vision algorithms. ECMWF analyses and convection-permitting Meso-NH simulations are used to examine large-scale circulation and the mesoscale environment, respectively. Synoptic-scale analysis revealed the Azores anticyclone centered slightly northwest of the Iberian Peninsula (IP), with its eastern sector directly affecting the northern IP under north/northeast winds. The hectometric-scale simulation demonstrated that orographically enhanced wind gusts over the northern Portuguese mountains substantially intensified near-surface fire-weather conditions when the winds were nearly easterly. Furthermore, strong low-level winds and atmospheric stability constrained vertical plume growth, favoring horizontal smoke transport. In addition, the study highlights that Arouca’s fire had 88% of its area affected with moderate to high severity. Overall, the results demonstrate that the interaction between large-scale NE circulation and local orography plays a decisive role in amplifying fire danger in northern Portugal, emphasizing the need for high-resolution atmospheric modeling to identify fire-prone regions under specific synoptic patterns.

This study investigates the Serra da Estrela megafire, located in Portugal, which occurred in August 2022, aiming to explore how changes in weather conditions can modify fire behavior in a fire burning during several days. The study uses a set of fire–atmosphere coupled simulations, referred as EXPn. The MesoNH atmospheric model was coupled to the ForeFire fire propagation model and configured with three nested domains, each comprising a 150 × 150 horizontal grid, with resolutions of 1500 m, 500 m, and 250 m. Regarding the experiments, EXP1 [IGN] corresponds to the ignition phase on 6 August, EXP2 [PYRO] to the development of pyro‐convective clouds on 10 August, whereas EXP3 [REAC] represents the fire reactivation period on 15 August. The first and third periods showed that smoke plumes were transported mainly within the boundary layer, whereas in the second period pyro‐convection reached the middle tropospheric levels, allowing the formation of pyro‐convective clouds. Such clouds were confirmed from radar and satellite observations. Beside the fire‐generated clouds, the study highlighted that propagation of the fire front can suddenly shift due to the interaction between weather, fire and complex terrain. This study confirmed the importance of the use of coupled atmosphere–fire models to represent fire dynamics and their impact on local meteorology. In a global perspective, such an approach provides a framework for assessing wildfire–atmosphere interactions in diverse environments, supporting improved prediction of large wildfires. Overall, it underscores the influence of evolving atmospheric conditions on fire behavior, which is critical for understanding the development of megafires capable of burning over several weeks. This study highlights the crucial need to expand our scientific understanding of fire behavior, enabling its application to other regions worldwide.

A major fire occurred in the wildland–urban interface in southern Portugal, on 13 July 2022, becoming uncontrolled due to weather conditions. This study investigates how atmospheric dynamics increased fire danger in Mainland Portugal during early July 2022. The synoptic circulation from European Centre for Medium-Range Weather Forecasts (ECMWF) analysis and mesoscale conditions from Meso-NH model simulation at 1.5 km resolution revealed atmospheric conditions before and during the fire. Fire risk was assessed using the Fire Weather Index (FWI) from Meso-NH outputs. A blocking pattern was configured by an upper-level low-pressure system in early July, remaining semi-stationary west of Mainland Portugal until 18 July. The counter-clockwise circulation of the cut-off low resulted in dry, warm air advection from North Africa, enhancing fire danger over the Iberian Peninsula. In southern Portugal, a jet-like wind with strong east/southeasterly flow from Gibraltar Strait favored rapid fire spread. This circulation below 1 km altitude from the Mediterranean Sea enhanced fire danger through strong winds, independent of the large-scale blocking pattern. This study presents an atmospheric scenario for evaluating fire danger in Southern Portugal, important for pre-firefighting management that complemented previous studies for the region. Also, high-resolution FWI calculations using Meso-NH emphasized the importance of improved temporal and spatial resolution for fire danger assessment.

The study discusses in an exploratory way the atmospheric conditions leading to several fires in the Pantanal on November 12, 2023. These episodes were marked by two periods of maximum fire expansion, first in the early afternoon and another in the evening. The study is based on a set of observations from satellites and weather stations, which helped to identify these fires and some meteorological conditions at the surface. In addition, the Fire Weather Index (FWI) in the Pantanal was analyzed for a 44-year period. However, this dataset was not enough to completely explain the fire behavior on that day. In this context, atmospheric modeling was applied to find the possible causes of fire behavior in these two periods. The Meso-NH model was run with two nested domains of horizontal resolutions of 2500 m and 500 m. The results showed a positive trend of FWI in the last decades and a clear seasonality for the maximum values in 2023. Also, the simulation indicated favorable weather conditions for fire ignition, given the high temperatures and low relative humidity. However, the wind gust field showed moderate gusty winds in both periods but caused by different forces. In the early afternoon, the larger scale circulation favored fire propagation, whereas in the evening fire spread by a gust front. Furthermore, the findings highlight the role of weather conditions on a sub-daily scale, with sudden changes in surface winds throughout the day, which is strongly recommended to be considered when examining fire danger and firefighting action plans in the region.

O estudo discute de forma exploratória as condições atmosféricas favoráveis à evolução dos incêndios florestais no Pantanal em 12 de novembro de 2023. Esses episódios foram marcados por dois períodos de rápida expansão do fogo, primeiro no início da tarde e outro à noite. O estudo usa um conjunto de observações de satélite e estações meteorológicas, as quais ajudaram a identificar o fogo e algumas condições meteorológicas na superfície. Além disso, o Fire Weather Index (FWI) no Pantanal foi analisado para um período de 44 anos. No entanto, esse conjunto de dados não foi suficiente para explicar completamente o comportamento do fogo naquele dia. Nesse contexto, a modelagem atmosférica foi aplicada para encontrar as possíveis causas do comportamento do fogo em dois períodos. O modelo Meso-NH foi configurado com dois domínios aninhados e resoluções horizontais de 2500 m e 500 m. Os resultados mostraram uma tendência positiva do FWI nas últimas décadas, bem como uma clara sazonalidade para os valores máximos no ano de 2023. A simulação indicou condições favoráveis à ignição do fogo, e o campo de rajadas de vento mostrou ventos moderados em ambos os períodos, mas causados por diferentes forçantes. No início da tarde, a circulação em grande escala favoreceu a propagação do fogo, enquanto à noite uma frente de rajada foi observada. O estudo destaca o papel das condições meteorológicas na escala sub-diária, em particular para mudanças repentinas do vento à superfície ao longo do dia. Esse resultado deve ser considerado ao examinar o perigo de fogo e o planejamento das ações de combate aos incêndios na região.

Wildfires are among the most powerful natural phenomena that have severe environmental and social impacts. Their increasing frequency and intensity, driven by climate change and unsustainable land-use practices, emphasise the importance of reliable fire risk assessment tools. Live fuel moisture content (LFMC) is a key variable for wildfire risk assessment and management, as it directly influences fire behaviour and propagation. However, retrieving this information is challenging owing to its high spatial and temporal variability, combined with time-consuming and expensive in situ sampling methods. This study introduces a novel approach to estimate the LFMC from a Land Surface model (LSM). This approach was achieved by integrating remote sensing, numerical modelling, and a machine learning (ML) framework. Land surface model simulations were conducted to generate surface and soil conditions, which served as predictors in an ML-based regressor model to estimate the LFMC. The proposed model demonstrated robust performance, with a coefficient of determination (r ² ) of 0.72 and an absolute root mean square error (RMSE) of 11.6%. This approach produces reliable LFMC estimates with high spatial resolution, which can be used in wildfire propagation models. Finally, this study highlights a novel model to produce LFMC information, which can significantly enhance wildfire management and support more informed decisions in fire prevention and action strategies. Graphical abstract

Portugal is one of the European countries that faces significant challenges with wildfires. While lightning-triggered natural fires constitute a minority compared to anthropogenic ones, accurate forecasting of lightning occurrences is crucial for effective prevention. The study assesses the ECMWF model's capability to predict lightning in Portugal over four fire seasons [2019-2022]. Observed lightning data was obtained from the national lightning detector network, aggregated into 0.5° and 1° resolutions over 3-hour periods. The evaluation employs statistical indices from a contingency table to analyze the model's performance. Results indicate an overestimation of lightning occurrences by the ECMWF model, with a Bias greater than 1. The success rate for lightning prediction was 57.7% for a horizontal resolution of 1° and 49% for 0.5°. Additionally, the temporal analysis reveals a time lag between both data, with the model starting to predict lighting before its occurrence and finishing the prediction earlier. These findings are complemented by analyzing the spatial lightning distribution, which led us to identify some weather patterns associated with lightning activity during the study period. For instance, lightning activity was associated with the Iberian thermal low development overlapped by an Upper Level Low and the passage of large-scale features, such as frontal systems. The insights gained from this study have implications for the ECMWF lightning forecast applicability in the context of forecasting natural forest fires in Portugal. The research was funded by the European Union through the CILIFO project (0753-CILIFO-5-E) and also by national funds through FCT Foundation for Science and Technology, I.P. under the PyroC.pt project (PCIF/MPG/0175/2019).