Burnt Forest Areas Research Articles

New metrics for governance in the era of earth observation data: Monitoring violations after wildfires

Abstract This study uses Earth observation data to measure illegal activities and investigates possible relationships with local governments. We have collected satellite images, digital maps, and geospatial data for over a decade to detect potential illegal constructions in protected burned forest areas in Sardinia, Italy. We create a database of buildings erected in these protected areas and connect it to administrative data on local election results. First, we examine the climatic and geographical factors that contribute to wildfires. Our findings indicate that fires tend to spread more in places with higher altitudes and greater distances from urban centers. This highlights the challenges that local authorities face in monitoring and intervening in areas that are less accessible. Next, we analyze the relationship between mayoral turnover and constructions erected in protected burned forests to highlight how this phenomenon influences voters' preferences. We observe that citizens express their disapproval of these illegal activities through the ballot box. Violations in the years leading up to local elections increase the likelihood of a change in mayoral leadership within a municipality.

FBA-DPAttResU-Net: Forest burned area detection using a novel end-to-end dual-path attention residual-based U-Net from post-fire Sentinel-1 and Sentinel-2 images

Forest burned area (FBA) detection using remote sensing (RS) data is critical for timely forest management and recovery attempts after wildfires. This study introduces a dual-path attention residual-based U-Net (DPAttResU-Net), a novel end-to-end deep learning (DL) model tailored for FBA detection using dual-source post-fire Sentinel-1 (S1) and Sentinel-2 (S2) satellite RS imagery. To better distinguish FBAs from other land cover types, DPAttResU-Net incorporates a dual-pathway structure to exploit complementary geometrical/physical and spectral features from S1 and S2, respectively. An integral component in the proposed architecture is the channel-spatial attention residual (CSAttRes) block, which emphasizes salient features through the channel and spatial attention modules, thus improving the burned area feature representation. To compare DPAttResU-Net to state-of-the-art DL models, experiments were conducted on benchmark FBA datasets collected over 12 areas, where ten datasets were used as training data and two datasets were used to test the trained DL models. The experimental results demonstrate the high proficiency of the proposed deep model in meticulously delineating FBAs. In further detail, DPAttResU-Net, with a PFN of 17.96 (%) in the first case and an IoU of 89.31 (%) in the second case, outperformed the existing U-Net-based models. Accordingly, through dual-path integration and attention mechanisms, DPAttResU-Net contributes to accurately identifying FBAs by preserving their geometrical details, making it a promising tool for post-wildfire forest management.

From conflict to coexistence: Reaffirming belonging and property rights through the ‘sons of village’ discourse in post-conflict Kenya

This article explores the puzzle of victims of political violence in Kenya committing to return to contested spaces of their prior victimhood and displacement. It considers how political violence has been brought to bear on understandings of property rights and belonging among Kikuyu victims of political violence in the Burnt Forest area of Uasin Gishu County in Kenya. It is reported that the iteration between the collapse of the multiparty Kenyan state commitment to protect Kikuyu land rights in their state settled areas and the partisan character of neo-customary tenure that restricts the admission of co-ethnic outsiders, induced Kikuyu spontaneous resistance to their spatial political confinement motivating their formulation of an organic discourse of belonging. The article introduces the ‘sons of village’ concept as a bottom-up framework for understanding the informal mechanisms for claiming property rights and belonging in contested spaces in Africa. By challenging notions of belonging rooted in contested histories and emphasizing credible links to land and space, this concept embodies inclusive citizenship with the potential to foster conciliatory relations between previously hostile groups in post-conflict scenarios. The 'sons of village' identification, therefore, offers a promising avenue for fostering positive peace in regions afflicted by chronic violence in Africa and beyond.

Anticipating Future Risks of Climate-Driven Wildfires in Boreal Forests

Extreme forest fires have historically been a significant concern in Canada, the Russian Federation, the USA, and now pose an increasing threat in boreal Europe. This paper deals with application of the wildFire cLimate impacts and Adaptation Model (FLAM) in boreal forests. FLAM operates on a daily time step and utilizes mechanistic algorithms to quantify the impact of climate, human activities, and fuel availability on wildfire probabilities, frequencies, and burned areas. In our paper, we calibrate the model using historical remote sensing data and explore future projections of burned areas under different climate change scenarios. The study consists of the following steps: (i) analysis of the historical burned areas over 2001–2020; (ii) analysis of temperature and precipitation changes in the future projections as compared to the historical period; (iii) analysis of the future burned areas projected by FLAM and driven by climate change scenarios until the year 2100; (iv) simulation of adaptation options under the worst-case scenario. The modeling results show an increase in burned areas under all Representative Concentration Pathway (RCP) scenarios. Maintaining current temperatures (RCP 2.6) will still result in an increase in burned area (total and forest), but in the worst-case scenario (RCP 8.5), projected burned forest area will more than triple by 2100. Based on FLAM calibration, we identify hotspots for wildland fires in the boreal forest and suggest adaptation options such as increasing suppression efficiency at the hotspots. We model two scenarios of improved reaction times—stopping a fire within 4 days and within 24 h—which could reduce average burned forest areas by 48.6% and 79.2%, respectively, compared to projected burned areas without adaptation from 2021–2099.

Temporal and Spatial Analyses of Forest Burnt Area in the Middle Volga Region Based on Satellite Imagery and Climatic Factors

Wildfires are important natural drivers of forest stands dynamics, strongly affecting their natural regeneration and providing important ecosystem services. This paper presents a comprehensive analysis of spatiotemporal burnt area (BA) patterns in the Middle Volga region of the Russian Federation from 2000 to 2022, using remote sensing time series data and considering the influence of climatic factors on forest fires. To assess the temporal trends, the Mann–Kendall nonparametric statistical test and Theil–Sen’s slope estimator were applied using the LandTrendr algorithm on the Google Earth Platform (GEE). The accuracy assessment revealed a high overall accuracy (>84%) and F-score value (>82%) for forest burnt area detection, evaluated against 581 reference test sites. The results indicate that fire occurrences in the region were predominantly irregular, with the highest frequency recorded as 7.3 over the 22-year period. The total forest BA was estimated to be around 280 thousand hectares, accounting for 1.7% of the land surface area or 4.0% of the total forested area in the Middle Volga region. Coniferous forest stands were found to be the most fire-prone ecosystems, contributing to 59.0% of the total BA, while deciduous stands accounted for 25.1%. Insignificant fire occurrences were observed in young forests and shrub lands. On a seasonal scale, temperature was found to have a greater impact on BA compared with precipitation and wind speed.

Deep learning-based burned forest areas mapping via Sentinel-2 imagery: a comparative study.

In order to evaluate the effects of forest fires on the dynamics of the function and structure of ecosystems, it is necessary to determine burned forest areas with high accuracy, effectively, economically, and practically using satellite images. Extraction of burned forest areas utilizing high-resolution satellite images and image classification algorithms and assessing the successfulness of varied classification algorithms has become a prominent research field. This study aims to indicate on the capability of the deep learning-based Stacked Autoencoders method for the burned forest areas mapping from Sentinel-2 satellite images. The Stacked Autoencoders, used in this study as an unsupervised learning method, were compared qualitatively and quantitatively with frequently used supervised learning algorithms (k-Nearest Neighbors (k-NN), Subspaced k-NN, Support Vector Machines, Random Forest, Bagged Decision Tree, Naive Bayes, Linear Discriminant Analysis) on two distinct burnt forest zones. By selecting burned forest zones with contrasting structural characteristics from one another, an objective assessment was achieved. Manually digitized burned areas from Sentinel-2 satellite images were utilized for accuracy assessment. For comparison, different classification performance and quality metrics (Overall Accuracy, Mean Squared Error, Correlation Coefficient, Structural Similarity Index Measure, Peak Signal-to-Noise Ratio, Universal Image Quality Index, and KAPPA metrics) were used. In addition, whether the Stacked Autoencoders method produces consistent results was examined through boxplots. In terms of both quantitative and qualitative analysis, the Stacked Autoencoders method showed the highest accuracy values.

Assessment for the severity of forest areas burnt by fire in the Phu Kradueng National Park by retrieving data from the Landsat 8 OLI Satellite

Forest fire is a problem that has been more and more severe due to the changing climate and other factors caused by humans. This study aimed at developing operation forms that are up-to-date for the analysis of and assessment of the severity of forest fire burnt fire in the Phu Kradueng National Park by retrieving data from the Landsat 8 OLI Satellite. The operation methods are as follows: 1) bring the data from the Landsat 8 OLI Satellite before and after forest fire incident to be adjusted on part of the wave radiation, 2) perform the analysis to find the Normalized Burn Ratio (NBR) and ∆NBR, 3) assess for the severity of forest areas burnt by forest fire in 7 levels, and 4) check for the accuracy of forest areas burnt by forest fire by using the method of visual interpretation into the data obtained from the Landsat 8 OLI Satellite that has passed the False Color Composite process from SWIR - NIR - Red wave ranges (RGB:754); the location was specified by creating 90 random points, dividing into forest area for 30 points, forest area burnt by forest fire for 30 points, and other areas that are not forest for 30 points. After that, consider the overall accuracy and Kappa statistics. It was found from the operation that NBR and ∆NBR can be well used to assess for the severity of the burning, and are consistent with the situations that occur. Also, there is consistency in the accuracy of the burnt forest area from forest fire by the method of visual interpretation.

Community engagement to inform development of strategies to improve referral for hypertension: perspectives of patients, providers and local community members in western Kenya

BackgroundHypertension is the leading cause of death and disability. Clinical care for patients with hypertension in Kenya leverages referral networks to provide basic and specialized healthcare services. However, referrals are characterized by non-adherence and delays in completion. An integrated health information technology (HIT) and peer-based support strategy to improve adherence to referrals and blood pressure control was proposed. A formative assessment gathered perspectives on barriers to referral completion and garnered thoughts on the proposed intervention.MethodsWe conducted a qualitative study in Kitale, Webuye, Kocholya, Turbo, Mosoriot and Burnt Forest areas of Western Kenya. We utilized the PRECEDE-PROCEED framework to understand the behavioral, environmental and ecological factors that would influence uptake and success of our intervention. We conducted four mabaraza (customary heterogenous community assemblies), eighteen key informant interviews, and twelve focus group discussions among clinicians, patients and community members. The data obtained was audio recorded alongside field note taking. Audio recordings were transcribed and translated for onward coding and thematic analysis using NVivo 12.ResultsSpecific supply-side and demand-side barriers influenced completion of referral for hypertension. Key demand-side barriers included lack of money for care and inadequate referral knowledge. On the supply-side, long distance to health facilities, low availability of services, unaffordable services, and poor referral management were reported. All participants felt that the proposed strategies could improve delivery of care and expressed much enthusiasm for them. Participants appreciated benefits of the peer component, saying it would motivate positive patient behavior, and provide health education, psychosocial support, and assistance in navigating care. The HIT component was seen as reducing paper work, easing communication between providers, and facilitating tracking of patient information. Participants also shared concerns that could influence implementation of the two strategies including consent, confidentiality, and reduction in patient-provider interaction.ConclusionsAppreciation of local realities and patients’ experiences is critical to development and implementation of sustainable strategies to improve effectiveness of hypertension referral networks. Incorporating concerns from patients, health care workers, and local leaders facilitates adaptation of interventions to respond to real needs. This approach is ethical and also allows research teams to harness benefits of participatory community-involved research.Trial registrationClinicaltrials.gov, NCT03543787, Registered June 1, 2018. https://clinicaltrials.gov/ct2/show/NCT03543787

Assessment of Spectral Vegetation Indices Performance for Post-Fire Monitoring of Different Forest Environments

Although wildfires are a common disturbance factor to the environment, some of them can cause significant environmental and socioeconomic losses, affecting ecosystems and people worldwide. The wildfire identification and assessment of their effects on damaged forest areas is of great importance for provision of effective actions on their management and preservation. Forest regrowth after a fire is a continuously evolving and dynamic process, and the accuracy assessment of different remote sensing indices for its evaluation is a complicated task. The implementation of this task cannot rely on the standard procedures. Therefore, we suggested a method involving delineation of dynamic boundaries between conditional categories within burnt forest areas by application of spectral reflectance characteristics (SRC). This study compared the performance of firmly established for fire monitoring differenced vegetation indices—Normalized Difference Vegetation Index (dNDVI) and Normalized Burn Ratio (dNBR) and tested the capabilities of tasseled cap-derived differenced Disturbance Index (dDI) for post-fire monitoring purposes in different forest environments (Boreal Mountain Forest (BMF), Mediterranean Mountain Forest (MMF), Mediterranean Hill Forest (MHF)). The accuracy assessment of the tree indices was performed using Very High Resolution (VHR) aerial and satellite data. The results show that dDI has an optimal performance in monitoring post-fire disturbances in more difficult-to-be-differentiated classes, whereas, for post-fire regrowth, the more appropriate is dNDVI. In the first case, dDI has an overall accuracy of 50%, whereas the accuracy of dNBR and dNDVI is barely 35% and 36%. Moreover, dDI shows better performance in 16 accuracy metrics (from 17). In the second case, dNDVI has an overall accuracy of 59%, whereas those of dNBR and dDI are 55% and 52%, and the accuracy metrics in which dNDVI shows better performance than the other two indices are 11 (from 13). Generally, the studied indices showed higher accuracy in assessment of post-fire disturbance rather than of the post-fire forest regrowth, implicitly at test areas—BMF and MMF, and contrary opposite result in the accuracy at MHF. This indicates the relation of the indices’ accuracy to the heterogeneity of the environment.

Characterization of global fire activity and its spatiotemporal patterns for different land cover types from 2001 to 2020

Fire is a widespread phenomenon that plays an important role in Earth's ecosystems. This study investigated the global spatiotemporal patterns of burned areas, daytime and nighttime fire counts, and fire radiative power (FRP) from 2001 to 2020. The month with the largest burned area, daytime fire count, and FRP presented a bimodal distribution worldwide, with dual peaks in early spring (April) and summer (July and August), while the month with the largest nighttime fire count and FRP showed a unimodal distribution, with a peak in July. Although the burned area showed decline at the global scale, a significant increase occurred in temperate and boreal forest regions, where nighttime fire occurrence and intensity have consistently increased in recent years. The relationships among burned area, fire count, and FRP were further quantified in 12 typical fire-prone regions. The burned area and fire count exhibited a humped relationship with FRP in most tropical regions, whereas the burned area and fire count constantly increased when the FRP was below approximately 220 MW in temperate and boreal forest regions. Meanwhile, the burned area and FRP generally increased with the fire count in most fire-prone regions, indicating an increased risk of more intense and larger fires as the fire count increased. The spatiotemporal dynamics of burned areas for different land cover types were also explored in this study. The results suggest that the burned areas in forest, grassland, and cropland showed dual peaks in April and from July to September while the burned areas in shrubland, bareland, and wetlands usually peaked in July or August. Significant increases in forest burned area were observed in temperate and boreal forest regions, especially in the western U.S. and Siberia, whereas significant increases in cropland burned area were found in India and northeastern China.

Assessment of Post-Fire Phenological Changes Using MODIS-Derived Vegetative Indices in the Semiarid Oak Forests

Wildfire has significant impact on plant phenology. The plants’ phenological variables, derived from time series satellite data, can be monitored and the changes in satellite imagery may be used to identify the beginning, peak, and end of the growing season. This study investigated the use of remote sensing data and land surface phenology (LSP) parameters to evaluate the impacts of fire. The LSP parameters included the start of growing season (SOS), the length of the growing season (LOS), the end of the growing season (EOS), maximum greenness of the season (Gmax), and minimum greenery in the season (Gmin) in the fire-impacted, semiarid oak forests of Iran. These LSP parameters were extracted from multitemporal normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI2) data, acquired from MODIS sensor images in Zagros of the Ilam province in western Iran. By extracting LSP indices from the NDVI and EVI2 data, the indices were compared between burned forest areas, areas surrounding the burned forests, and unburned areas and for timesteps representing pre-fire, fire (i.e., year of fire), and post-fire (i.e., 2 years) conditions. It was found that for the burned area, there were significant differences in Gmax and the day that Gmax occurred. Furthermore, there was also a significant difference in Gmin between the pre- and post-fire conditions when NDVI was used and a significant difference between Gmax when EVI2 was used. The results also showed that in both time series there was a significant difference between the burned and control area in terms of Gmax. In general, the results showed that the fire had a negative effect on LSP, but in the two years after the fire, there were signs of forest restoration. This study provides necessary information to inform forest and resource conservation and restoration programs.

An Unsupervised Saliency-Guided Deep Convolutional Neural Network for Accurate Burn Mapping from Sentinel-1 SAR Data

SAR data provide sufficient information for burned area detection in any weather condition, making it superior to optical data. In this study, we assess the potential of Sentinel-1 SAR images for precise forest-burned area mapping using deep convolutional neural networks (DCNN). Accurate mapping with DCNN techniques requires high quantity and quality training data. However, labeled ground truth might not be available in many cases or requires professional expertise to generate them via visual interpretation of aerial photography or field visits. To overcome this problem, we proposed an unsupervised method that derives DCNN training data from fuzzy c-means (FCM) clusters with the highest and lowest probability of being burned. Furthermore, a saliency-guided (SG) approach was deployed to reduce false detections and SAR image speckles. This method defines salient regions with a high probability of being burned. These regions are not affected by noise and can improve the model performance. The developed approach based on the SG-FCM-DCNN model was investigated to map the burned area of Rossomanno-Grottascura-Bellia, Italy. This method significantly improved the burn detection ability of non-saliency-guided models. Moreover, the proposed model achieved superior accuracy of 87.67% (i.e., more than 2% improvement) compared to other saliency-guided techniques, including SVM and DNN.

Application of Remote Sensing and GIS techniques for detecting burned areas and severity. A case study of the National Park "Dajti Mountain", Albania.

Assessment of forest areas affected by wildfire is crucial for designing appropriate management strategies to support post-wildfire restoration. This study integrates Remote Sensing and GIS data to map burned areas and severity, and regeneration of vegetation in a Mediterranean forest type ecosystem (National Park "Dajti Mountain", NPDM), in Albania. Landsat 8 satellite imagery was employed to calculate various spectral indices such as the Normal Burn Ratio Index (NBR), NBR2, the Normalized Difference Vegetation Index (NDVI) and the Enhanced Vegetation Index (EVI). Burn severity levels were defined by using the dNBR thresholds developed by Key and Benson (2006). The accuracy of burn severity map produced was evaluated by relating field-based Composite Burn Index (CBI) and satellite-derived metrics (dNBR) from Landsat-8. By means of dNBR and dNBR2 we detected and mapped several burned forest areas within the NPDM, at the sites of Shkallë, Qafëmolle, Ibë, Tujan, Derje, Selbë, Surrel and Dajt, which were affected by wildfire during the year 2017. The dNBR produced the best results for burned areas mapping and burn severity assessment (91.7%) over the dNBR2 (89.8%). The dNBR and dNBR2 index maps showed that a total of 103.59 and 105.72 hectares of forests was affected by wildfire. Areas with different levels of burn severity were detected: 17.29 and 23.80% unburned, 43.36 and 45% low, 15.11 and 12.13% moderate, 24.93 and 21.2% high. Overall, the dNBR2 index produced lower percentages of wildfire-affected areas at high and moderate rates compared to the dNBR index while for unburned areas the dNBR2 index resulted in higher percentages. Vegetation recovery during the subsequent growing season was generally good as revealed by the high dNDVI and dEVI values, indicating the reactivation of photosynthetic activity. This information is useful for forest managers/specialists to design relevant strategies for the proper rehabilitation/management of burned forest areas in the future.

Rethinking the focus on forest fires in federal wildland fire management: Landscape patterns and trends of non-forest and forest burned area

For most of the 20th century and beyond, national wildland fire policies concerning fire suppression and fuels management have primarily focused on forested lands. Using summary statistics and landscape metrics, wildfire spatial patterns and trends for non-forest and forest burned area over the past two decades were examined across the U.S, and federal agency jurisdictions. This study found that wildfires burned more area of non-forest lands than forest lands at the scale of the conterminous and western U.S. and the Department of Interior (DOI). In an agency comparison, 74% of DOI burned area occurred on non-forest lands and 78% of U.S. Forest Service burned area occurred on forested lands. Landscape metrics revealed key differences between forest and non-forest fire patterns and trends in total burned area, burned patch size, distribution, and aggregation over time across the western U.S. Opposite fire patterns emerged between non-forest and forest burns when analyzed at the scale of federal agency jurisdictions. In addition, a fire regime departure analysis comparing current large fire probability with historic fire trends identified certain vegetation types and locations experiencing more fire than historically. These patterns were especially pronounced for cold desert shrublands, such as sagebrush where increases in annual area burned, and fire frequency, size, and juxtaposition have resulted in substantial losses over a twenty-year period. The emerging non-forest fire patterns are primarily due to the rapid expansion of non-native invasive grasses that increase fuel connectivity and fire spread. These invasions promote uncharacteristic frequent fire and loss of native ecosystems at large-scales, accelerating the need to place greater focus on managing invasive species in wildland fire management. Results can be used to inform wildfire management and policy aimed at reducing uncharacteristic wildfire processes and patterns for both non-forest and forest ecosystems as well as identify differing management strategies needed to address the unique wildfire issues each federal agency faces.

Spatiotemporal patterns of fire-driven forest mortality in China

Wildfire is one of the most prevalent natural or human-induced disturbances across the world, which can kill trees directly during the active combustion phase or cause delayed forest mortality through interactions with other disturbance agents. With its diverse range of climate and forest ecosystems, China also has diverse forest fire regimes. However, the spatial and temporal patterns of fire-driven forest mortality on the national scale remain largely unexplored. In this study, we employed satellite observations and statistical models to investigate the spatiotemporal patterns and underlying drivers of fire-induced forest mortality across China. Our results demonstrate pronounced temporal patterns of forest burned area and fire-driven forest mortality. On the seasonal scale, forest fires mostly occurred in the winter and spring, whereas the forest mortality is highest from May to July. Both forest burned area and fire-driven mortality exhibited a significant decreasing trend during 2003–2016. Forest burned areas are mainly distributed in Northeast China, Southwest China and South China. There exists a tight spatial consistency between areas with high mortality and burned area with East China as an exception. Correlation analysis and multiple linear regression analysis show that fire size, fire spread rate, fire duration and drought intensity had significant impacts on postfire forest mortality, with fire size and fire spread rate accounting for 38.3% and 37.1%, respectively, of the model’s explanatory power. Our results highlight the importance of fire regime impacts on postfire forest mortality. Future climate change that drives more intense fire regimes might thus lead to enhanced forest mortality and increase the loss of forest ecosystem services.

Forest Burned Area Detection Using a Novel Spectral Index Based on Multi-Objective Optimization

Forest fires cause environmental and economic damage, destroy large areas of land and displace entire communities. Accurate extraction of fire-affected areas is of vital importance to support post-fire management strategies and account for the environmental impact of fires. In this paper, an analytical burned area index, called ABAI, was proposed to map burned areas from the newly launched Sentinel-2 images. The innovation of this method is to separate the fire scars from other typical land covers by formulating different objective functions, which involved three main components: First, spectral differences between the burned land and other land covers were characterized by analyzing the spectral features of the existing burned area indices. Then, for each type of land cover, we formed an objective function by linear combination of bands with the values of band ratios. Second, all the objective functions and possible constraints were formulated as a multi-objective optimization problem, and then it was solved using a linear programming approach. Finally, the ABAI spectral index was achieved with the optimizing coefficients derived from the multi-objective problem. To validate the effectiveness of the proposed spectral index, three experimental datasets, clipped from Sentinel-2 images at different places, were tested and compared with baseline indices, such as normalized burned area (NBR) and burned area index (BAI) methods. Experimental results demonstrated that the injection of a green band to the spectral index has led to good applicability in burned area detection, where the ABAI can avoid most of the confusion presented by shadows or shallow water. Compared to other burned area indices, the proposed ABAI achieved the best classification accuracy, with the overall accuracy being over 90%. Visually, our approach significantly outperforms other spectral indexed methods, especially in confused areas covered by water bodies and shadows.

Machine Learning-Based Forest Burned Area Detection with Various Input Variables: A Case Study of South Korea

Recently, an increase in wildfire incidents has caused significant damage from economical, humanitarian, and environmental perspectives. Wildfires have increased in severity, frequency, and duration because of climate change and rising global temperatures, resulting in the release of massive volumes of greenhouse gases, the destruction of forests and associated habitats, and the damage to infrastructures. Therefore, identifying burned areas is crucial for monitoring wildfire damage. In this study, we aim at detecting forest burned areas occurring in South Korea using optical satellite images. To exploit the advantage of applying machine learning, the present study employs representative three machine learning methods, Light Gradient Boosting Machine (LightGBM), Random Forest (RF), and U-Net, to detect forest burned areas with a combination of input variables, namely Surface Reflectance (SR), Normalized Difference Vegetation Index (NDVI), and Normalized Burn Ratio (NBR). Two study sites of recently occurred forest fire events in South Korea were selected, and Sentinel-2 satellite images were used by considering a small scale of the forest fires. The quantitative and qualitative evaluations according to the machine learning methods and input variables were carried out. In terms of the comparison focusing on machine learning models, the U-Net showed the highest accuracy in both sites amongst the designed variants. The pre and post fire images by SR, NDVI, NBR, and difference of indices as the main inputs showed the best result. We also demonstrated that diverse landcovers may result in a poor burned area detection performance by comparing the results of the two sites.

Изменения структуры и биоразнообразия лиственничных лесов в верхнем течении р. Хилок (бассейн оз. Байкал) под воздействием пожаров

The study results of changes in the larch forests’ structure and biodiversity in the upper Khilok River basin (right tributary of the Selenga River) affected by the climate change and forest fires in the period from 1996 to 2018 are presented. It was found that the area of closed forests decreased by an average of 51 %, the burnt forest area and burnt timber increased by 46%, dwarf birches decreased by 12 %, the natural reforestation of the sample areas is 90 % poor and unsatisfactory. It was revealed that due to fires in forest communities, the total number of plant species and lichens decreases, indices of species wealth and biodiversity concentration decrease in plant communities, and the ratio of species of latitudinal geographical groups (forest, steppe and meadow) changes. It is shown that with further climate aridization, an increase in the area of dwarf birches communities after fires can lead to the replacement of forests with shrubby, and later grassy communities. It is recommended to continue research on changes in the structure and biological diversity of plant communities in the selected model areas due to an increase in the amount of spring-summer precipitation in the region.

Remote Sensing of Forest Burnt Area, Burn Severity, and Post-Fire Recovery: A Review

Wildland fires dramatically affect forest ecosystems, altering the loss of their biodiversity and their sustainability. In addition, they have a strong impact on the global carbon balance and, ultimately, on climate change. This review attempts to provide a comprehensive meta-analysis of studies on remotely sensed methods and data used for estimation of forest burnt area, burn severity, post-fire effects, and forest recovery patterns at the global level by using the PRISMA framework. In the study, we discuss the results of the analysis based on 329 selected papers on the main aspects of the study area published in 48 journals within the past two decades (2000–2020). In the first part of this review, we analyse characteristics of the papers, including journals, spatial extent, geographic distribution, types of remote sensing sensors, ecological zoning, tree species, spectral indices, and accuracy metrics used in the studies. The second part of this review discusses the main tendencies, challenges, and increasing added value of different remote sensing techniques in forest burnt area, burn severity, and post-fire recovery assessments. Finally, it identifies potential opportunities for future research with the use of the new generation of remote sensing systems, classification and cloud performing techniques, and emerging processes platforms for regional and large-scale applications in the field of study.

Catastrophic PM2.5 emissions from Siberian forest fires: Impacting factors analysis

With increased forest fires due to climate change, PM2.5 emissions also intensified. Record PM2.5 emissions according to Copernicus Atmosphere Monitoring Service in Russia amounted to 8 megatons (Mt) in 2021, which is 78% higher than the average level of 2004–2021 (4.5 Mt). Seven federal subjects (the constituent entities) with vast forest areas without fire protection produced 86% of emissions (6.8 Mt) in 2021, the major losses (6.1 Mt) in Yakutia (Sakha Republic). The ambient temperature in Eastern Siberia is increasing, especially in months of winter and spring seasons (up to +3.6 °C) in 1990–2020 compared to 1901–2020 (CEDA Archive); climate change has affected meteorological conditions leading to increased forest fires. The results of the SARIMAX model study for PM2.5 emissions considering meteorological factors using ERA5 and burnt forest area using MODIS (MCD64A1), establishing a significant dependence of PM2.5 emissions on the lack of precipitation and the associated parameters of complete and potential evaporation. This influence long before the fire season (up to 9 months), as it affects the snow cover and the dryness of the fuel by the beginning of forest fires. In turn, high PM2.5 emission values are accompanied by a drop in 2 m air temperature and surface solar radiation downwards due to the aerosol saturation with suspended particles. The average COR for seven federal subjects was 0.79, with the highest forecast result in Yakutia (0.95), indicating the maximum propensity for record emissions due to weather conditions. In combination with forest management without fire protection, meteorological parameters have caused an increase in PM2.5 emissions in recent years in Siberia. The forest needs other ways to manage under the pressures of climate change to reduce environmental pollution associated with PM2.5 emissions from vast Siberian fires.