Impacts on and damage to European forests from the 2018–2022 heat and drought events

Natural hazards and earth system sciences, Journal Year: 2025, Volume and Issue: 25(1), P. 77 - 117

Published: Jan. 6, 2025

Abstract. Drought and heat events in Europe are becoming increasingly frequent due to human-induced climate change, impacting both human well-being ecosystem functioning. The intensity effects of these vary across the continent, making it crucial for decision-makers understand spatial variability drought impacts. Data on drought-related damage currently dispersed scientific publications, government reports, media outlets. This study consolidates data European forests from 2018 2022, using Europe-wide datasets including those related crown defoliation, insect damage, burnt forest areas, tree cover loss. data, covering 16 countries, were analysed four regions, northern, central, Alpine, southern, compared with a reference period 2010 2014. Findings reveal that all zones experienced reduced vitality elevated temperatures, varying severity. Central showed highest vulnerability, coniferous deciduous trees. southern zone, while affected by loss, demonstrated greater resilience, likely historical exposure. northern zone is experiencing emerging impacts less severely, possibly site-adapted boreal species, Alpine minimal impact, suggesting protective effect altitude. Key trends include (1) significant loss zones; (2) high levels despite 2021 being an average year, indicating lasting previous years; (3) notable challenges central Sweden bark beetle infestations; (4) no increase wildfire severity ongoing challenges. Based this assessment, we conclude (i) highly vulnerable heat, even resilient ecosystems at risk severe damage; (ii) tailored strategies essential mitigate change forests, incorporating regional differences resilience; (iii) effective management requires harmonised collection enhanced monitoring address future comprehensively.

Language: Английский

---

Mapping forest fire susceptibility using spatially explicit ensemble models based on the locally weighted learning algorithm

Ecological Informatics, Journal Year: 2021, Volume and Issue: 63, P. 101292 - 101292

Published: April 8, 2021

Language: Английский

---

A Systematic Review of Applications of Machine Learning Techniques for Wildfire Management Decision Support

Inventions, Journal Year: 2022, Volume and Issue: 7(1), P. 15 - 15

Published: Jan. 21, 2022

Wildfires threaten and kill people, destroy urban rural property, degrade air quality, ravage forest ecosystems, contribute to global warming. Wildfire management decision support models are thus important for avoiding or mitigating the effects of these events. In this context, paper aims at providing a review recent applications machine learning methods wildfire support. The emphasis is on summary with classification according case study type, method, location, performance metrics. considers documents published in last four years, using sample 135 (review articles research articles). It concluded that adoption may enhancing different fire phases.

Language: Английский

---

Natural disturbances risks in European Boreal and Temperate forests and their links to climate change – A review of modelling approaches

Forest Ecology and Management, Journal Year: 2022, Volume and Issue: 509, P. 120071 - 120071

Published: Feb. 10, 2022

It is expected that European Boreal and Temperate forests will be greatly affected by climate change, causing natural disturbances to increase in frequency severity. To detangle how, through forest management, we can make less vulnerable the impact of disturbances, need include risks such our decision-making tools. The present review investigates: i) how most important forestry-related are linked ii) different modelling approaches assess their applicability for large-scale management planning. Global warming decrease frozen soil periods, which increases root rot, snow, ice wind damage, cascading into an increment bark beetle damage. Central Europe experience a precipitation temperature, lowers tree defenses against beetles rot infestations. Ice wet snow damages Northern forests, reduce Southern forests. However, lack cover may cases frost-damaged seedlings. increased temperatures drought together with fuel from other likely enhance wildfire risk, especially For approaches, thirty-nine disturbance models were assessed categorized according required input variables models' outputs. Probability usually common all model however, predict effects seem scarce.

Language: Английский

---

A multi-hazard map-based flooding, gully erosion, forest fires, and earthquakes in Iran

Scientific Reports, Journal Year: 2021, Volume and Issue: 11(1)

Published: July 21, 2021

Abstract We used three state-of-the-art machine learning techniques (boosted regression tree, random forest, and support vector machine) to produce a multi-hazard (MHR) map illustrating areas susceptible flooding, gully erosion, forest fires, earthquakes in Kohgiluyeh Boyer-Ahmad Province, Iran. The earthquake hazard was derived from probabilistic seismic analysis. mean decrease Gini (MDG) method implemented determine the relative importance of effective factors on spatial occurrence each four hazards. Area under curve (AUC) plots, based validation dataset, were created for maps generated using algorithms compare results. model had highest predictive accuracy, with AUC values 0.994, 0.982, 0.885 respectively. Approximately 41%, 40%, 28%, 3% study area are at risk earthquakes, floods,

Language: Английский

---

Forest fire susceptibility assessment using google earth engine in Gangwon-do, Republic of Korea

Geomatics Natural Hazards and Risk, Journal Year: 2022, Volume and Issue: 13(1), P. 432 - 450

Published: Feb. 2, 2022

Forest fires are one of the most frequently occurring natural hazards, causing substantial economic loss and destruction forest cover. As Gangwon-do region in Korea has abundant resources ecological diversity as Korea's largest area, spatial data on fire susceptibility urgently required. In this study, a map (FFSM) was constructed using Google Earth Engine (GEE) three machine learning algorithms: Classification Regression Trees (CART), Random (RF), Boosted (BRT). The factors related to climate, topography, hydrology, human activity were constructed. To verify accuracy, area under receiver operating characteristic curve (AUC) used. AUC values 0.846 (BRT), 0.835 0.751 (CART). Factor importance analysis performed identify important occurrence Gangwon-do. results show that factor is slope. A slope approximately 17° (moderately steep) considerable impact fires. Human interference other affect established FFSM can support future efforts resource protection environmental management planning

Language: Английский

---

Flood, landslides, forest fire, and earthquake susceptibility maps using machine learning techniques and their combination

Natural Hazards, Journal Year: 2023, Volume and Issue: 116(3), P. 3797 - 3816

Published: Feb. 9, 2023

Language: Английский

---

Mapping Forest Fire Risk Zones Using Machine Learning Algorithms in Hunan Province, China

Sustainability, Journal Year: 2023, Volume and Issue: 15(7), P. 6292 - 6292

Published: April 6, 2023

Forest fire is a primary disaster that destroys forest resources and the ecological environment, has serious negative impact on safety of human life property. Predicting probability fires drawing risk maps can provide reference basis for control management in Hunan Province. This study selected 19 factors based satellite monitoring hotspot data, meteorological topographic vegetation social data from 2010–2018. It used random forest, support vector machine, gradient boosting decision tree models to predict Province RF algorithm create map quantify potential risk. The results show performs best compared SVM GBDT algorithms with 91.68% accuracy, 91.96% precision, 92.78% recall, 92.37% F1, 97.2% AUC. most important drivers are meteorology vegetation. There obvious differences spatial distribution seasonal risks Province, winter spring seasons high risks. medium- high-risk areas mostly concentrated south Hunan.

Language: Английский

---

Forest fire risk mapping with Landsat 8 OLI images: Evaluation of the potential use of vegetation indices

Ecological Informatics, Journal Year: 2024, Volume and Issue: 79, P. 102461 - 102461

Published: Jan. 7, 2024

Fire is one of the most important natural catastrophes threatening forest ecosystem. The severity and frequency fires are increasing daily due to increase in population vulnerable areas effects global climate change. Creating fire risk maps using them take required protective actions prevent will decrease adverse fires. This study focused on producing comparing based four vegetation indices, Normalized Burn Ratio (NBR) index, Thermal (NBRT) Difference Vegetation Index (NDVI), Water (NDWI) data gathered with use remote sensing devices. Muğla Regional Directorate Forestry, which Mediterranean zone has experienced mega-fires, was selected as case area. were prepared for indices from Landsat 8 OLI satellite images. Receiver operating characteristic curves 195 ignition points that occurred 2021 July 5 end year used assess accuracy maps. Most locations (>90%) high- extremely high-risk according NBR, NDWI, NDVI. fact almost all revealed area sensitive draw up highly accurate predicting where might occur. results showed under curve 0.842 0.835 0.812 NBRT, 0.810 NBR approach more precise than other models providing information Risk created could help decision-makers precautions minimize damage.

Language: Английский

---

Modelling current and future forest fire susceptibility in north-eastern Germany

Natural hazards and earth system sciences, Journal Year: 2025, Volume and Issue: 25(1), P. 383 - 401

Published: Jan. 27, 2025

Abstract. Preventing and fighting forest fires has been a challenge worldwide in recent decades. Forest alter structure composition; threaten people's livelihoods; lead to economic losses, as well soil erosion desertification. Climate change related drought events, paired with anthropogenic activities, have magnified the intensity frequency of fires. Consequently, we analysed fire susceptibility (FFS), which can be understood likelihood occurrence certain area. We applied random (RF) machine learning (ML) algorithm model current future FFS federal state Brandenburg (Germany) using topographic, climatic, anthropogenic, soil, vegetation predictors. was modelled at spatial resolution 50 m for (2014–2022) scenarios (2081–2100). Model accuracy ranged between 69 % (RFtest) 71 (leave one year out, LOYO), showing moderately high reliability predicting FFS. The results underscore importance parameters modelling on regional level. This study will allow managers environmental planners identify areas are most susceptible fires, enhancing warning systems prevention measures.

Language: Английский

---