Wildfire risk in a changing climate: Evaluating fire weather indices and their global patterns with CMIP6 multi-model projections

Weather and Climate Extremes, Год журнала: 2025, Номер unknown, С. 100751 - 100751

Опубликована: Фев. 1, 2025

Язык: Английский

---

A framework for defining fire danger to support fire management operations in Australia†

International Journal of Wildland Fire, Год журнала: 2024, Номер 33(3)

Опубликована: Март 18, 2024

Background Development of the Australian Fire Danger Rating System began in 2017 with a project aimed at demonstrating feasibility new fire danger rating system through Research Prototype (AFDRSRP) that accounted for variability vegetation types, was nationally applicable, modular and open to continuous improvement. Aims In this manuscript, we identify define transition points categories AFDRSRP. We discuss user responses categorisation during live trial evaluation AFDRSRP reflect on limitations potential improvements. Methods A review available literature, broad consultation stakeholders reanalysis impact data were used determine suitable thresholds categorising within Key results transitions behaviour result application different management strategies or are associated variation serious consequences impacts. Conclusions The incorporated best science, supported by well-defined framework defining making it across jurisdictions range fuel types. Implications allows managers assess accuracy appropriateness forecasted danger.

Язык: Английский

---

Australian Fire Danger Rating System: implementing fire behaviour calculations to forecast fire danger in a research prototype†

International Journal of Wildland Fire, Год журнала: 2024, Номер 33(4)

Опубликована: Апрель 10, 2024

Background The Australian Fire Danger Rating System (AFDRS) was implemented operationally throughout Australia in September 2022, providing calculation of fire danger forecasts based on peer-reviewed behaviour models. system is modular and allows for ongoing incorporation new scientific research improved datasets. Aims Prior to operational implementation the AFDRS, a Research Prototype (AFDRSRP), described here, built test input data systems evaluate performance potential outputs. Methods spread models were selected aligned with fuel types process that captured bioregional variation characteristics. National spatial datasets created identify history alignment existing weather forecast layers. Key results AFDRSRP demonstrated improvements over McArthur Forest Grass due its use models, as well more accurately reflecting fuels. Conclusions design robust allowed updates prior AFDRS.

Язык: Английский

---

Live trial performance of the Australian Fire Danger Rating System – Research Prototype†

International Journal of Wildland Fire, Год журнала: 2024, Номер 33(4)

Опубликована: Март 28, 2024

Background The Australian Fire Danger Rating System program (AFDRS) has built a new fire danger rating system for Australia. A live trial of the system’s Research Prototype (AFDRSRP), based on behaviour thresholds, was run and evaluated between October 2017 March 2018. Aims Live results are critically analysed, knowledge gaps recommendations future work discussed. Methods bushfire experts assessed wildfires prescribed burns across range vegetation types weather conditions. Forecast ratings calculated using: (1) AFDRSRP; (2) Forest Index (FFDI) Grassland (GFDI) were compared against derived by expert opinion each evaluation (n = 336). Key Overall performance AFDRSRP superior to FFDI/GFDI (56 vs 43% correct), with tendency over-predict rather than under-predict potential. also demonstrated its value assess in fuel not conforming current grassland or forest models; e.g. fuels that grouped use mallee-heath, spinifex shrubland spread models. Conclusions successful, outperforming existing operational system. Implications Identified improvements would further enhance performance, ensuring readiness implementation.

Язык: Английский

---

Interpretation of seasonal fire outlooks

Journal of Southern Hemisphere Earth System Science, Год журнала: 2025, Номер 75(1)

Опубликована: Март 18, 2025

The Australian Fire Danger Rating System (AFDRS) is a nationally consistent approach to forecasting fire danger for all major vegetation types found in Australia. AFDRS climate outlooks (Fire Outlooks, FDOs) extending out 3 months ahead are the first such operational products of their kind world. use Bureau’s seasonal model Community Climate Earth simulator – Seasonal (ACCESS-S2). FDOs currently available agencies, and partner agencies involved land management prevention activities. To make sound planning decisions, should be used with other sources intelligence understand which components might driving risk. It prudent consult temperature rainfall as both these contributing factors conditions, but have differing data foundations (hindcast periods) that need understood correct interpretation. Previous comparative analysis showed hindcast period warmer during shoulder seasons some regions; thus, high chance above average not reflected expected outlooks. For this reason, it has been important provide users advice on how best interpret alongside In work, we continued determine differs over periods subsequent implications when interpreting strategic context.

Язык: Английский

---

Exploration of geo-spatial data and machine learning algorithms for robust wildfire occurrence prediction

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Март 28, 2025

Wildfires play a pivotal role in environmental processes and the sustainable development of ecosystems. Timely responses can significantly reduce damages consequences caused by their spread. Several critical issues wildfire behavior analysis include fire occurrence forecasting, early detection, spread prediction. In this study, we focus on which is valuable tool for facilitating earlier intervention. Conventional approaches primarily rely computation indices based weather conditions. However, solutions that utilize more comprehensive data, remote sensing information, artificial intelligence (AI) algorithms may offer substantial advantages rapid decision-making extensive territory monitoring. The wide variety spatial parameters great diversity geographical regions influence complicate task. Consequently, there no unified approach predicting occurrences using data AI techniques. goal study to explore potential various available - meteorological, geo-spatial, anthropogenic machine learning (ML) algorithms. We developed pipeline acquisition subsequent ML-based algorithm development. includes following algorithms: Random Forest, XGBoost, Autoencoder, ConvLSTM, Attention Multilayer Perceptron, RegNetX. addition, several metrics assess quality models case highly imbalanced spatio-temporal data. To conduct collected unique dataset covering large central Russia, incorporating than 17,000 verified events over period 10 years. findings underscore necessity developing individual ML tailored each region, taking into account specific features correlated with probability occurrence. achieved models, as measured F1-score, varies from 0.7 0.87 depending demonstrating integrating such emergency response systems.

Язык: Английский

---

Meteorological analysis of an extreme pyroconvective wildfire at Dunalley-Forcett, Australia

Natural Hazards, Год журнала: 2025, Номер unknown

Опубликована: Март 27, 2025

Язык: Английский

---

An efficient and comprehensive field protocol for assessing fuel characteristics for fire behaviour modelling in Australian open forests

MethodsX, Год журнала: 2025, Номер 14, С. 103345 - 103345

Опубликована: Май 1, 2025

Knowledge of fuel characteristics and their spatial temporal distribution is increasingly important as fire managers rely on this information to quantify risk, plan prescribed burning activities, forecast danger predict wildland behaviour effects. Current inventory approaches used in Australia largely visual assessment methods that are subjective lack the consistency accuracy required for management applications. We describe a protocol various strata considered Australian modelling applications, namely: litter suspended dead fuels; downed wood debris; live understorey; bark; overstorey canopy. The method provides about:•Cover height (or depth) each strata;•Mass fine fuels litter, understorey layers (dead diameter (d) ≤ 0.6 cm, d 0.4 cm); and•Mass size class woody (d>0.6 cm). integrates variety sampling including destructive particles, line intersect fuel, indirect relying double techniques estimate understorey, bark canopy fuels. can be adapted enable application situations with distinct requirements. Data collected using will have direct use developing models forest dynamics evaluating outputs from remote sensing these

Язык: Английский

---

Data-driven recommendations for enhancing real-time natural hazard warnings

One Earth, Год журнала: 2025, Номер 8(5), С. 101274 - 101274

Опубликована: Май 1, 2025

Язык: Английский

---

Predicting Forest Fire Area Growth Rate Using an Ensemble Algorithm

Forests, Год журнала: 2024, Номер 15(9), С. 1493 - 1493

Опубликована: Авг. 26, 2024

Due to its unique geographical and climatic conditions, the Liangshan Prefecture region is highly prone large fires. There an urgent need study growth rate of fire-burned areas fill research gap in this region. To address issue, uses Grey Wolf Optimizer (GWO) algorithm optimize hyperparameters eXtreme Gradient Boosting (XGBoost) model, constructing a GWO-XGBoost model. Finally, optimized ensemble model (GWO-XGBoost) used create fire warning map for Sichuan Province, China, filling forest studies area. This comprehensively selects factors such as monthly climate, vegetation, terrain, socio–economic aspects incorporates reanalysis data from assessment systems Canada, United States, Australia features construct dataset. After collinearity tests filter redundant Pearson correlation analysis explore related burned area rate, Synthetic Minority Oversampling Technique (SMOTE) oversample positive class samples. The GWO XGBoost which then compared with XGBoost, Random Forest (RF), Logistic Regression (LR) models. Model evaluation results showed that AUC value 0.8927, best-performing Using SHapley Additive exPlanations (SHAP) method quantify contribution each influencing factor indicates Ignition Component (IC) States National Fire Danger Rating System contributes most, followed by average temperature population density. indicate southern part key prevention

Язык: Английский

---