We applied logistic regression and Random Forest to evaluate drivers of fire occurrence on a provincial scale. Potential driving factors were divided into two groups according scale influence: ‘climate factors’, which operate regional scale, ‘local includes infrastructure, vegetation, topographic socioeconomic data. The analysed separately then significant from both together. Both models identified factors, ranked in terms relative importance. Results show that climate are the main forests...

Frequent and intense anthropogenic fires present meaningful challenges to forest management in the boreal of China. Understanding underlying drivers human-caused fire occurrence is crucial for making effective scientifically-based plans. In this study, we applied logistic regression (LR) Random Forests (RF) identify important biophysical factors that help explain likelihood Chinese forest. Results showed were more likely occur at areas close railways significantly influenced by types....

AimsThe pattern and driving factors of forest fires are interest for fire occurrence prediction management. The aims the study were: (i) to describe history human-caused by season size burned area over time; (ii) identify spatial patterns test existence 'hotspots' determine their exact locations in Daxing'an Mountains; (iii) that distribution possibility occurrence.

We applied a classic logistic regression (LR) model together with geographically weighted (GWLR) to determine the relationship between anthropogenic fire occurrence and potential driving factors in Chinese boreal forest test whether explanatory power of LR could be increased by considering geospatial information geographical human using GWLR model. Three tests, “all variables”, “significant “cross-validation”, were compare performance models. Our results confirmed importance distance...

Establishing an efficient PM2.5 prediction model and in-depth knowledge of the relationship between predictors in are great significance for preventing controlling pollution policy formulation Yangtze River Delta (YRD) where there is serious air pollution. In this study, spatial pattern concentration YRD during 2003–2019 was analyzed by Hot Spot Analysis. We employed five algorithms to train, verify, test 17 years data YRD, we explored drivers exposure. Our key results demonstrated: (1) High...

Forest city (FC) usually refers to an urban area with high forest coverage. It is a green model of development that has been strongly advocated for by governments many nations. fire prominent threat in FC development, but the causes fires FCs are different and more complex than pure forested areas since socio-economic factors human activity involved ignition spread fire. The large increasing number lives being exposed wildfire hazard highlights need understand characteristics these so...

Fires in urban-forest ecosystems (UFEs) are frequent with complex causes, posing a serious hazard to human lives and infrastructure. Thus, quantifying wildfire risks UFEs their spatial pattern is quintessential develop appropriate fire management strategies. The aim of this study was explore (geographically weighted logistic regression, GWLR) versus non-spatial (logistic LR) modelling approaches determine the relationship between forest occurrence driving factors Yichun, typical ecosystem...

Understanding the drivers of wildfire occurrence is great value for fire prevention and management, but due to variation in research methods, data sources, resolution those studies, it challenging conduct a large-scale comprehensive comparative qualitative analysis on topic. China has diverse vegetation types topography, undergone rapid economic social development, experiences high frequency wildfires, making one ideal locations research. We applied Random Forests modelling approach explore...

Wildfire is a major disturbance that affects large area globally every year. Thus, better prediction of the likelihood wildfire occurrence essential to develop appropriate fire prevention measures. We applied global negative Binomial (NB) and geographically weighted regression (GWNBR) models determine relationship between its drivers factors in boreal forests Great Xing’an Mountains, northeast China. Using geo-weighted techniques consider geospatial information meteorological, topographic,...

Climate determines the spatiotemporal distribution pattern of forest fires by affecting vegetation and extent drought. Thus, analyzing dynamic change fire season its response to climate will play an important role in targeted adjustments management policies practices. In this study, we studied variations occurrence Fujian Province, China using Mann–Kendall trend test correlation analysis analyze Moderate Resolution Imaging Spectroradiometer (MODIS) data from 2001 2016 meteorological data....

It is of great significance to understand the drivers PM2.5 and fire carbon emission (FCE) relationship between them for prevention, control, policy formulation severe exposure in areas where biomass burning a major source. In this study, we considered northern Laos as area research, utilized space cluster analysis present spatial pattern FCE from 2003–2019. With use random forest structural equation model, explored their drivers. The key results during target period study were follows: (1)...

Understanding the changes and driving factors of forest fire can provide scientific basis for prevention management fire. In this study, we analyzed in Zhejiang Province during 2001-2016 based on trend analysis Logistic regression model with MODIS satellite point data combined meteorological (daily ave-rage wind speed, daily average temperature, relative humidity, temperature difference, cumulative precipitation), human activities (distance from road, distance railway, resident, population,...

The effect of driving factors on forest fire occurrence at various risk levels beyond average is great interest to managers in practice. Using data collected Fujian province, China, global quantile regression (QR) and geographically weighted (GWQR) were applied investigate the spatially varying relationships between environmental different quantiles (e.g. 0.50, 0.75, 0.90 0.99) occurrence. These results indicated that: (1) each quantile, coefficients both QR GWQR negative for elevation,...

Four regression techniques, including two global models (i.e., Poisson and negative binominal) geographically weighted (GWR) (GWPR) binominal (GWNBR)) were used to explore which was the most suitable method for predicting number of forest fires investigate spatially varying relationships between environmental factors in Fujian province, Southeast China. Our results showed that GWR fitted fire count data better than models, yielded more realistic spatial distributions model predictions....

Forest fires have a significant impact on human life, property safety, and ecological environment. Deve-loping high-quality forest fire risk maps is beneficial for preventing fires, guiding resource allocation firefighting, assisting in suppression efforts, supporting decision-making. With multi-criteria decision analysis (MCDA) method based geographic information systems (GIS) literature review, we assessed the main factors influencing occurrences of Youxi County, Fujian Province. We...

Given the increasing importance of effectively identifying synergistic changes between PM2.5 and O3 comprehensively analyzing their impact on air quality management in China, we employ Sen+Mann–Kendall (Sen+M-K) trend test this study to examine temporal spatial variation trends Yangtze River Delta (YRD), from 2003 2020. We identified regions where these pollutants exhibited established pathways potential drivers, using geographically weighted random forest algorithms structural equation...