Search

Anyone who has tried lighting a campfire on a windy day can appreciate how difficult it could be. However, despite real-life experience and despite laboratory experiments which have demonstrated that fire ignition risk dramatically decreases beyond a certain wind threshold, current fire weather indices (FWIs) do not take this effect into account and assume a monotonic relation between wind velocity and ignition risk. In this paper, we perform a global analysis which empirically quantifies the probability of ignition as a function of wind velocity. Using both traditional methods (a logistic regression and a generalized additive model) and machine learning techniques, we find that beyond a threshold of approximately 3–4 m/s, the ignition risk substantially decreases. The effect holds when accounting for additional factors such as temperature and relative humidity. We recommend updating FWIs to account for this issue. [ABSTRACT FROM AUTHOR]

Wildfires are a major natural hazard that lead to deforestation, carbon emissions, and loss of human and animal lives every year. Effective predictions of wildfire occurrence and burned areas are essential to forest management and firefighting. In this paper we apply various machine learning (ML) methods on a 0.25° monthly resolution global dataset of wildfires. We test the prediction accuracies of four different fire occurrence classifiers: random forest (RF), eXtreme Gradient Boosting (XGBoost), multilayer perceptron (MLP) neural network, and a logistic regression. Our best ML model predicts wildfire occurrence with over 90% accuracy, compared to approximately 70% using a logistic regression. We then train ML regression models to predict the size of burned areas and obtain an MAE score of 3.13 km 2 , compared to 7.48 km 2 using a linear regression. To the best of our knowledge, this is the first study to be conducted in such resolution on a global dataset. We use the developed models to shed light on the influence of various factors on wildfire occurrence and burned areas. We suggest building upon these results to create ML-based fire weather indices. [ABSTRACT FROM AUTHOR]

Copyright of Canadian Journal of Forest Research is the property of Canadian Science Publishing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Forest fires are expected to be more frequent and more intense with climate change, including in temperate and mountain forest ecosystems. In the Alps, forest vulnerability to fire resulting from interactions between climate, fuel types, vegetation structure and tree resistance to fire is little understood. This paper aims at identifying trends in the vulnerability of Alpine forest ecosystems to fire at different scales (tree species, stand level and biogeographic level) and according to three different climatic conditions (cold season, average summer and extremely dry summer). To explore Alpine forest vulnerability to fire, we used surface fuel measurements, forest inventory and fire weather data to simulate fire behaviour and ultimately post-fire tree mortality across 4438 forest plots in the French Alps. The results showed that cold season fires (about 50% of the fires in the French Alps) have a limited impact except on low-elevation forests of the Southern Alps (mainly Oak, Scots pine). In average summer conditions, mixed and broadleaved forests of low elevations suffer the highest mortality rates (up to 75% in coppices). Finally, summer fires occurring in extremely dry conditions promote high mortality across all forest communities. Lowest mortality rates were observed in high forest stands composed of tree species presenting adaptation to surface fires (e.g. thick bark, high canopy) such as Larch forests of the internal Alps. This study provides insights on the vulnerability of the main tree species and forest ecosystems of the French Alps useful for the adaptation of forest management practices to climate changes. [ABSTRACT FROM AUTHOR]

The goal of decreasing wildfire hazard as much as possible, using minimal fuel treatments, has led to increasing scholarly interest in fuel reduction spatial optimisation. Most models in the field rest on the assumption of a known wind direction and a corresponding dominant direction of fire spread, and plan firebreaks in perpendicular directions. This strategy is effective when the wind blows in the hypothesised direction, but is quite ineffective when the wind direction is parallel to the firebreaks. In this article, we re-examine this assumption using a global fire dataset covering more than a decade. We perform a variety of circular statistical analyses including circular variance and principal component analysis (PCA). We find that the direction of fire spread in most regions is not limited to a single direction. We also find that the wind direction during fire weather is characterised by a high variance in a substantial fraction of regions around the globe. We validate this finding with a dataset comprised of over a hundred meteorological stations in Israel. We conclude that forest management should consider regional historical data of wind directions and fire spread directions, but also should plan firebreaks so that they are effective in various fire scenarios. [ABSTRACT FROM AUTHOR]

Copyright of Canadian Journal of Forest Research is the property of Canadian Science Publishing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Fire Ecology is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

• Dead fuel moisture content (DFMC) is a central determinant of wildfire risk. • DFMC values are shaped by the accumulated effect of meteorological factors. • We measured DFMC values in six sites in Israel over a four-year period. • Machine Learning applied on this novel dataset enable accurate DFMC prediction. • The trained models provide insights of the physical processes shaping DFMC values. Ten-hour dead fuel moisture content (DFMC10) is an important determinant of wildfire risk, as it provides a reasonable proxy to vegetation flammability. Thus far DFMC10 prediction by physical models has shown limited accordance with empirical measurements. DFMC10 can be measured using fuel-sticks, but to date these are only used in part of the meteorological stations. The goal of this paper is to improve the prediction accuracy of DFMC10 in the Mediterranean climate of Israel, where we collected four-year data from fuel sticks located in six meteorological stations. We show that Machine Learning (ML) models have the potential of accurately predicting DFMC10 based on historical values of temperature, relative humidity, rain, and additional meteorological factors. Our best ML model predicts low (<28%) DFMC10 values with an accuracy of 0.67% (MAE). The model even obtained a good MAE score (1.53%) when predicting data in stations it was never trained on. [ABSTRACT FROM AUTHOR]

China's forest cover has increased by approximately 10% as a result of sustainable forest management since the late 1970s. The forest ecosystem area affected by fire is increasing at an alarming rate of approximately 600,000 ha per year. The northeastern part of China, with a forest cover of 41.6%, has the greatest percentage of acres affected by forest fires. This study combines field and satellite weather data to determine factors that influence dead fuel moisture content (FMC). It assesses the use of the Canadian forest fire weather index to determine the daily forest fire danger in a typical temperate forest in Northeastern China during autumn. Based on the Wilcoxon test for paired samples, the observed and predicted values of FMC showed similar variation in eight of eleven sampling sites (72.7%), with a p value > 0.05. Three sampling plots presented lower predicted values of FMC than observed values (27.3%), with a p value < 0.05. The calculation of fire risk using the Canadian Forest Fire Weather Rating System (CFFDRS) in Maoer Mountain forest ecosystems presented low, medium or high risk; thus, the CFFDRS is suitable for determining fire danger in our study region. Along with these results, this study served to compare the use of FMC-metre field data and China Weather Station data to evaluate fire danger. The results of this study led us to suggest the multiplication of meteorological stations in fire-prone regions. [ABSTRACT FROM AUTHOR]

Fires burning under marginal fire-weather conditions tend to be patchy in terms of their spatial coverage. This patchiness is partially driven by variability in the ignitability of the fuel bed. An understanding of fuel-bed ignitability through space and time would help fire managers to more effectively carry out prescribed burns to achieve desired levels of burn coverage in Eucalyptus forests. We sought to identify the key fuel-bed attributes influencing ignitability under marginal weather conditions. We recorded ignition successes and failures at 45 points within 5 operational prescribed burns and used the data to build logistic regression models to predict the probability of ignition as a function of fuel-bed attributes. Models were ranked using an information theoretic approach. The four highest ranked models explained 48–54% of the variance in ignitability. Surface fine-fuel moisture content (FFMC) and overall fuel hazard (i.e. fuel arrangement) were the strongest predictors of ignitability, occurring in all four highest ranking models. Both surface FFMC and overall fuel hazard were negatively related to ignition likelihood, contradicting a commonly assumed positive relationship between fuel hazard and flammability. Our field method to measure ignition success could be applied across more prescribed burns to develop operationally useful models of ignitability. We examined fuel bed ignitibility in Eucalyptus forests. Field data from prescribed burns were used to model ignition probability as a function of fuel bed attributes. Surface fine-fuel moisture content and overall fuel hazard were the strongest predictors of the likelihood of ignition. [ABSTRACT FROM AUTHOR]

Prescribed burning is a widely accepted wildfire hazard reduction technique; however, knowledge of its effectiveness remains limited. To address this, we employ simulations of a widely used fire behaviour model across the ecologically diverse Australian island state of Tasmania. We simulate three broad scenarios: (1) no fuel treatment, (2) a maximal treatment, with the most possible prescribed burning within ecological constraints, and (3) 12 hypothetically more implementable state-wide prescribed-burning plans. In all simulations, we standardised fire-weather inputs to represent regionally typical dangerous fire-weather conditions. Statistical modelling showed that an unrealistically large maximal treatment scenario could reduce fire intensity in three flammable vegetation types, and reduce fire probability in almost every vegetation type. However, leverage analysis of the 12 more-realistic implementable plans indicated that such prescribed burning would have only a minimal effect, if any, on fire extent and that none of these prescribed-burning plans substantially reduced fire intensity. The study highlights that prescribed burning can theoretically mitigate wildfire, but that an unrealistically large area would need to be treated to affect fire behaviour across the island. Rather, optimisation of prescribed burning requires careful landscape design at the local scale. Such designs should be based on improved fire behaviour modelling, empirical measurement of fuels and analysis of actual wildfires. [ABSTRACT FROM AUTHOR]

Copyright of Canadian Journal of Forest Research is the property of Canadian Science Publishing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Identifying what factors control fire severity in large fires is critical for understanding fire impacts and planning pre- and post-fire management. Here, we determined the role of pre-fire stand structure, directional topography, and burning conditions on fire severity in a large fire (12,697 ha) in Central Spain that burned a Pinus pinaster forest on July 2005. Fire severity was estimated using RdNBR based on Landsat 5 TM images. Forest stand structure was reconstructed by systematically sampling the burned area ( n = 236). Burning conditions were established using weather information and a map of fire progression, based on which fire rate of spread and propagation direction were calculated. Topographic features in the direction of the fire-front were derived from a digital elevation model. Boosted regression tree (BRT) analysis was employed to relate each group of variables or the entire set to RdNBR. Fire severity was best explained by burning conditions (cross-validation correlation [CVC] 0.56), followed by pre-fire stand structure (CVC 0.34), and directional topography (CVC 0.17). Combining the three sets of variables, CVC increased to 0.71. Higher fire severity occurred in areas burning upslope, with high fire rate of spread, with heterogeneous and dense stands of P. pinaster and Quercus pyrenaica in the understory, receiving high solar radiation, among other characteristics. Fire severity was the result of interactive relationships between burning conditions, pre-fire stand structure, and directional topography. Thus, determining factors controlling fire severity from static stand structure or topography, as is often done, may not be appropriate. [ABSTRACT FROM AUTHOR]

The Canadian fire weather index system (CFFWIS) has been adopted for fire danger rating throughout Europe as part of the European Forest Fire Information System. However, its performance has not been thoroughly assessed, especially in environments less prone to fire. In this study, we characterised fire activity between 1995 and 2009 in the sub-Mediterranean Karst forest management area of SW Slovenia. Five fire danger classes (very low, low, moderate, high and very high) were derived from percentile analysis of the CFFWIS Fire Weather Index. These classes were found to be related with fire activity descriptors. Logistic regression prediction of fire-days based on CFFWIS indices had low accuracy, and better discrimination was achieved by classification tree modelling. Fire activity was found to be driven by current weather conditions rather than by drought. Our findings highlight the potential of fire danger rating to guide fire management but also the limitations imposed by relatively low incidence of fire and the spatial scale of analysis. [ABSTRACT FROM AUTHOR]

To provide insights concerning changes in fire regime in north-eastern Cambodia over the course of the Holocene, and discuss implications of these long-term data for fire management in the present day. Southern Ratanakiri Province, north-eastern Cambodia. The lake sites sampled here are embedded in a mosaic of mostly open, strongly deciduous dipterocarp forest, with patches of riparian, semi-evergreen and evergreen forests. Background information on the environmental and cultural setting comes from informal and semi-structured interviews of local villagers to determine present-day burning patterns and customs. Primary data come from analysis of changes in charcoal concentration within sediments from small, closed basin lakes. Charcoal data are compared with changes in pollen and sediment physical characteristics, and to present-day local customs, to infer or speculate on changes in human use of fire. Interviews with local people reveal two general types of human-induced fires, one type for swidden cultivation in denser forests, the other type for clearance of ground layer vegetation in more open forests. A 9300-year sediment record of microscopic charcoal deposition shows strongest fire activity ending by 8000 years ago, and the remainder of the early Holocene reflecting a strong summer monsoon and low fire activity. Beginning c. 5500 years ago, forest disturbance and fire activity increased. A subtle change in the record at c. 3500 years ago and more marked change at c. 2500 years ago suggest that fire frequency, and maybe human control over fire, became more important during that period and continuing up to the present. With this type of empirical data from only one site, it is impossible to make accurate conclusions about long-term human impacts from burning. However, this record does show that present-day charcoal input from fire activity is among the lowest for the last 9300 years. Considered together with other changes in the record and with present-day customs, there is a suggestion that anthropogenic fire is an adaptation to the monsoonal environment, and may be conservative of forest cover in open forest formations. This long-term perspective on the role of indigenous land-use customs in landscape evolution should be considered in forest management and biological conservation, as it differs significantly from the traditional rationale for policies of fire suppression in tropical forests. [ABSTRACT FROM AUTHOR]

Wildfires burn tens of thousands of hectares of forest, chaparral and grassland in Mediterranean countries every year, giving rise to landscape, ecologic, economic, and public safety concerns. On the Greek island of Chios and in many other Mediterranean landscapes, areas affected by fire are difficult to access and control due to rugged terrain, requiring wildfire preparedness and response plans that support fire fighting. This study utilized open source data and a weighted linear combination to extract factors that determine wildfire risk. Landsat satellite imagery and publicly available geospatial data were used to create a Geographic Information System and a multi-criteria analysis to develop a methodology for spatially modeling fire risk on Chios, a Greek island with frequent fire occurrence. This study focused on the static, structural component of the risk assessment to produce a spatial distribution of fire risk as a thematic map. Fire weather conditions were accounted for using Fuel Moisture Content, which reflected dryness of dead fuels and water deficit of live biomass. To assess the results, historic fire data representing actual occurrence of fire incidents were compared with probable fire locations predicted by our GIS model. It was found that there was a good agreement between the ground reference data and the results of the created fire risk model. The findings will help fire authorities identify areas of high risk for wildfire and plan the allocation of resources accordingly. This is because the outputs of the designed fire risk model are not complex or challenging to use in Chios, Greece and other landscapes. [ABSTRACT FROM AUTHOR]

Public lands provide many ecosystem services and support diverse plant and animal communities. In order to provide these benefits in the future, land managers and policy makers need information about future climate change and its potential effects. In particular, weather extremes are key drivers of wildfires, droughts, and false springs, which in turn can have large impacts on ecosystems. However, information on future changes in weather extremes on public lands is lacking. Our goal was to compare historical (1950–2005) and projected mid‐century (2041–2070) changes in weather extremes (fire weather, spring droughts, and false springs) on public lands. This case study looked at the lands managed by the U.S. Forest Service across the conterminous United States including 501 ranger district units. We analyzed downscaled projections of daily records from 19 Coupled Model Intercomparison Project 5 General Circulation Models for two climate scenarios, with either medium‐low or high CO2− equivalent concentration (RCPs 4.5 and 8.5). For each ranger district, we estimated: (1) fire potential, using the Keetch‐Byram Drought Index; (2) frequency of spring droughts, using the Standardized Precipitation Index; and (3) frequency of false springs, using the extended Spring Indices. We found that future climates could substantially alter weather conditions across Forest Service lands. Under the two climate scenarios, increases in wildfire potential, spring droughts, and false springs were projected in 32–72%, 28–29%, and 13–16% of all ranger districts, respectively. Moreover, a substantial number of ranger districts (17–30%), especially in the Southwestern, Pacific Southwest, and Rocky Mountain regions, were projected to see increases in more than one type of weather extreme, which may require special management attention. We suggest that future changes in weather extremes could threaten the ability of public lands to provide ecosystem services and ecological benefits to society. Overall, our results highlight the value of spatially‐explicit weather projections to assess future changes in key weather extremes for land managers and policy makers. [ABSTRACT FROM AUTHOR]