Fire Weather Index System Research Articles

Enhanced prediction of extreme fire weather conditions in spring using the Hot-Dry-Windy Index in Alberta, Canada

Background Fire weather indices forecast fire behaviour and provide valuable information for wildland fire prevention, preparedness, and suppression. However, these indices do not directly account for atmospheric conditions aloft. The province of Alberta, Canada has experienced extreme fire weather conditions during spring for decades, leading to the continued occurrence of disastrous wildland fires. Aims We examined the Hot-Dry-Windy Index (HDWI) and spread days over the first 4 days of 80 large wildland fires that started in May 1990–2019 in Alberta. Methods HDWI values were calculated using ERA5 reanalysis data from the 1000, 975 and 950 hPa levels. Differences between HDWI distributions on spread days and non-spread days were examined using permutation tests. Initial Spread Index was also examined as it is considered an important Fire Weather Index System value for wildland fire spread during spring in Alberta. Key results Higher median HDWI values were observed on spread days than non-spread days, where median Initial Spread Index values showed little to no difference. Conclusions This analysis suggests that HDWI can contribute to the prediction of significant spring wildland fire spread in Alberta. Implications Forecasted HDWI and HDWI climatologies may provide additional decision support for wildland fire management agencies.

The Role of Field Measurements of Fine Dead Fuel Moisture Content in the Canadian Fire Weather Index System—A Study Case in the Central Region of Portugal

The Canadian Fire Weather Index System (CFWIS), empirically developed for forests in Canada, estimates the fuel moisture content (mf) at different depths and loads through meteorological parameters. While it is often suggested that adapting an existing fire danger rating system like CFWIS for a new environment requires developing new relationships or modifying existing ones, it is worth considering if such adaptations are always necessary. Based on a dataset of field measurements for surface litter (Pinus pinaster) carried out in the central region of Portugal (2014–2023), we propose a correction of mf based on the Fine Fuel Moisture Code (FFMC) of the CFWIS. This moisture correction was used to determine the Initial Spread Index (ISI) directly and, subsequently, the Fire Weather Index (FWI). Fire records from the study region were used to analyze the performance of the corrected indices. We found that the moisture correction led to higher values and potentially more accurate indices under dry conditions but did not provide a significant improvement in predicting the number of fires and burned areas compared to the original indices. The results suggest that, in relation to fire activity, the CFWIS is sufficiently robust to variations in the fuel moisture content in the study region.

Forecasting Hourly Wildfire Risk: Enhancing Fire Danger Assessment Using Numerical Weather Prediction

Abstract Wildfire agencies use fire danger rating systems (FDRSs) to deploy resources and issue public safety measures. The most widely used FDRS is the Canadian fire weather index (FWI) system, which uses weather inputs to estimate the potential for wildfires to start and spread. Current FWI forecasts provide a daily numerical value, representing potential fire severity at an assumed midafternoon time for peak fire activity. This assumption, based on typical diurnal weather patterns, is not always valid. To address this, we developed an hourly FWI (HFWI) system using numerical weather prediction. We validate HFWI against the traditional daily FWI (DFWI) by comparing HFWI forecasts with observation-derived DFWI values from 917 surface fire weather stations in western North America. Results indicate strong correlations between forecasted HFWI and the observation-derived DFWI. A positive mean bias in the daily maximum values of HFWI compared to the traditional DFWI suggests that HFWI can better capture severe fire weather variations regardless of when they occur. We confirm this by comparing HFWI with hourly fire radiative power (FRP) satellite observations for nine wildfire case studies in Canada and the United States. We demonstrate HFWI’s ability to forecast shifts in fire danger timing, especially during intensified fire activity in the late evening and early morning hours, while allowing for multiple periods of increased fire danger per day—a contrast to the conventional DFWI. This research highlights the HFWI system’s value in improving fire danger assessments and predictions, hopefully enhancing wildfire management, especially during atypical fire behavior.

Factors affecting severity of wildfires in Scottish heathlands and blanket bogs

Temperate heathlands and blanket bogs are globally rare and face growing wildfire threats. Ecosystem impacts differ between low and high severity fires, where severity reflects immediate fuel consumption. This study assessed factors influencing fire severity in Scottish heathlands and blanket bogs, including the efficacy of the Canadian Fire Weather Index System (CFWIS). Using remote sensing, we measured the differenced Normalised Burn Ratio at 92 wildfire sites from 2015 to 2021. We used Generalised Additive Mixed Models to investigate the impact of topography, habitat wetness, CFWIS components and 30-day weather on severity. Dry heath exhibited higher severity than wet heath and blanket bog, and slope, elevation and south facing aspect were positively correlated to severity. Weather effects were less clear due to data scale differences, yet still indicated weather's significant role in severity. Rainfall had an increasingly negative effect from approximately 15 days before the fire, whilst temperature had an increasingly positive effect. Vapour Pressure Deficit (VPD) was the weather variable with highest explanatory value, and predicted severity better than any CFWIS component. The best-explained fire severity model (R2 = 0.25) incorporated topography, habitat wetness wind and VPD on the day of the fire. The Drought Code (DC), predicting organic matter flammability at ≥10 cm soil depth, was the CFWIS component with the highest predictive effect across habitats. Our findings suggest that wildfires in wet heath and blanket bogs are typically characterised by low severity, but that warmer, drier weather may increase the risk of severe, smouldering fires which threaten peatland carbon stores.

Synoptic-scale drivers of fire weather in Greece

The identification of the large-scale atmospheric circulation patterns which are associated with extreme fire weather is of great importance for developing early warning systems, management strategies, and for increasing awareness and preparedness of all the involved entities, including both the public and practitioners. Such a forecasting approach is currently missing in Greece and many other countries. Furthermore, considering climate projections over the Mediterranean, which indicate an environment more conducive to wildfire activity, the need for timely forecasting of extreme fire weather becomes increasingly urgent. Here, we present an alternative fire weather forecasting framework using ERA5 reanalysis data of atmospheric variables and fire weather indices of the Canadian Forest Fire Weather Index System (CFFWIS) during the period June–October from 1979 to 2019. Within this framework, we define the critical fire weather patterns (CFWPs) of Greece associated with different levels of fire weather severity by applying Self-Organizing-Maps (SOMs) on mid-tropospheric geopotential height. We quantify the fire weather conditions associated with each CFWP. Using a set of CFFWIS indices and key fire weather variables, our SOM-based analysis reveals five distinct CFWPs linked to different levels and characteristics of fire weather severity. The lowest fire weather severity is associated with lower than average geopotential heights, and anomalous cold and moist weather. The highest fire weather severity is associated with higher than average geopotential heights, and anomalous hot, dry, and windy conditions, suggesting the potential for wind-driven wildfires. Our analysis yields elevated fire weather severity linked to a CFWP, when hot and dry conditions are accompanied by atmospheric instability, suggesting the potential for plume-driven wildfires and the potential for pyroconvection. The main advantage of this forecasting framework is that it could be used for providing valuable information regarding the upcoming fire weather conditions even up to 7–12 days in advance depending on the atmospheric predictability.

Improving the fire weather index system for peatlands using peat-specific hydrological input data

Abstract. The Canadian Fire Weather Index (FWI) system, even though originally developed and calibrated for an upland Jack pine forest, is used globally to estimate fire danger for any fire environment. However, for some environments, such as peatlands, the applicability of the FWI in its current form, is often questioned. In this study, we replaced the original moisture codes of the FWI with hydrological estimates resulting from the assimilation of satellite-based L-band passive microwave observations into a peatland-specific land surface model. In a conservative approach that maintains the integrity of the original FWI structure, the distributions of the hydrological estimates were first matched to those of the corresponding original moisture codes before replacement. The resulting adapted FWI, hereafter called FWIpeat, was evaluated using satellite-based information on fire presence over boreal peatlands from 2010 through 2018. Adapting the FWI with model- and satellite-based hydrological information was found to be beneficial in estimating fire danger, especially when replacing the deeper moisture codes of the FWI. For late-season fires, further adaptations of the fine fuel moisture code show even more improvement due to the fact that late-season fires are more hydrologically driven. The proposed FWIpeat should enable improved monitoring of fire risk in boreal peatlands.

Assessment of wildfires forecast performance in Albania: Case study

The harmful impacts of climate change caused by wildfires are substantially harming the people of mainland Europe, as well as damaging species biodiversity and the ecosystem. It can be minimised by improving the effectiveness of fire risk forecasting and mitigation strategies. The aim of this paper was to investigate the accuracy of forest fire forecasts in Albania produced by the FWI (Fire Weather Index) system. During the summer of 2022, observations and data were collected on expected and actual fires in the prefectures of Albania, which were previously divided into four categories according to the level of fire risk: high, moderate, low, and zero. It was determined that, in the summer of 2022, Albania happened a grand total of 620 wildfires. The data were analysed using two indicators: the probability of fire occurrence for a particular prefecture and the number of fires per prefecture. The analysis revealed varying degrees of accuracy in fire predictions across different prefectures, with higher precision observed in high-risk regions but diminishing as the risk level decreased. The most dependable indicator of forecast accuracy, reaching 75%, was observed in high-risk areas during the month of August. Predicting fire localization within moderate-risk zones consistently achieved results above 50% but fell short of the 60% threshold. Overall, the results confirm the effectiveness of using data to predict the probability of fires for prefectures with a high and increased level of the relevant threat. This will make it possible to more effectively deploy and mobilise the resources needed to overcome them and substantially reduce the losses associated with them

Satellite observed response of fire dynamics to vegetation water content and weather conditions in Southeast Asia

Fires greatly affect terrestrial ecosystems, climate, human health, and economy. Under the threat of global warming, assessing fire occurrence frequency and fire danger is critical to mitigate fire’s negative impacts. The state-of-the-art satellite microwave observation provides the unique opportunity to monitor vegetation water content (VWC) for multiple vegetation strata, yet still lacks associated investigation in fire dynamics. Here, we applied the multiple microwave Emissivity Difference Vegetation Indexes (EDVIs) as VWC proxies of different vegetation strata to explore their influence on fire occurrence frequency (given by the number of fires, NOF) and the cumulative fire radiant energy (FRE) in mainland Southeast Asia during 2015–2019; and compared to that with the commonly used fire weather indices from the Canadian Forest Fire Weather Index system. Then, we identified the optimal regression models to depict the relationship of fires with EDVIs and fire weather indices, respectively; and incorporated both kinds of indices in models to retrospectively predict NOF and FRE. Results show that: (1) the monthly mean NOF and FRE nonlinearly decrease with EDVI in mainland Southeast Asia, and 40–60% of the variances are explained by EDVI, which further improve to 60–70% by incorporating EDVIs of two different vegetation layers. (2) In comparison, NOF and FRE nonlinearly increase with fire weather indices, and connections are stronger than that with EDVI; nevertheless, combing fire weather indices with EDVI, especially with EDVIs of two vegetation layers provides almost the highest explanation for fires. (3) The log-linear and/or log-quadratic models are identified to outperform the other models in depicting the fire-EDVI and fire-fire weather indices relationships; and incorporating EDVI and fire weather indices in models generally improves the prediction accuracy. These findings reveal the differentiated VWC effects on NOF and FRE among vegetation strata and quantify the strong nonlinear response of fires to VWC and weather conditions in mainland Southeast Asia, which could provide new insight for fire prediction and fire danger assessments with the utility of microwave observations.

Evaluation of CMIP6 model performances in simulating fire weather spatiotemporal variability on global and regional scales

Abstract. Weather and climate play an important role in shaping global wildfire regimes and geographical distributions of burnable area. As projected by the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC-AR6), in the near future, fire danger is likely to increase in many regions due to warmer temperatures and drier conditions. General circulation models (GCMs) are an important resource in understanding how fire danger will evolve in a changing climate, but, to date, the development of fire risk scenarios has not fully accounted for systematic GCM errors and biases. This study presents a comprehensive global evaluation of the spatiotemporal representation of fire weather indicators from the Canadian Forest Fire Weather Index System simulated by 16 GCMs from the sixth Coupled Model Intercomparison Project (CMIP6). While at the global scale, the ensemble mean is able to represent variability, magnitude and spatial extent of different fire weather indicators reasonably well when compared to the latest global fire reanalysis, there is considerable regional and seasonal dependence in the performance of each GCM. To support the GCM selection and application for impact studies, the evaluation results are combined to generate global and regional rankings of individual GCM performance. The findings highlight the value of GCM evaluation and selection in developing more reliable projections of future climate-driven fire danger, thereby enabling decision makers and forest managers to take targeted action and respond to future fire events.

Evaluating the Drought Code for lowland taiga of Interior Alaska using eddy covariance measurements

Background The Drought Code (DC) of the Canadian Fire Weather Index System (CFWIS) has been intuitively regarded by fire managers in Alaska, USA, as poorly representing the moisture content in the forest floor in lowland taiga forests on permafrost soils. Aims The aim of this study was to evaluate the DC using its own framework of water balance as cumulative additions of daily precipitation and substractions of actual evaporation. Methods We used eddy covariance measurements (EC) from three flux towers in Interior Alaska as a benchmark of natural evaporation. Key results The DC water balance model overpredicted drought for all 14 site-years that we analysed. Errors in water balance cumulated to 109 mm by the end of the season, which was 54% of the soil water storage capacity of the DC model. Median daily water balance was 6.3 times lower than that measured by EC. Conclusions About half the error in the model was due to correction of precipitation for canopy effects. The other half was due to dependence of the actual evaporation rate on the proportional ‘fullness’ of soil water storage in the DC model. Implications Fire danger situational awareness is improved by ignoring the DC in the CFWIS for boreal forests occurring on permafrost.

Assessment of lake-level fluctuation as an indicator of fire activity in boreal Canada

Fuel moisture content is a highly variable property that dictates the ignition and combustion potentials of wildfire fuels. The Canadian Fire Weather Index (FWI) System—a subset of the established Canadian Forest Fire Danger Rating System—relies solely on weather observations to track fuel moisture and compute indices of fire potential. Ancillary environmental measures, such as lake levels in non-regulated lake systems, may integrate climatic and landscape factors that better reflect environmental moisture deficits and thereby complement the FWI System. Accordingly, we investigated whether inter-annual variation in lake levels has potential to serve as an indicator of regional fire activity using a quasi-binomial regression approach. This study combines annual measurements of fire activity, lake level, and other climatic variables from the Boreal Shield and Boreal Plain ecozones of Canada from 1980 to 2019 to determine whether variation in regional area burned can be explained by lake-level fluctuation, and whether this explanatory power differs by ecozone. Results indicated that lake-level fluctuation is a significant predictor of regional area burned for many lakes (28–48%). It often performs as well as or better than traditional fire-weather metrics (such as temperature and precipitation), and this relationship is stronger on the Boreal Shield compared to the Boreal Plain. Lake-level fluctuation provides information on fuel availability unaccounted for by traditional fire-weather metrics, suggesting that this variable could be added to the suite of environmental indicators of regional fire activity in Canada.

Summer and Fall Extreme Fire Weather Projected to Occur More Often and Affect a Growing Portion of California throughout the 21st Century

Annual burned area has increased in California over the past three decades as a result of rising temperatures and a greater atmospheric demand for moisture, a trend that is projected to continue throughout the 21st century as a result of climate change. Here, we implement a bias-correction and statistical downscaling technique to obtain high resolution, daily meteorological conditions for input into two fire weather indices: vapor pressure deficit (VPD) and the Canadian Fire Weather Index System (FWI). We focus our analysis on 10 ecoregions that together account for the diverse range of climates, ecosystems, topographies, and vegetation types found across the state of California. Our results provide evidence that fire weather conditions will become more extreme and extend into the spring and fall seasons in most areas of California by 2100, extending the amount of time vegetation is exposed to increased atmospheric demand for moisture, and heightening the overall risk for the ignition and spread of large wildfire. The ecoregion-level spatial scale adopted for this study increases the spatial specificity of fire weather information, as well as the resolution with which fire and land managers can implement strategies and counter-measures when addressing issues related to climate change.

The Performance of ECMWF Ensemble Prediction System for European Extreme Fires: Portugal/Monchique in 2018

At the beginning of August 2018, Portugal experienced a severe heat episode over a few days that consequently increased the probability of wildfire events. Due to the advection of an anomalous very hot and dry air mass, severe fire-prone meteorological conditions were forecasted mainly over southern Portugal, in the Monchique region. Together with the significant fuel amount accumulated since the last extreme wildfire in August 2003, all the unfavorable conditions were set to drive a severe fire over this region. The Monchique fire started on 3 August 2018, being very hard to suppress and lasting for seven days, with a burnt area of 27,000 ha. Regarding the need to have operational early warning tools, this work aims to evaluate the reliability of fire probabilistic products, up to 72 h ahead, together with the use of fire radiative power products, as support tools in fire monitoring and resource activities. To accomplish this goal, we used the fire probabilistic products of the Ensemble Prediction System, provided by the Copernicus Atmosphere Monitoring Service. Among available fire danger rating systems, the Fire Weather Index and the Fine Fuels Moisture Code of the Canadian Forest Fire Weather Index System were selected to assess the meteorological fire danger. The assessment of the fire intensity was based on the Fire Radiative Energy released, considering the Fire Radiative Power, delivered in near real-time, by EUMETSAT Land Surface Analysis Satellite Applications Facility. The exceptional fire danger over southern Portugal that favors the ignition of the Monchique fire and its severity was essential driven by two important factors: (i) the anomalous fire weather danger, before and during the event; (ii) the accumulated fuel amount, since the last severe event occurred in 2003, over the region. Results show that the selected fire probabilistic products described the meteorological fire danger observed well, and the LSA-SAF products revealed the huge amount of fire energy emitted, in line with the difficulties faced by authorities to suppress the Monchique fire.

Optimal Cross‐Validation Strategies for Selection of Spatial Interpolation Models for the Canadian Forest Fire Weather Index System

AbstractThe Canadian forest fire weather index (FWI) system requires spatially continuous, gridded weather data for temperature, relative humidity, wind speed, and precipitation. Reliable estimates of the Canadian FWI system components are needed to ensure the safety of communities, resources, and ecosystems. The quality of the interpolated input weather variables are typically evaluated using error estimates from cross‐validation. These error estimates are used for selecting between spatial interpolation methods for generating the continuous weather surfaces. Leave‐one‐out cross‐validation (LOOCV) is the most commonly used method, but it is biased in spatially clustered weather station networks. Accurate error estimation is important for selecting the optimal interpolation method and evaluating how well an interpolated surface represents true patterns in a weather variable. Other cross‐validation methods may better account for bias relating to clustered weather station networks. We present a comparison of cross‐validation methods for evaluating spatial interpolation models of weather variables for generating the inputs to the Canadian FWI system with the objective of determining whether they identify the same spatial interpolation model as having the lowest error. We found that LOOCV, shuffle‐split, stratified shuffle‐split, and a modified buffered leave‐one‐out procedure generally identified the same spatial interpolation models as having the lowest error. Spatial k‐fold favored spatial interpolation models with extrapolation ability. Our findings indicate that the most computationally efficient cross‐validation approach can be used for automatically selecting spatial interpolation models for weather surface generation, which will improve the quality of historical daily FWI maps.

Crown fuel consumption in Canadian boreal forest fires

Predictive crown fuel consumption models were developed using empirical data from experimental burning projects. Crown fuel load for foliage, bark, branchwood and stemwood were calculated for live overstorey and understorey trees in each plot using nationally derived tree biomass algorithms. Standing dead tree branchwood and stemwood biomass were similarly calculated. Crown bulk density values were calculated for all non-stemwood fuel components. Factors that affect the initiation and spread of crown fires (live crown base height, foliar moisture content, surface fuel consumption, critical surface fire spread rate, critical surface fire intensity) and components of the Canadian Forest Fire Weather Index System were not statistically significant variables. Crown bulk density was moderately correlated with crown fuel consumption but was not an influential factor. A new crown fuel consumption model was developed by regression analysis using fuel load of overstorey tree foliage and standing dead tree branchwood, and fire rate of spread through crown fraction burned. A simpler model was developed using only overstorey tree foliage fuel load and fire rate of spread. Both models provide forest and fire management agencies with enhanced ability to determine crown fuel consumption, fire behaviour and carbon emissions in boreal fires using basic forest inventory or biomass/carbon datasets.

Automatic forest fire danger rating calibration: Exploring clustering techniques for regionally customizable fire danger classification

Effective wildfire management begins with fire prevention and the assessment of the local fire danger. Typical approaches to fire danger classification rely heavily on manual analysis for specific regions based on expert knowledge. In this paper, a novel approach is proposed for the automatic calibration of fire danger classes based on the Canadian Forest Fire Weather Index System (CFFWIS) applied for specific regions. The proposed automatic calibration method is based on clustering algorithms, namely k-means, fuzzy c-means, Gaussian mixture models, and data-clouds, which are used to identify clusters in datasets composed of elements from CFFWIS and wildfire historical records. The clusters are associated with fire danger classes which are separated by proposed thresholds based on the fire weather index values contained within each cluster. Exhaustive experiments ensured an accurate comparison of performance with the analysis of our fire danger classes against the classes defined by the European Forest Fire Information System (EFFIS), also based on the CFFWIS. These experiments consider individual information from each of the selected European regions from a total of 769 regions with available data, with validation aimed at the analysis of the large fires in a general context, and a case study for Portuguese regions.

Global increase in wildfire risk due to climate-driven declines in fuel moisture.

There is mounting concern that global wildfire activity is shifting in frequency, intensity, and seasonality in response to climate change. Fuel moisture provides a powerful means of detecting changing fire potential. Here, we use global burned area, weather reanalysis data, and the Canadian fire weather index system to calculate fuel moisture trends for multiscale biogeographic regions across a gradient in vegetation productivity. We quantify the proportion of days in the local fire season between 1979 and 2019, where fuel moisture content is below a critical threshold indicating extreme fire potential. We then associate fuel moisture trends over that period to vegetation productivity and comment on its implications for projected anthropogenic climate change. Overall, there is a strong drying trend across realms, biomes, and the productivity gradient. Even where a wetting trend is observed, this often indicates a trend toward increasing fire activity due to an expected increase in fuel production. The detected trends across the productivity gradient lead us to conclude global fire activity will increase with anthropogenic climate change.

Development of Google Earth Engine Fire Weather Index Calculator for Indonesian Fire Danger Rating System

Tropical peatlands make up almost ten percent of the land surface in Indonesia, making peat fires detrimental not only for global atmospheric carbon levels, but also to public health and socioeconomic activities in the region. Indonesian Fire Danger Rating System (FDRS) was developed based on the Canadian Forest Fire Weather Index System (CFFWIS), using three different fuel codes and three indices representing fire behaviour. Daily Fire Weather Index (FWI) calculation is done by the Meteorological Climatological and Geophysical Agency (BMKG) with data from its synoptic weather stations network. Distribution of such weather stations are sparse, therefore this paper reports on the development of Fire Weather Index calculator on Google Earth Engine, using high resolution weather data, provided by weather model and remote-sensing open datasets. The resulting application is capable of generating daily maps of FWI components to be used by the Indonesian Fire Danger Rating System.

Fire prediction with logistic regression on territory of Bosnia and Herzegovina

The Canadian Fire Weather Index system [1] has been used worldwide by many countries as classic approach in fire prediction. It represents system that account for the effects of fuel moisture and weather conditions on fire behaviour. It numerical outputs are based on calculation of four meteorological elements: air temperature, relative humidity, wind speed and precipitation in last 24h. In this paper meteorological data in combination with Canadian Fire Weather Index system (CFWI) components is used as input to predict fire occurrence using logistic regression model. As logistic regression is a supervised machine learning method it’s based on user input in the form of dataset. Dataset is collected using NASA GES DISC Giovanni web-based application in the form of daily area-averaged time series in period of 31.7.2010 to 31.7.2020, it’s analysed and pre-processed before it is used as input for logit model. CFWI components values are not imported but calculated in run-time based on pre-processed meteorological data. As a result of this research windows application was developed to assist fire managers and all those involved in studying the fire behaviour.

A Fire Regime Zoning System for China

Understanding fire regimes is central to fire management. In our study, we analyzed the fire weather and forest fire regime of China using fire data from satellite remote sensing and statistics from 2000 to 2020. The fire weather index system was calculated from observed weather data for 2007 to 2017. Using qualitative and quantitative methods, we created a zoning system for China based on the spatial distribution characteristics of fire regimes and vegetation. The fire seasons varied between regions because of differences in vegetation, climate and ignition sources. The fire seasons in the north were spring and autumn. In the south and southwest forest regions, the fire seasons were winter and spring. Most forest fires occurred in southern China, but the average burned area per fire was lower compared with fires in the northeast. The zoning system includes 13 forest fire regime zones with specific fire characteristics according to quantitative variables. These zones are further divided into 17 fire regime units based on qualitative variables. Each fire regime unit has unique characteristics for regime, climate and vegetation type. Human activity was the main cause of fires, especially in south China, where the population density is high. Fire management should be tailored to each fire regime type based on fire characteristics and management targets.