Characteristics and Controls of Extremely Large Wildfires in the Western Mediterranean Basin. Fernandes, P. M., Barros, A. M. G., Pinto, A., & Santos, J. A. Journal of Geophysical Research: Biogeosciences, 121(8):2141–2157, August, 2016.
doi  abstract   bibtex   
Large fires account for a disproportionally high percentage of area burned with potentially severe environmental and socioeconomic impacts. This study characterizes extremely large fires (ELFs; 2500-24,843\,ha) in Portugal (1998-2013) and the concomitant fuel and weather conditions, analyzing the response of ELF size to their variation. ELF burned less shrubland-grassland (33\,% of the total ELF area) than forest (59\,% of total), the latter primarily composed by pine and pine-eucalypt. High fuel hazard was the norm, as indicated by median values of 0.98 for fuel load as a fraction of potential (maximum) load and time since fire $>$14\,years over 91\,% of the burned area. ELF occurred under anticyclonic circulation patterns, especially ridging, and 78\,% of them coincided with extreme fire danger days (corresponding to infrequent conditions) in conjunction with unstable atmosphere. Containment time, fire growth rate, and energy release metrics varied by 1 more order of magnitude than ELF size, hence indicating that size alone is insufficient to describe extreme fires. Distinct combinations between ambient weather conditions, atmospheric instability, and drought defined three categories of ELF as defined by size. Quantile regression indicated that increasingly larger fires showed gradually stronger responses to fire weather severity, highlighting the difficulty in restraining fire spread in flammable landscapes in the absence of extensive fuel treatments. Data limitations inherent to the methods used are discussed, and improvements to further advance the understanding of extreme fires are suggested.
@article{fernandesCharacteristicsControlsExtremely2016,
  title = {Characteristics and Controls of Extremely Large Wildfires in the Western {{Mediterranean Basin}}},
  author = {Fernandes, Paulo M. and Barros, Ana M. G. and Pinto, Anita and Santos, Jo{\~a}o A.},
  year = {2016},
  month = aug,
  volume = {121},
  pages = {2141--2157},
  issn = {2169-8953},
  doi = {10.1002/2016jg003389},
  abstract = {Large fires account for a disproportionally high percentage of area burned with potentially severe environmental and socioeconomic impacts. This study characterizes extremely large fires (ELFs; 2500-24,843\,ha) in Portugal (1998-2013) and the concomitant fuel and weather conditions, analyzing the response of ELF size to their variation. ELF burned less shrubland-grassland (33\,\% of the total ELF area) than forest (59\,\% of total), the latter primarily composed by pine and pine-eucalypt. High fuel hazard was the norm, as indicated by median values of 0.98 for fuel load as a fraction of potential (maximum) load and time since fire {$>$}14\,years over 91\,\% of the burned area. ELF occurred under anticyclonic circulation patterns, especially ridging, and 78\,\% of them coincided with extreme fire danger days (corresponding to infrequent conditions) in conjunction with unstable atmosphere. Containment time, fire growth rate, and energy release metrics varied by 1 more order of magnitude than ELF size, hence indicating that size alone is insufficient to describe extreme fires. Distinct combinations between ambient weather conditions, atmospheric instability, and drought defined three categories of ELF as defined by size. Quantile regression indicated that increasingly larger fires showed gradually stronger responses to fire weather severity, highlighting the difficulty in restraining fire spread in flammable landscapes in the absence of extensive fuel treatments. Data limitations inherent to the methods used are discussed, and improvements to further advance the understanding of extreme fires are suggested.},
  journal = {Journal of Geophysical Research: Biogeosciences},
  keywords = {*imported-from-citeulike-INRMM,~INRMM-MiD:c-14174965,~to-add-doi-URL,burnt-area,empirical-equation,fire-weather-index,mediterranean-region,portugal,regression,statistics,wildfires},
  lccn = {INRMM-MiD:c-14174965},
  number = {8}
}

Downloads: 0