Mapping the probability of large fire occurrence in northern Arizona, USA

Brett G Dickson, John W. Prather, Yaguang Xu, Haydee M. Hampton, Ethan N. Aumack, Thomas D Sisk

Research output: Contribution to journalArticle

82 Citations (Scopus)

Abstract

In the southwestern U.S., wildland fire frequency and area burned have steadily increased in recent decades, a pattern attributable to multiple ignition sources. To examine contributing landscape factors and patterns related to the occurrence of large (≥20 ha in extent) fires in the forested region of northern Arizona, we assembled a database of lightning- and human-caused fires for the period 1 April to 30 September, 1986-2000. At the landscape scale, we used a weights-of-evidence approach to model and map the probability of occurrence based on all fire types (n = 203), and lightning-caused fires alone (n = 136). In total, large fires burned 101,571 ha on our study area. Fires due to lightning were more frequent and extensive than those caused by humans, although human-caused fires burned large areas during the period of our analysis. For all fires, probability of occurrence was greatest in areas of high topographic roughness and lower road density. Ponderosa pine (Pinus ponderosa)-dominated forest vegetation and mean annual precipitation were less important predictors. Our modeling results indicate that seasonal large fire events are a consequence of non-random patterns of occurrence, and that patterns generated by these events may affect the regional fire regime more extensively than previously thought. Identifying the factors that influence large fires will improve our ability to target resource protection efforts and manage fire risk at the landscape scale.

Original languageEnglish (US)
Pages (from-to)747-761
Number of pages15
JournalLandscape Ecology
Volume21
Issue number5
DOIs
StatePublished - Jul 2006

Fingerprint

event
lightning
road
ability
resources
evidence
roughness
vegetation
resource
modeling

Keywords

  • Fire risk
  • Lightning
  • Ponderosa pine
  • Topographic roughness
  • Weights of evidence
  • Wildland fire

ASJC Scopus subject areas

  • Earth and Planetary Sciences (miscellaneous)
  • Nature and Landscape Conservation
  • Ecology
  • Geography, Planning and Development

Cite this

Mapping the probability of large fire occurrence in northern Arizona, USA. / Dickson, Brett G; Prather, John W.; Xu, Yaguang; Hampton, Haydee M.; Aumack, Ethan N.; Sisk, Thomas D.

In: Landscape Ecology, Vol. 21, No. 5, 07.2006, p. 747-761.

Research output: Contribution to journalArticle

Dickson, Brett G ; Prather, John W. ; Xu, Yaguang ; Hampton, Haydee M. ; Aumack, Ethan N. ; Sisk, Thomas D. / Mapping the probability of large fire occurrence in northern Arizona, USA. In: Landscape Ecology. 2006 ; Vol. 21, No. 5. pp. 747-761.
@article{e9429be45a414a0989419b3df4a895c0,
title = "Mapping the probability of large fire occurrence in northern Arizona, USA",
abstract = "In the southwestern U.S., wildland fire frequency and area burned have steadily increased in recent decades, a pattern attributable to multiple ignition sources. To examine contributing landscape factors and patterns related to the occurrence of large (≥20 ha in extent) fires in the forested region of northern Arizona, we assembled a database of lightning- and human-caused fires for the period 1 April to 30 September, 1986-2000. At the landscape scale, we used a weights-of-evidence approach to model and map the probability of occurrence based on all fire types (n = 203), and lightning-caused fires alone (n = 136). In total, large fires burned 101,571 ha on our study area. Fires due to lightning were more frequent and extensive than those caused by humans, although human-caused fires burned large areas during the period of our analysis. For all fires, probability of occurrence was greatest in areas of high topographic roughness and lower road density. Ponderosa pine (Pinus ponderosa)-dominated forest vegetation and mean annual precipitation were less important predictors. Our modeling results indicate that seasonal large fire events are a consequence of non-random patterns of occurrence, and that patterns generated by these events may affect the regional fire regime more extensively than previously thought. Identifying the factors that influence large fires will improve our ability to target resource protection efforts and manage fire risk at the landscape scale.",
keywords = "Fire risk, Lightning, Ponderosa pine, Topographic roughness, Weights of evidence, Wildland fire",
author = "Dickson, {Brett G} and Prather, {John W.} and Yaguang Xu and Hampton, {Haydee M.} and Aumack, {Ethan N.} and Sisk, {Thomas D}",
year = "2006",
month = "7",
doi = "10.1007/s10980-005-5475-x",
language = "English (US)",
volume = "21",
pages = "747--761",
journal = "Landscape Ecology",
issn = "0921-2973",
publisher = "Springer Netherlands",
number = "5",

}

TY - JOUR

T1 - Mapping the probability of large fire occurrence in northern Arizona, USA

AU - Dickson, Brett G

AU - Prather, John W.

AU - Xu, Yaguang

AU - Hampton, Haydee M.

AU - Aumack, Ethan N.

AU - Sisk, Thomas D

PY - 2006/7

Y1 - 2006/7

N2 - In the southwestern U.S., wildland fire frequency and area burned have steadily increased in recent decades, a pattern attributable to multiple ignition sources. To examine contributing landscape factors and patterns related to the occurrence of large (≥20 ha in extent) fires in the forested region of northern Arizona, we assembled a database of lightning- and human-caused fires for the period 1 April to 30 September, 1986-2000. At the landscape scale, we used a weights-of-evidence approach to model and map the probability of occurrence based on all fire types (n = 203), and lightning-caused fires alone (n = 136). In total, large fires burned 101,571 ha on our study area. Fires due to lightning were more frequent and extensive than those caused by humans, although human-caused fires burned large areas during the period of our analysis. For all fires, probability of occurrence was greatest in areas of high topographic roughness and lower road density. Ponderosa pine (Pinus ponderosa)-dominated forest vegetation and mean annual precipitation were less important predictors. Our modeling results indicate that seasonal large fire events are a consequence of non-random patterns of occurrence, and that patterns generated by these events may affect the regional fire regime more extensively than previously thought. Identifying the factors that influence large fires will improve our ability to target resource protection efforts and manage fire risk at the landscape scale.

AB - In the southwestern U.S., wildland fire frequency and area burned have steadily increased in recent decades, a pattern attributable to multiple ignition sources. To examine contributing landscape factors and patterns related to the occurrence of large (≥20 ha in extent) fires in the forested region of northern Arizona, we assembled a database of lightning- and human-caused fires for the period 1 April to 30 September, 1986-2000. At the landscape scale, we used a weights-of-evidence approach to model and map the probability of occurrence based on all fire types (n = 203), and lightning-caused fires alone (n = 136). In total, large fires burned 101,571 ha on our study area. Fires due to lightning were more frequent and extensive than those caused by humans, although human-caused fires burned large areas during the period of our analysis. For all fires, probability of occurrence was greatest in areas of high topographic roughness and lower road density. Ponderosa pine (Pinus ponderosa)-dominated forest vegetation and mean annual precipitation were less important predictors. Our modeling results indicate that seasonal large fire events are a consequence of non-random patterns of occurrence, and that patterns generated by these events may affect the regional fire regime more extensively than previously thought. Identifying the factors that influence large fires will improve our ability to target resource protection efforts and manage fire risk at the landscape scale.

KW - Fire risk

KW - Lightning

KW - Ponderosa pine

KW - Topographic roughness

KW - Weights of evidence

KW - Wildland fire

UR - http://www.scopus.com/inward/record.url?scp=33746164335&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33746164335&partnerID=8YFLogxK

U2 - 10.1007/s10980-005-5475-x

DO - 10.1007/s10980-005-5475-x

M3 - Article

AN - SCOPUS:33746164335

VL - 21

SP - 747

EP - 761

JO - Landscape Ecology

JF - Landscape Ecology

SN - 0921-2973

IS - 5

ER -