USDA Forest Service
 

Fire and Environmental Research Applications Team

 
 

Fire and Environmental Research Applications Team
Pacific Wildland Fire Sciences Laboratory

400 N 34th Street, Suite 201
Seattle, WA 98103

(206) 732-7800

Logo of the Pacific Northwest Research Station

United States Department of Agriculture Forest Service.

USDA Link Forest Service Link

 

Magnifying glass iconSpatial Pattern and Structure in Low-Severity Fire Regimes

FERA researchers developed and maintain a spatially explicit fire-history database (Figure 1) from seven watersheds on the Okanogan-Wenatchee and Colville National Forests.  With funding from the Joint Fire Science Program we produced a multi-scale analysis of the relationships between climate and spatio-temporal patterns in historical fire regimes in the inland Pacific Northwest.  We continue to use these data to explore spatial pattern and structure in the time series of fire events recorded by the 5000+ fire-scarred trees in the 7 watersheds.  These data present us with a unique opportunity to explore landscape ecology theory and the spatial structure of contagious disturbance.  Our current focus is in three areas:

I. Spatio-Temporal Information in Fire-Scar Records

Every recorder tree represents a temporal record, often 200 yr or longer, of fire on the landscape.  With hundreds of recorder trees in a watershed, each one georeferenced, we have a multivariate dataset with annual resolution whose information content far exceeds that of contemporary fire atlases and instrumental data.  The multivariate structure is lost when the data are used for aggregate statistics such as composite fire return intervals (Falk et al. 2007).  In contrast, we built variogram-like models, using the Sorensen’s Distance to represent the dissimilarity of individual fire-scar records and how it changes with the distance between trees.  We have shown (Kennedy and McKenzie 2010) that the Sorensen’s Distance has a probabilistic interpretation in terms of the likelihood of one tree’s recording a fire conditional on another’s doing so.  This provides us with a robust analytical tool for establishing scaling laws and making inferences about spatial patterns of historical fires.Location map and maps fo 6  fire history research sites across Washington state.

Figure 1. Fire history sites east of the crest of the Cascade Mountains, Washington

II. Topographic Complexity and Scaling Relationships

Topography in eastern Washington mountains varies from very complex and rugged near the crest of the Cascade Range to gentler in the Okanogan Highlands and northeastern part of the state.  We developed multivariate and univariate measures of topographic complexity and found that it was directly related to power-law behavior in the Sorensen’s Distance variograms (Figure 2).  Power laws arise in many natural phenomena, and researchers have had varying degrees of success in attributing physical, biological, or social mechanisms to them.  We hypothesize here that in the topographically more complex watersheds, barriers to fire spread and related patchiness of fuels provide complex fine-scale controls on the spatial patterns of fires, giving rise to power laws in the relationship between dissimilarity in fire-scar records and their separation in space.

Variograms

Figure 2. Sorensen variograms in log-log space (i.e., linear relation suggests power law). Topographically complex landscapes are more linear; shorter fire return intervals decrease variance. Data are from Kellogg et al. (2008)

III. Neutral-Model Simulations to Replicate Fire History

We developed a raster model that probabilistically spreads fire across a gridded landscape and scars simulated recorder trees randomly located across the landscape.  On this “neutral” landscape we tuned two parameters -- the probability of fire spreading from a burning cell to its neighbors and the probability of a recorder tree in the cell scarring if the cell burns -- to replicate the multivariate structure of fire scar records represented by the Sorensen’s Distance variograms, using Monte Carlo significance tests.  We were able to match the empirical variograms in each of the seven watersheds, suggesting that stochastic processes of varying complexity (depending on topographic complexity) were sufficient to reproduce the observed fire histories.  Individual simulated fires had spatial properties expected from real fires in complex (irregular spatial structure and jagged perimeters) and simple (regular patterns and smooth perimeters) topography.


Publications

Falk, D.A., C. Miller, D. McKenzie, and A.E. Black. 2007. Cross-scale analysis of fire regimes. Ecosystems 10:809-823.
Abstract and full text [.html]

Kellogg, Lara-Karena B., D. McKenzie, D.L. Peterson, and A.E. Hessl. 2008. Spatial models for inferring topographic controls on historical low-severity fire in the eastern Cascade Range of Washington, USA. Landscape Ecology 23:227-240.
Abstract and full text [.html]

Kennedy, M.C.; McKenzie, D. 2010. Using a stochastic model and cross-scale analysis to evaluate controls on historical low-severity fire regimes. Landscape Ecology. 25: 1561-1573.
Abstract and full text [.html]

McKenzie, D., A.E. Hessl, and L.-K.B. Kellogg. 2006. Using neutral models to identify constraints on low-severity fire regimes. Landscape Ecology 21:139-152.
Abstract and full text [.html]

McKenzie, D.; Kennedy, M.C. 2011. Scaling laws and complexity in fire regimes. In: McKenzie, D.; Miller, C.; Falk, D.A., eds. The Landscape Ecology of Fire. Springer-Verlag, The Netherlands: 27-49. Chapter 2.
Full text [.pdf]

McKenzie, D.; Kennedy, M.C. 2012. Power laws reveal phase transitions in landscape controls of fire regimes. Nature Communications. doi:10.1038/ncomms1731


Project Leads: Don McKenzie and Maureen Kennedy

Disclaimers | FOIAPrivacy Policy | Quality of InformationPrint This Page
USDA logo which links to the department's national site. Forest Service logo which links to the agency's national site.