Predicting spatial patterns of fire on a southern California landscape

Abstract

Humans influence the frequency and spatial pattern of fire and contribute to altered fire regimes, but fuel loading is often the only factor considered when planning management activities to reduce fire hazard. Understanding both the human and biophysical landscape characteristics that explain how fire patterns vary should help to identify where fire is most likely to threaten values at risk. We used human and biophysical explanatory variables to model and map the spatial patterns of both fire ignitions and fire frequency in the Santa Monica Mountains, a human-dominated southern California landscape. Most fires in the study area are caused by humans, and our results showed that fire ignition patterns were strongly influenced by human variables. In particular, ignitions were most likely to occur close to roads, trails, and housing development but were also related to vegetation type. In contrast, biophysical variables related to climate and terrain (January temperature, transformed aspect, elevation, and slope) explained most of the variation in fire frequency. Although most ignitions occur close to human infrastructure, fires were more likely to spread when located farther from urban development. How far fires spread was ultimately related to biophysical variables, and the largest fires in southern California occurred as a function of wind speed, topography, and vegetation type. Overlaying predictive maps of fire ignitions and fire frequency may be useful for identifying high-risk areas that can be targeted for fire management actions.

Human-caused fire occurrence modelling in perspective: a review

“…Usually, lower elevation (Kalabokidis et al 2007;Sebastián-López et al 2008;Kwak et al 2012;Narayanaraj and Wimberly 2012;Liu and Wimberly 2015) and smaller slope gradient (Preisler et al 2004;Syphard et al 2008;Dondo Bühler et al 2013;Oliveira et al 2014;Argañaraz et al 2015) increase HCF occurrence. Since surface temperature and humidity are affected by terrain, these may be reflecting climatic conditions.…”

“…The location of human activities is highly dependent on site-related variables that determine the number and distribution of human sources of ignition. Human presence can be analysed from explicit spatial factors such as proximity to, or density of, infrastructure such as roads (Dickson et al 2006;Yang et al 2008Yang et al , 2015Gralewicz et al 2012b;Hegeman et al 2014;Syphard and Keeley 2015;Zhang et al 2016;Mhawej et al 2016;Vilar et al 2016b), tracks (Pew and Larsen 2001;Romero-Calcerrada et al 2008, trails (Syphard et al 2008;Vilar del Hoyo et al 2011;Arndt et al 2013) and railways (Sturtevant and Cleland 2007;Guo et al 2015;Alcasena et al 2016), all of which are associated with an increase in fire occurrence. For example, in Spain (MAGRAMA 2015), the United States (Morrison 2007) and south-eastern Australia (Penman et al 2013), more than half of HCFs start along road systems.…”

Modelling long-term fire regimes of southern California shrublands

“…1c). In the SMM, 155 of the 161 fires from 1981 to 2003 were anthropogenic in origin, the remaining six were due to lightning strikes (National Park Service 2005), and anthropogenic ignitions have been shown to preferentially occur close to roads (Keeley and Fotheringham 2003;Syphard et al 2008). We tested (i) spatially homogeneous and (ii) spatially correlated ignition probabilities.…”

Modelling natural disturbances in forest ecosystems: a review

“…Human-caused ignitions depend on the presence of people and their respective activities. Fire ignition as a function of human and/or biophysical explanatory variables is often modelled using generalized linear models such as logistic, Poisson or negative binomial regression (e.g., Wotton et al, 2003;Martinez et al, 2009;Syphard et al, 2008), generalized linear mixed models (Díaz-Avalos et al, 2001;González-Olabarria et al, 2010), through direct gradient analyses (e.g., Viedma et al, 2009), weight of evidence (e.g., Romero-Calcerrada et al, 2008), using neural network models (e.g., Vega-García and Chuvieco, 2006), or fuzzy logic (Loboda and Csiszar, 2007). However, many widely applied dynamic landscape models, simulating individual fire events explicitly, are based on descriptive parameters of the fire regime only, e.g., average return intervals and maximum (and sometimes also minimum) fire sizes (e.g., Mladenoff and He, 1999).…”