Fire Frequency, Area Burned, and Severity: A Quantitative Approach to Defining a Normal Fire Year

Carl H Key,James A Lutz,Jan W Van Wagtendonk,Crystal A Kolden,Jonathan T Kane

doi:10.4996/fireecology.0702051

Carl H Key, James A Lutz + Show 3 more

Open Access

https://doi.org/10.4996/fireecology.0702051

Copy DOI

Abstract

Fire frequency, area burned, and fire severity are important attributes of a fire regime, but few studies have quantified the interrelationships among them in evaluating a fire year. Although area burned is often used to summarize a fire season, burned area may not be well correlated with either the number or ecological effect of fires. Using the Landsat data archive, we examined all 148 wildland fires (prescribed fires and wildfires) >40 ha from 1984 through 2009 for the portion of the Sierra Nevada centered on Yosemite National Park, California, USA. We calculated mean fire frequency and mean annual area burned from a combination of field- and satellite-derived data. We used the continuous probability distribution of the differenced Normalized Burn Ratio (dNBR) values to describe fire severity. For fires >40 ha, fire frequency, annual area burned, and cumulative severity were consistent in only 13 of 26 years (50 %), but all pair-wise comparisons among these fire regime attributes were significant. Borrowing from long-established practice in climate science, we defined “fire normals” to be the 26 year means of fire frequency, annual area burned, and the area under the cumulative probability distribution of dNBR. Fire severity normals were significantly lower when they were aggregated by year compared to aggregation by area. Cumulative severity distributions for each year were best modeled with Weibull functions (all 26 years, r2 ≥ 0.99; P < 0.001). Explicit modeling of the cumulative severity distributions may allow more comprehensive modeling of climate-severity and area-severity relationships. Together, the three metrics of number of fires, size of fires, and severity of fires provide land managers with a more comprehensive summary of a given fire year than any single metric.

Highlights

Fire frequency, fire extent, and fire severity are three of the seven fire regime attributes of particular importance to ecologists and land managers (Sugihara et al 2006) who closely monitor changes and trends in natural processes
We propose a set of fire regime metrics—annual fire frequency, annual area burned, and cumulative fire severity distribution—that can be monitored on a purely statistical basis using data available from the Monitoring Trends in Burn Severity (MTBS) database (Eidenshink et al 2007)
The years with the most extensive annual area burned are associated with high cumulative severity, and the years with low annual area burned are associated with ͳ ‫ ݕ‬ൌ ͳ ൅ ݁ିቀ௫ି௕௫బቁ

Summary

Introduction

Fire extent, and fire severity are three of the seven fire regime attributes of particular importance to ecologists and land managers (Sugihara et al 2006) who closely monitor changes and trends in natural processes. Changing fire regimes impact efforts to manage (Kolden and Brown 2010) and suppress fire. Few frameworks have been proposed to help land managers quantitatively examine these changes over time or between areas. We propose a set of fire regime metrics—annual fire frequency, annual area burned, and cumulative fire severity distribution—that can be monitored on a purely statistical basis using data available from the Monitoring Trends in Burn Severity (MTBS) database (Eidenshink et al 2007)

Methods

Results

Conclusion