COS 237-3 - Global trends in individual fire events

Thursday, August 18, 2022

10:30 AM – 10:45 AM EDT

Location: 516B

Presenting Author(s)

TT

Ty Tuff, n/a

Data Scientist
Earth Lab, CIRES, University of Colorado, United States

Coauthor(s)

AM

Adam L. Mahood, n/a

Postdoctoral Researcher
Earth Lab, CIRES, University of Colorado
Boulder, Colorado, United States
ES

E. Natasha Stavros, n/a

Director
Analytics Hub, Earth Lab, CIRES, University of Colorado, United States
AD

Amy DeCastro, n/a

Ph.D. Candidate
Department of Geography, University of Colorado at Boulder, United States
RN

R. Chelsea Nagy, n/a

Interim Director
Earth Lab, CIRES, University of Colorado
Boulder, CO, United States
JS

Jilmarie Stephens, n/a

Postdoctoral Researcher
Earth Lab, CIRES, University of Colorado, United States
MC

Megan E. Cattau, n/a

Associate Professor
Boise State University, United States
TH

Thomas E. Hanson, n/a

Ph.D. Candidate
Department of Anthropology, Univesity of Colorado at Boulder, United States
VI

Virginia Iglesias, n/a

Research Scientist
Earth Lab, CIRES, University of Colorado, United States
MC

Maxwell C. Cook, n/a

Ph.D. Student
Department of Geography, University of Colorado at Boulder
Longmont, Colorado, United States
JB

Jennifer K. Balch, n/a

Director
Earth Lab, CIRES, University of Colorado, United States

Background/Question/Methods

In 2020, members of Earth Lab published an algorithm to delineate individual fire events using the MODIS fire data products. The result of that work originally produced a dataset describing the size and duration of every fire in the USA on a daily basis, but over the last two years, we have been able to scale that effort to include the entire globe and produce an event-delineated fire dataset for every flammable surface on Earth. We analyze this global dataset using a series of median-based linear models and cluster analyses to highlight the traits that unify fire behavior globally and identify climate-fire combinations that are increasing or decreasing over time.

Results/Conclusions

Our results show that fires mirror many of the global trends that have previously been used to describe ecological systems, like metabolic scaling and latitudinal gradients in productivity. For example, a log-log plot of fire size vs. fire growth rate shows agreement between the data and a positive, sublinear trend line with a slope of 0.68, just as we see in metabolic rates of species or cities scaling across multiple orders of magnitude. These results support the idea that fuel and climate are fundamental components to fire behavior and that the ultimate size and duration of an individual fire is driven by constraints imposed by the ecology of their fuel and their compatibility with the local climate. Trend analysis shows regional patterns highlight the relative influences of human activity and climate on fire activity. Many countries in the Boreal region, for example, are experiencing an increase in fire frequency, but many countries in equatorial regions are experiencing a decline; this is likely due to the outsized influence of human activity. We explore the mechanisms driving those human-impact counter-examples as well as those who seem to buck the global trend because of their unique geography, ecology, or climate.