The geography of fire: A paleo perspective
Marlon, Jennifer R.
Fire is a fundamental, transformative, yet poorly understood process in the Earth system; it can radically reorganize ecosystems, alter regional carbon and energy balances, and change global climate. Short-term fire histories can be reconstructed from satellite (seasonal- to interannual-scales), historical (decadal scales), or dendrochronological records (for recent centuries), but only sedimentary charcoal records enable an analysis of the complex interactions between climate, vegetation and people that drive fire activity over longer temporal scales. This dissertation describes the compilation, synthesis and analysis of a global paleofire dataset and its application to understanding past, current, and future changes in fire activity. Specifically, I co-led efforts to compile charcoal records around the world into a single database, and to conduct three meta-analyses to understand the controls on fire at multiple spatial and temporal scales. The first meta-analysis reconstructed global biomass burning since the Last Glacial Maximum (LGM) 21,000 years ago. Results from this study demonstrated that global fire activity is low when conditions are cool and high when conditions are warm. This fundamental relationship between climate and fire is due in large part to associated changes in vegetation productivity. The second meta-analysis examined fire activity in North America during past abrupt climate changes and looked for evidence of continental-scale wildfires associated with a hypothesized comet impact ∼13,000 years ago. This analysis found a correlation between increased fire activity and abrupt climate change, but provided no evidence for continental-scale wildfires. A final meta-analysis disentangled the climate and human influences on global biomass burning during the past 2000 years; it found a close relationship between climate change and biomass burning until ∼1750 A.D., when human activities became a primary driver of global fire activity. Together, these three meta-analyses demonstrate that climate change is the primary control of global fire activity over long time scales. In general, global fire activity increases when the Earth's climate warms and decreases when climate cools. The paleofire data and analyses suggest that the rapid climate changes projected for coming decades will lead to widespread increases in fire frequency and biomass burning. This dissertation includes previously published and unpublished co-authored material.