Dr. Jennifer Marlon is awarded an NSF Earth Sciences Postdoctoral Fellowship to carry out a research and education plan at the University of Wisconsin - Madison. Dr. Marlon will use paleoenvironmental data and a dynamic global vegetation-fire model to study past climate change impacts on fire-vegetation interactions. She will first conduct meta-analyses of charcoal and pollen data from lake sediments to reconstruct centennial- to millennial-scale patterns of fire and vegetation change at regional to continental scales. A cutting-edge, fire-enabled dynamic global vegetation model will then be used to simulate paleofire activity under climate conditions very different from today (i.e., 6000 and 21,000 years ago). Results from the charcoal and pollen analyses will be used to assess model performance through data-model comparisons. Finally, results from the synthesis and data-model comparison will be used to design and implement sensitivity experiments with the coupled vegetation-fire model. The aim of the experiments will be to advance our understanding of how climate and vegetation control fire regimes, and of the range of responses in vegetation and fire to potential future climate changes.
The likely impacts of global warming on vegetation and wildfires at broad spatial scales are largely unknown. Yet, we know that projected climate changes will have major consequences for ecosystem dynamics, including disturbance regimes, and that feedbacks from changing fire regimes in particular will affect global biogeochemical cycles, as well as atmospheric chemistry and physics. Such consequences, in turn, threaten ecosystem services, such as the provisioning of clean air and water, and thus human health and livelihoods. Paleofire records indicate that fire regimes are highly sensitive to climate changes, particularly when the changes are abrupt, and regardless of the direction of change. This research will improve our understanding of natural variability in fire regimes and thus improve our ability to model future changes in fire. Results will also provide ecologists and natural resource managers with critical context for current fire-regime changes in North America. Outreach and education activities associated with this research include the development of standardized and integrated datasets of past fire and vegetation changes that will be made available to the broader community on a public website. Dr. Marlon will also develop course material on fire, paleoecology and paleoclimatology for undergraduates and graduates at the host institution.
Wildfire occurrence and size has been increasing in the western U.S., Canada, and in parts of Europe, Asia, Australia, and elsewhere over the past few decades. Damages from these fires, their gas and particulate emissions, and the costs of managing them, have had significant impacts not only on people, communities and economies, but also on natural resources, ecosystems, and even global climate. This fellowship supported research on past fires because these can tell us how and why current wildfires are unique, and what forces are shaping the changes. Long fire-history records (paleofire data) in particular, give us a unique window into the slow "background" processes that take decades and longer to unfold. Such processes, which include changes in temperature, precipitation, vegetation composition, and land use, can have subtle but strong impacts on wildfires, so understanding them improves our ability to detect, evaluate, and manage fire. Results from this project include 1) a detailed reconstruction of fire history and climate change in the Western U.S. during the past 3000 years; 2) a global analysis of fire and its relationships to people and climate during the Holocene (i.e., the past 12,000 years, when agriculture was developing and expanding); and 3) an analysis of fire activity in eastern Asia during the Holocene based on black carbon and charcoal records across the region. In the Western U.S., three independent sources of fire history data were analyzed with temperature and drought to establish a predictive model of fire activity over the past 1400 years (Figure 1). The model was used to identify a large gap, or "fire deficit," between observed levels of buring in the West and those expected given prevailing climate conditions, which are warm and dry compared to past centuries. Current "megafires" in the West are sympotmatic of the large fire deficit, which resulted in large part from land-use changes during the past century. The study also suggests that wildfires will grow in coming decades if temperatures and drought continue to increase. The global analysis of Holocene fire activity yielded regional, continental, and global reconstructions of biomass burning. Levels of burning have generally tracked fluctuations in temperature and moisture throughout the Holocene, but human impacts on fire can also be detected in the records, particularly in recent centuries. Interestingly, maximum biomass burning occurred about 2000 years ago in the global record, and this finding is difficult to explain given what we know about changes in climate and human activity around this time. New global fire, climate and vegetation models, as well as archaeological data syntheses will be needed to understand what caused the high levels of burning 2000 years ago. A third analysis in eastern Asia was conducted in part to determine whether soot could be clearly identified and measured in lake sediments. I worked with colleagues in China to compare soot measurements from Lake Daihai with a regional reconstruction of fire activity from 36 nearby charcoal records. Our reconstructions showed strong similarities, providing support for the new analytical methods as well as evidence for a close link between fire activity and moisture availability in eastern Asia. The fellowship had broader societal impacts through its support of both domestic and international collaborations, opportunities for graduate students, the development of reports on public understanding of climate change, and the expansion of a Global Charcoal Database. The database (available online at gpwg.org) is one of the largest global data sets of Quaternary paleoenvironmental records, and one of the few data sets that has global coverage reflecting a single environmental process (fire). Importantly, the fellowship is also stimulating new research and collaborations on many aspects of fire that are only poorly understood, including estimates of global fire emissions and the causes of "megafires."
