Terrestrial Records of Holocene Climate Change: Fire, climate and humans

Science Center Objects

Large wildfires have raged across the western Americas in the past decade including the Las Conchas, New Mexico fire that burned 44,000 acres in a single day in 2011 (Orem and Pelletier, 2015, Geomorphology 232: 224-238, and references therein), the 2016 Fort McMurray, Alberta fire that required evacuating an entire city, and the 2015 Alaskan fire season that burned more than 5 million acres (Alaska Interagency Coodination Center). These fires are caused by a changing climate resulting in hotter, drier conditions across much of the western North America, and are augmented by land-use practices resulting in more potential forest fuel. It is essential to place these fires in a longer temporal context to examine if recent fires are anomalous or if they have occurred in the past under diverse climate conditions.


Figure 1: Dark aerosols from recent wildfires may increase the melting of the Juneau Icefield.

Dark aerosols from recent wildfires may increase the melting of the Juneau Icefield.

Recent peat fires in Indonesia demonstrate that biomass burning causes carbon dioxide emissions that can be as much as 50% of those from fossil-fuel combustion and so are highly likely to influence future climate change (Stocker, and others, 2013, Technical Summary, in Climate Change 2013: The Physical Science Basis: Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change: Cambridge University Press; Global Fire Emissions Database) However, aerosols—and especially fire aerosols—continue to be one of the least understood aspects of the modern climate system and even less is known about their past influence (Stocker, and others, 2013). The impact of these aerosols when they fall back to the surface is unknown. Wildfire aerosols may accelerate glacier melt through depositing dark material on reflective glaciers, thereby decreasing the glacier's albedo. The effect of fire aerosols on the water quality local and regional watersheds is also not well known.

Project Lead Natalie Kehrwald and collaborators developed methods to determine molecular markers of fire activity in ice and lake cores (e.g. Gambaro and others 2008, Analytical Chemistry 80 (5), 1649–1655; Kehrwald and others, 2012, Tellus Series B – Chemical and Physical Meteorology, 64, 18196; Zennaro and others, 2014, Climate of the Past, 10, 5, 1905-1924 and Zennaro and others, 2015, Geophysical Research Letters, 42, 12, 5023-5033). Levoglucosan is a specific biomarker in that it can only be produced by cellulose burning at temperatures centered around ~250°C (Kuo and others, 2011, Chemosphere, 85, 5, 797-805). Levoglucosan is incorporated into smoke plumes, transported in the atmosphere, and deposited through wet and dry deposition. The presence of levoglucosan can be used to determine fire histories in paleoclimate records through the Holocene (Zennaro and others, 2014 and Zennaro and others, 2015). High resolution records of levoglucosan in short ice cores can indicate whether fire aerosols are accelerating glacier melt. Investigating levoglucosan concentrations in modern streams can help determine if fire aerosols affect surface water quality.

Gilkey Glacier, Juneau Icefield, Alaska

Interactions between glaciers, bedrock, and surface debris on the Gilkey Glacier, Juneau Icefield, Alaska.


Photo flat glacier surface, Juneau Icefield, Alaska

An ideal ice core site is the highest, flattest glacier in a region. In 2016, a transect of 7-9 m ice cores was drilled on the Matthes Glacier, Juneau Icefield to determine if recent fires are affecting the glacier surface. (Photo: Lucas Foglia, used with permission


Pueblo del Arroyo, Chaco Canyon

Pueblo del Arroyo, Chaco Canyon. (Photo: National Park Service)

Humans have had control over fire for at least the last 1 million years (Berna and others, 2012, PNAS, 109, 20, E1215-E1220). This "control" is relative, however, as human ignitions can accidentally cause wildfires such as the 2016 Nederland, Colorado fire that started from an improperly doused campfire. In areas that are not naturally fire-prone, biomass burning residue in lake cores is sometimes used as an indicator of the presence of humans in an area. New techniques using specific biomarkers can determine if and when humans lived in a region, and if the presence of humans coincides with increased fire activity.

However, specific biomarkers can help determine the timing and interpretation of climatic and anthropogenic events. For example, the timing of arrival of humans to Chaco Canyon, New Mexico is not well known. Humans may have arrived in multiple waves, and certain groups may have had a greater effect on the landscape through increased biomass burning. Changing climate conditions coupled with deforestation may have influenced the ability of residents to continue living in the canyon. Fecal sterols indicate the presence of humans in an area, and by investigating these markers in tandem with levoglucosan (Battistel and others, 2015, Analytical and Bioanalytical Chemistry, 28, 8505-8514) the interactions and impacts of fire, climate, and human activity in a region such as Chaco Canyon can be reconstructed.

The climate of the southwestern U.S. is projected to become hotter and drier in the upcoming decades, mirroring climate conditions that lead in part to the decline of civilizations such as the one in Chaco Canyon. Lake core records can help establish the roles of humans and climate in such societal shifts. The investigation of organic tracers is expanding the limits of proxy information and provides data for one of the least understood aspects of the climate system, with implications for how past societies responded to a changing climate.


Llewellyn Glacier, Juneau Icefield

The rapidly degrading Llewellyn Glacier, Juneau Icefield.

This project investigates the interactions between a warming climate, increasing fire activity and impacts on people initially by addressing the following questions:

  1. Are recent wildfires accelerating the melt of the Juneau Icefield or other glaciers?

    Biomass burning deposits aerosols on snow surfaces. Dark aerosols from fossil fuel burning are known to increase glacier surface melt over sub-seasonal timescales, yet the effect of wildfire aerosols on glacier surfaces is currently unknown. We are collecting data on levoglucosan concentrations in short (7-9 meter) snow cores from the Juneau Icefield in conjunction with stable isotopes, major ions, and dust concentrations to determine if and how wildfire aerosols affect surface melt relative to other factors such as warmer temperatures and increased dust deposition.

  2. What role did climate and land use change associated with burning vegetation for land clearance play in the human history of Chaco Canyon, New Mexico?

    Scientists do not definitively know when humans first arrived in Chaco Canyon. Currently, the estimated date of arrival is based in part on the increase in fire activity and/or a shift in pollen suggesting anthropogenic land use change. Combining fecal sterol concentrations with levoglucosan data can help determine the timing of the arrival of humans and discover if increased fire activity coincides with the presence of people. Data from lake cores demonstrate the changing vegetation in the region, but currently do not include specific markers for determining the presence of humans. Investigating fecal sterols as a marker of human presence in Chaco Canyon sediment cores can specifically determine when people lived in an area, and provide a value-added component to previous USGS studies.


In July and August, 2016, Natalie Kehrwald (USGS GECSC), Sarah Fortner (Wittenberg University), Shad O’Neel (USGS), and a team of students used a Kovacs drill to obtain a transect of 7 to 9 meter short cores across the Juneau Icefield, Alaska. This project is a joint effort with the Juneau Icefield Research Program (JIRP). Project staff and collaborators are returning to the Juneau Icefield in June and July, 2017 in order to obtain more ice cores.

Location of the Juneau Icefield in relation to other Alaskan glaciers

Location of the Juneau Icefield in relation to other Alaskan glaciers. (Map background: ESRI World Imagery)


Drilling firn cores on the Juneau Icefield, Alaska

Molly Peek and Chris Miele drilling and processing firn cores on the Juneau Icefield, Alaska.


Chemical markers provide fire data across a wide array of time scales and types of information. These biomarkers are especially valuable as they are present in both ice and lake cores. We use levoglucosan, and its isomers mannosan and galactosan, to reconstruct past fire history from ice cores and lake sediments.

Table of specific fire markers

Specific fire markers such as levoglucosan provide fire history information over wide spatial and temporal resolutions. (Figure modified from Conedera and others, QSR, 2009)

Our previous research demonstrates the applicability of using levoglucosan in ice cores to determine past biomass burning and the role of humans in altering fire activity (figure below). The NEEM, Greenland ice core contains fire biomarkers throughout the Holocene. The peak in fire activity centered around ~2500 years before present cannot be explained by climate variables such as increased temperature or summer insolation. Model results for the potential source areas do not show an increase in area burnt unless human activity is taken into account. We ascribe this peak fire activity to land clearance for agriculture (see Zennaro and others, 2015 for full details).

One of the most enticing questions in fire science is determining how fires ignited. Defining if a natural ignition such as lightning or human ignition such as an out-of-control campfire caused a wildfire is currently not possible. However, biomarkers such as fecal sterols provide direct proof that humans were in an area. Combining the presence of fecal sterols with levoglucosan can show the presence of humans in a local area and any possible correction with changes in fire activity (Battistel and others, 2016, The Holocene, 1-11).

Figure showing  The NEEM, Greenland ice core records fire activity over thousands of years.

The NEEM, Greenland ice core records fire activity over thousands of years. The map illustrates the possible source areas and transport paths for the fire products. Levoglucosan concentrations indicate peak biomass burning occurring ~2500 years before present where this increased combustion occurs separately from associated climate variables such as temperature (see Zennaro and others, 2015 for full details).

Laboratory Methods

Through a collaboration with Larry Barber and Jeramy Jasmann (National Research Program, Boulder, Colorado), new methods are being developed to determine the specific fire biomarkers levoglucosan, mannosan and galactoan in ice and lake cores using a gas chromatography coupled to tandem mass spectrometry. This method allows for separating levoglucosan and its isomers as well as using initial sample sizes of 1 mL or less. Current methods for determining levoglucosan in snow and ice use different instrumentation that may not be able to separate levoglucosan from its isomers, or may use a larger initial sample (Gambaro and others, 2008, Analytical Chemistry 80 (5), 1649–1655; Kuo and others, 2011; You and others, 2016, Talanta, 148, 534-538). This collaborative work is expanding the fire history methods to incorporate analysis of fecal sterols in lake sediments in order to determine both biomass burning and the presence of humans (Battistel and others, 2015).

Group photo of Glacial Biogeochemistry Team

2016 Glacial Biogeochemistry Team: Front row (left to right): Sarah Fortner, Kiana Ziola, Auri Clark, Natalie Kehrwald. Back row (left to right): Kristen Aruell, Chris Miele, Annie Zaccharin, Kit Cunningham, Annie Holt, Molly Peek.