Union [U]

U31A
 MC:Hall D  Wednesday  0800h

Origin of Late Holocene (Pre-Industrial) Increases in Atmospheric CO2 and CH4 I Posters


Presiding:  W F Ruddiman, University of Virginia; M Crucifix, Institut d""Astronomie et de Geophysique

U31A-0001

Evidence for the Postconquest Demographic Collapse of the Americas in Historical CO2 Levels

* Mannstein, H Hermann.Mannstein@dlr.de, Deutsches Zentrum für Luft- und Raumfahrt, Institute of Atmospheric Physics, Oberpfaffenhofen, Wessling, D-82234, Germany
Faust, F X fxfaust1@yahoo.com, Universidad del Cauca, Dept. of Antropology, Calle 5 No. 4-70, Popayan, 25, Colombia

In this talk we promote the hypothesis that the massive demographic collapse of the native populations of the Americas triggered by the European colonization brought about the abandonment of large expanses of agricultural fields soon recovered by forests, which in due turn fixed atmospheric CO2 in significant quantities. This hypothesis is supported by measurements of atmospheric CO2 levels in ice cores from Law Dome, Antarctica. Changing the focus from paleoclimate to global population dynamics and using the same causal chain, the measured drop in historic atmospheric CO2 levels can also be looked upon as further, strong evidence for the postconquest demographic collapse of the Americas.

U31A-0002

What was the probability of entering into glaciation ?

* Michel, C michel.crucifix@uclouvain.be, Institut d'Astronomie et de Geophysique G. Lemaitre, Universite catholique de Louvain, 2, chemin du Cyclotron, Louvain-la-Neuve, BE-1348, Belgium
Jonathan, R , Departement of Mathematics, University of Bristol, University Walk, Bristol, BS8 1TW, United Kingdom

In 2003 Bill Ruddiman formulated the hypothesis that the civilisations of antiquity exerted a perturbation significant enough to modify the natural trajectory of climate evolution. Instead of lowering of CO2 and glacial inception climate trajectory turned out to follow a long interglacial. Ruddiman's hypothesis is partly founded on an analogy principle. It is based on the observation that glacial inception occurred in the past when the circumstances of insolation and ice volume were similar to today. On the other hand, the absence of a visible signature of the supposed anthropogenic forcing on the carbon isotopic record shows that the direct human influence on the net atmospheric carbon balance must have been fairly small. How can these two observations be reconciled ? Ruddiman (2007) argued that the effects of the human perturbation may have been amplified by natural feedbacks. This is plausible. After all, most theories of glacial-interglacial cycles admit the existence of bifurcations and internal instabilities in the climate system; nothing allows us to reject the idea that the Holocene was close enough to instability to allow a human perturbation to tip the scales towards a long interglacial. On the other hand, the argument is risky. A dynamical system that amplifies a small perturbation is unpredictable. Therefore, the analogy principle may no longer be a valid argument. In this context, we consider adequate to address Ruddiman's hypothesis from a probabilistic (Bayesian) prospective. What was the probability of entering glacial inception during the Holocene, given prior information ? Did the human perturbation modify this probability ? To this end, we use the reduced-order model of glacial dynamics by Saltzman and Maasch (1991). Depending on the chosen parameters, this models may exhibit a limit cycle between glacial and interglacial climates, phase-locked by the orbital forcing. We use SPECMAP and Vostok data to calibrate the model on observations. Once calibrated, the model is used to predict the future of climate, starting either at 8 kyr BP or from the present day. In either case, the prediction is a rise in CO2 to 280 ppmv at present and then a slow decrease. Glacial inception approximately occurs in 50,000 years.

U31A-0003

The archaeobotany of Asian rice expansion and the development of wet-field cultivation

* Fuller, D d.fuller@ucl.ac.uk, University College London, 31-34 Gordon Square, Institute of Archaeology, London, WC1H 0PY, United Kingdom

Archaeobotanical evidence provides direct data on past human diet and agriculture, including a geographical and chronological framework for studying the expansion of rice agriculture. The growth of systematic archaeobotanical sampling in recent years has allowed for the past presence of rice to be seen in relation to cultivation of other crops and associated weeds. The weed flora provides a basis for inferring the nature of cultivation systems, whether rain-fed dry rice or wetland "paddy" rice, a key distinction for considerations of past methane production. Nevertheless, current data is very unevenly distributed. This poster will summarize available evidence for the origins and spread of rice in South Asia (India and Pakistan), and mainland and Island Southeast Asia deriving from an earlier Chinese domestication. Where possible, such as in India or China, the potential of the weed flora remains for distinguishing wetland rice crops will be summarized. In broad terms, although the origins of rice use and cultivation begins by or during the Middle Holocene (6000- 3000 BC), rice cultivation spread outside the regions of the wild progenitor after this time. Two phases of rice expansion can be distinguished. Phase 1, between 3000 and 1500 BC, introduced rice to Southeast Asia, probably under wetland cultivation, and the spread of dry rice over northern India and Pakistan. Phase 2, taking place between 1000 and 0 BC, sees the spread of rice throughout the Southern Indian Peninsula, with weed evidence suggesting irrigated wetland rice. Similarly, this period sees the spread of intensive paddy agriculture through Korea and Japan, but in Southeast Asia is probably related to a spread of rice in upland, dry field systems.

U31A-0004

Effects of Syn-pandemic Fire Reduction and Reforestation in the Tropical Americas on Atmospheric Carbon Dioxide During European Conquest

* Nevle, R J rnevle@bcp.org, Bellarmine College Preparatory, 960 West Hedding St., San Jose, CA 95126-1215, United States
Bird, D K dbird@pangea.stanford.edu, Department of Geological and Environmental Sciences, Stanford University, 450 Serra Mall, Braun Hall, Building 320, Stanford, CA 94305-2115, United States

A new reconstruction of the Late Holocene biomass burning history of the tropical Americas is consistent with expanding fire use by Mesoamerican and Amazonian agriculturalists from 2000-500 BP and a subsequent period of fire reduction due to indigenous demographic collapse. Our reconstruction synthesizes published data from 50 charcoal accumulation records obtained from stratified lacustrine sediments and from soils, including soil charcoal records recovered from archeological sites. Synthesis of stratigraphic charcoal records yields indexes of the mean rate of regional charcoal accumulation and of variability in charcoal accumulation among sites during 500-year increments since 3500 BP. The age distribution of dated soil charcoal particles from non-archeological sites provides an independent measure of variation in regional charcoal accumulation; whereas age distribution of soil charcoal dates from archeological sites records variation in charcoal accumulation related to anthropogenic biomass burning. We observe that the charcoal accumulation indexes derived from stratigraphic records begin to increase at 2000 BP, remain high until 500 BP, and then decline to near-minimum values during the 500-year period subsequent to European contact. Similarly, the age distributions of soil charcoal dated from both non-archeological and archeological sites indicate increases in charcoal accumulation from 2000 to 500 BP followed by decline. An index of the inter- site variability in charcoal accumulation obtained from the stratigraphic records attains a maximum during the time period between 1000 and 500 BP and a near-minimum value afterward. We interpret the covariation between measures of charcoal accumulation derived from archeological and non-archeological sites as a consequence of the expansive influence of anthropogenic activity on the regional fire regime. Increases in regional charcoal accumulation apparent in both the stratigraphic and soil charcoal records beginning at 2000 BP correlate with expanding indigenous population, agriculture, and fire use in the tropical Americas. The rise in inter-site variability in charcoal accumulation after 2000 BP is consistent with a demographic shift toward sedentary agrarian communities and localized increases in charcoal accumulation in densely populated centers. Declines in regional charcoal accumulation and inter-site variability after 500 BP suggest a correlative cause related to reduction in anthropogenic biomass burning resulting from pandemic-driven population collapse. Published reconstructions of Pre-Columbian demography indicate that during European conquest, pandemics killed ~90% of the indigenous American population (~60 million), estimated to represent ~20% of the 16th century global population. Our predictive calculations suggest that fire reduction in the tropical Americas is associated with massive forest regeneration on ~5 x 105 km2 of land and sequestration of 5-10 Gt C into the terrestrial biosphere, which can account for 13- 50% of the ~2% global reduction in atmospheric CO2 levels and the 0.1‰ increase in δ13C of atmospheric CO2 from 1500 to 1700 CE recorded in Antarctic ice cores and tropical sponges. New archeological discoveries revealing extensive networks of geoglyphs and urban polities in Pre-Columbian Amazonia suggest that our estimates of reforestation, and consequent effects on atmospheric CO2, may be conservative.

U31A-0005

Holocene Concentrations of Methane in the Atmosphere are in Part Proportional to Concentrations of Sulfur Dioxide and Inversely Proportional to the Oxidizing Capacity of the Atmosphere

* Ward, P L peward@wyoming.com, Teton Tectonics, P.O. Box 4875, Jackson, WY 83001, United States

The atmosphere cleans itself by oxidizing pollutants. The primary oxidant is the hydroxyl radical (OH) formed by photodissociation of ozone in the near ultra-violet. Ozone and OH are in limited supply. Sulfur dioxide (SO2) absorbs near ultraviolet light limiting production of OH and reacts immediately with any available OH, forming sulfuric acid. Methane reacts more slowly with OH and will typically not be oxidized until there is little SO2. Thus a high concentration of methane indicates low oxidizing capacity. The rate at which SO2 is injected into the atmosphere controls oxidizing capacity and climate change in four ways: 1. Moderate rate: Large volcanic eruptions (VEI >=6) lower global temperatures for a few years when they are separated by years to decades so the oxidizing capacity of the atmosphere can fully recover. In 1991, Pinatubo volcano in the Philippines erupted 20 Mt SO2 and 491 Mt H2O, the largest volcanic eruption since 1912. The SO2 was oxidized primarily by OH to form a 99% pure aerosol of sulfuric acid and water at an elevation of 20-23 km. This aerosol reflected sunlight, lowering the world's temperature on average 0.4°C for three years. Ozone levels were reduced by 10%. Methane increased by 15 ppb for a year. The e-folding time for SO2 was 35 days. 2. High rate: When large eruptions occur once to several times per year, there is insufficient oxidizing capacity leading to increases in methane and other greenhouse gases and global warming. There were 15 times in the Holocene when large volcanoes erupted on average at least every year for 7 to 21 years. Man is now putting as much SO2 from burning fossil fuels into the atmosphere every year as one large volcano, causing current global warming. The two previous times were from 818-838 AD, the onset of the Medieval Warming Period, and from 180-143 BC, the onset of the Roman Warm Period. 3. Low rate: When there are no large eruptions for decades, the oxidizing capacity can catch up, cleaning the atmosphere, removing most of the methane and other pollutants. A clean atmosphere leads to cooling and drought. The 8.2 ka event is a classic example, but similar decadal droughts around 6.2, 5.8, 5.4, 4.2, and 2.9 ka caused the demise of major civilizations. 4. Extreme rate: Whereas large volcanic eruptions produce 10-1000 km3 of andesitic and silicic tephra, flood basalt eruptions produce as much as 3,000,000 km3 of basalt containing 10 to 100 times more SO2 per km3. The result is runaway global warming, widespread acid rain, and mass extinctions. The link between SO2 and global warming is good news because we have developed many efficient technologies that burn fossil fuels with less SO2 emission and scrub SO2 out of smoke stacks. Efforts to reduce acid rain have been successful in reducing manmade emissions of SO2 by >20% since 1980 and thereby reducing methane concentrations. Sudden increases in methane during the Pleistocene Dansgaard-Oeschger events follow sudden increases in volcanism. High rainfall especially in the Sahara and high methane concentrations in the early Holocene are clearly related to increased volcanism that brought about the end of the Ice Age. Increases in global warming at 3170 BC, 161 BC, and 828 AD are contemporaneous with short-term increases in methane. The rapid increase in SO2 from burning fossil fuels since 1850 can explain much of the corresponding rapid increase in methane. But during the last 5000 years, volcanism has been relatively constant and thus it can not explain the observed gradual increase in methane.

http://www.tetontectonics.org

U31A-0006

Holocene Greenhouse Gases And Asian Monsoon Climate

* Zhou, X zhouxin@mail.iggcas.ac.cn, Institute of Geology and Geophysics, Chinese Academy of Sciences, 19, Bei Tu Cheng Xi Road, Chaoyang District P.O. Box 9825, Beijing, 100029, China
Wu, H haibin_wu2001@yahoo.com, Institut des sciences de l' environnement, Université du Québec à Montréal, Montréal, QC H3C 3P8, Canada
Guo, Z ztguo@mail.iggcas.ac.cn, Institute of Geology and Geophysics, Chinese Academy of Sciences, 19, Bei Tu Cheng Xi Road, Chaoyang District P.O. Box 9825, Beijing, 100029, China

Atmospheric methane concentration gradually decreased during the first half of the Holocene and then reversed the trend since ~5,000 years ago. This has been variously attributed to natural or/and anthropogenic factors. The development of early rice farming in the world was roughly consistent in time with the methane reverse while the extent of irrigated lands and the amount of methane emission remain to be determined. It was also proposed that the late Holocene increase in the concentration of greenhouse gases would have offset part of the cooling trend due to decreased northern insolation and may have prevented the onset of the early stage of a new glaciation, as is supported by some climate model experiments. If this cooling occurred, it would have been equally catastrophic to the development of human civilization. As one of the tests to this hypothesis, we synthesized climate data from the Asian monsoon zone to examine the Holocene trends of monsoon precipitation and temperature changes: if the late Holocene changes of greenhouse gases had significant climate effects, we would expect a relatively stable temperature associated with a declined trend of monsoon rainfall in response to the insolation changes. The results effectively showed divergent trends of precipitation and temperature in the late Holocene for a number of the reconstructions, but regional complexity is also clear and the cause remains to be addressed. More accurate reconstructions of climate parameters are of particular importance.

U31A-0007

Effects of Pre-industrial Agricultural Expansion and Epidemics on the Climate and the Carbon Cycle

Pongratz, J julia.pongratz@zmaw.de, International Max Planck Research School on Earth System Modelling, Bundesstrasse 53, Hamburg, 20146, Germany
Pongratz, J julia.pongratz@zmaw.de, Max Planck Institute for Meteorology, Bundesstrasse 53, Hamburg, 20146, Germany
* Raddatz, T thomas.raddatz@zmaw.de, Max Planck Institute for Meteorology, Bundesstrasse 53, Hamburg, 20146, Germany
Reick, C christian.reick@zmaw.de, Max Planck Institute for Meteorology, Bundesstrasse 53, Hamburg, 20146, Germany
Claussen, M martin.claussen@zmaw.de, Meteorological Institute, University of Hamburg, Bundesstrasse 55, Hamburg, 20146, Germany
Claussen, M martin.claussen@zmaw.de, Max Planck Institute for Meteorology, Bundesstrasse 53, Hamburg, 20146, Germany

To assess the effects of anthropogenic land cover change on the pre-industrial climate and carbon cycle we apply a new, detailed reconstruction of land cover for the last millennium in a general circulation model. A transient simulation including the marine and terrestrial carbon cycle suggests that the agricultural expansion increased the atmospheric CO2 concentration by about 3.5 ppm between AD 800 and the late pre- industrial period. Taking into account land cover change prior to the last millennium, up to 5 ppm of the Holocene CO2 increase may be attributed to changes in vegetation and soil carbon as a consequence of agricultural activity. This value is smaller but of similar magnitude than the estimates by Ruddiman (2007). In contrast to his study the ocean is simulated to be a sink rather than a source of carbon at least during the last millennium, leaving much of the observed pre-industrial CO2 increase unexplained. On a regional scale, epidemics have the potential to change land cover by allowing natural vegetation to regrow on abandoned agricultural areas. While the land cover reconstruction indicates only small absolute changes in agricultural areas after European conquest of the Americas, it indicates forest regrowth on about 0.18 million km2 in Europe as a consequence of the medieval Black Death. For this event, simulations of radiative forcing show that the energy balance is significantly altered by the changes in surface albedo. This suggests that local to regional climate may be modified by the biogeophysical effects of vegetation changes induced by epidemics. First results, however, indicate that the amount of carbon taken up by the regrowing vegetation may not suffice to counterbalance the emissions of expanding agriculture in the other parts of the world. The effect on atmospheric CO2 concentrations may thus be small. Ensemble simulations are planned to compare the effects of epidemics on atmospheric CO2 with natural variability.

U31A-0008

Anthropogenic Deforestation and its Effect on the Carbon Cycle of Europe Over the Past Three Millennia

* Kaplan, J O jed.kaplan@epfl.ch, Ecole Polytechnique Federale de Lausanne, Institute of Environmental Science and Technology, EPFL-ENAC-ARVE Station 2, Lausanne, 1015, Switzerland
Krumhardt, K M kristen.krumhardt@epfl.ch, Ecole Polytechnique Federale de Lausanne, Institute of Environmental Science and Technology, EPFL-ENAC-ARVE Station 2, Lausanne, 1015, Switzerland

Over the past three millennia, both climate and anthropogenic land use and land cover change (LULUC) have substantially affected the European landscape. Anthropogenic deforestation for agriculture and pasture has been the most significant of these land cover changes, though climate variability itself may have had an impact on European ecosystems. In this study we attempt to quantify the influence of both LULUC and climate change on the carbon cycle of Europe during preindustrial time, and speculate on the ramifications for global atmospheric composition and biogeochemical feedbacks to the climate system. To quantify the effect of millennial-scale climate change and LULUC on the carbon cycle over the past three millennia, we assembled spatially explicit datasets of these quantities and ran a dynamic global vegetation model (LPJ-DGVM) in a number of experiments and sensitivity tests on a high-resolution grid for Europe. Climate data needed to run LPJ were synthesized from gridded datasets of mean monthly temperature and precipitation based on multiproxy climate reconstructions. Though it is certain that many European countries were substantially deforested before 1850, no coherent data set of the progression of deforestation that occurred during preindustrial time was available to us. We have therefore created a 10km, annually resolved gridded time series of European LULUC for the past three millennia by digitizing and synthesizing a database of population history for Europe and finding a relationship between population density, land quality for agricultural and pastoral activities, and anthropogenic deforestation. With these input data, we ran a series of experiments and sensitivity tests with LPJ to simulate the effect that changes in climate, LULUC and length- of-run (starting the run at 1700, 1850 or 1900) have on European carbon storage and its trajectory at year 2000. Climate variability in Europe over the past three millennia years caused modest reductions in carbon stored in living biomass, which outweigh increases in soil carbon, leading to a net loss of ca. 10 Pg of carbon over the most recent 500 years. In contrast, the time-history of increasing LULUC intensity over the past three millennia years leads to substantial reductions in both living biomass and soil carbon, including ca. 145 Pg over the most recent 500 years. Combining the effects of climate and land cover change results in a smaller total reduction in terrestrial carbon storage compared to LULUC only. Thus, climate change appears to ameliorate the amount of carbon lost after anthropogenic deforestation, as cooler temperatures suppress microbial respiration of soil organic matter. Length-of-run experiments indicate that the terrestrial biosphere is sensitive to the time history of both climate and LULUC.

U31A-0009

Global Anthropogenic CO2 Emissions Through Vegetation Clearance for Agriculture During the Last 6000 Years

* Hickler, T Thomas.Hickler@nateko.lu.se, Department of Physical Geography and Ecosystems Analysis, Sölvegatan 12, Lund, S-22362, Sweden
Olofsson, J Jorgen.Olofsson@nteko.lu.se, Department of Physical Geography and Ecosystems Analysis, Sölvegatan 12, Lund, S-22362, Sweden
Miller, P A Paul.Miller@nateko.lu.se, Department of Physical Geography and Ecosystems Analysis, Sölvegatan 12, Lund, S-22362, Sweden
Sykes, M T Martin.Sykes@nateko.lu.se, Department of Physical Geography and Ecosystems Analysis, Sölvegatan 12, Lund, S-22362, Sweden

The mechanisms underlying the development of atmospheric CO2 over the Holocene and the potential role of anthropogenic greenhouse gas forcing in pre-industrial times are still highly debated. We developed a global gridded data set of human land use for the last 6000 years, including permanent and shifting cultivation. The data set was mainly based on archaeological evidence on the global distribution of different types of human societies (empires and agricultural groups), the HYDE data base of land use since 1700, global population estimates, and assumptions concerning cultivation area per person. A dynamic global vegetation model (LPJ) was run with and without human land-use, and the difference in terrestrial carbon storage was assumed to represent the total anthropogenic carbon release to the atmosphere. Modeled total carbon release during the industrial period (A.D. 1850-1990) was 148 gigatons of carbon (GtC), of which 33 GtC originated from non-permanent agriculture. For pre-industrial times (4000 B.C. - A.D. 1850), the net carbon release was 79 GtC from permanent agriculture and 35 GtC from non-permanent agriculture. Modeled carbon release between 4000 and 0 B.C. was considerably lower than would be required for a substantial influence on the climate system. However, the extent of vegetation clearing before the year 1700 is highly uncertain. We suggest that various lines of evidence, and pollen analyses in particular, should be explored in order to test the hypothesis that many areas that were forested at the beginning of the industrial revolution had been cleared earlier. Even though the carbon storage in vegetation might have been restored in such areas, soil carbon storage could have been negatively affected. In summary, our results suggest that a substantial early anthropogenic impact on atmospheric CO2 is unlikely, but important uncertainties remain. We are currently addressing some of the uncertainties through a sensitivity analyses of the modeling approach presented here.

http://www.springer.com/earth+sciences/journal/334

U31A-0010 [WITHDRAWN]

Potential Impacts of Paleohydrological Changes on Holocene Methane Fluxes in Boreal and Subarctic Peatlands, James Bay, Quebec, Canada

* Garneau, M garneau.michelle@uqam.ca, GEOTOP - Department of Geography, Universite du Quebec a Montreal, C.P.8888 Succ. Centre-Ville, Montreal, QC H3C 3P8, Canada
Ali, A adam.ali@uqat.ca, GEOTOP - Department of Geography, Universite du Quebec a Montreal, C.P.8888 Succ. Centre-Ville, Montreal, QC H3C 3P8, Canada
Tremblay, L louis@flbtechnologie.ca, GEOTOP - Department of Geography, Universite du Quebec a Montreal, C.P.8888 Succ. Centre-Ville, Montreal, QC H3C 3P8, Canada
Pelletier, L pelletier.luc@uqam.ca, GEOTOP - Department of Geography, Universite du Quebec a Montreal, C.P.8888 Succ. Centre-Ville, Montreal, QC H3C 3P8, Canada
Asnong, H asnongh@hotmail.com, GEOTOP - Department of Geography, Universite du Quebec a Montreal, C.P.8888 Succ. Centre-Ville, Montreal, QC H3C 3P8, Canada

In boreal and subarctic region of the La Grande river watershed, James Bay, Quebec, Canada, peatlands cover closed to 15 % of the terrestrial surface. Multi proxy analysis results (plant macrofossils and Testate amoebae) from minerotrophic peatland have demonstrated important variations on the regional water table position since peat started to accumulate in the region ca 7400 cal BP. Macrofossil assemblages indicate that sites were first colonized by black spruce (Picea mariana Ait Muhl.) and Sphagnum spp which paludified with a regional rise of moisture at approx. 4500 BP. Drier conditions registered around 3900 cal BP induced a shift in vegetation and Testate amoeba assemblages for a relatively short period which was followed at approximately 3000 cal BP by an important increase in moisture. This shift in hydrological conditions involved drastic changes in the vegetation cover from Picea mariana and Sphagnum fuscum assemblages to sedges (Carex spp.) and wet Sphagnum species such as S. majus, S. subsecundum, S. pulchrum. This rise in the water table position could have induced enhance methane release to the atmosphere when considering the present-day methane fluxes/water table depth/vegetation cover relationship.

U31A-0011

Climate and Human Influences on Global Biomass Burning Over Past Millennia

* Marlon, J , University of Oregon, Dept. of Geography, Eugene, OR 97403, United States
(GPWG), G , School of Geographical Sciences, University of Bristol, Bristol, UK BS8 1SS, United Kingdom

Human-induced biomass burning has been proposed as one of the causes of the anomalous increases in atmospheric CO2 and CH4 during the millennia prior to the industrial era (as compared with previous interglacial periods). In particular, increased anthropogenic biomass burning (and rice irrigation) has been invoked to explain the methane anomaly (Ruddiman, 2007). We compiled sedimentary charcoal records spanning six continents to document global trends in biomass burning since the Last Glacial Maximum (LGM) 21,000 years ago. We also investigated global and regional trends in biomass burning during the last 2000 years in greater detail, and compared them with trends in CO2 and CH4. Burning increased from the LGM to the early Holocene (10,000 cal yr BP) as a result of increasing temperature and vegetation productivity. Biomass burning varied less during the Holocene; it declined slightly from 9000 to 7000 cal yr BP, and increased slightly from 7000 to 2000 cal yr BP. From 2000 to 250 cal yr BP declining global temperatures caused a gradual decline in biomass burning. Large human impacts on fire are not evident until 1750 A.D. at the global scale, when a rapid surge in burning occurred due to land clearance and land-cover changes associated with European colonization and population growth. Around 1900 A.D., biomass burning declined abruptly despite exponential growth in population and warming. This decline is consistent with the rapid expansion of agriculture, intensive grazing, urbanization and fire suppression. There seems to be little correlation between biomass burning and CO2 over the past millennium, but our reconstructions of biomass burning are consistent with unusual trends in δ 13CH4 from ice core data (Ferretti et al., 2005). These findings suggest that biomass burning has been driven primarily by climate changes rather than human influences until 1750 A.D.

U31A-0012

A Centrifuge-Based Technique for Dry Extraction of Air for Ice Core Studies of Carbon Dioxide.

Grachev, A M gracheal@geo.oregonstate.edu, Department of Geosciences, Oregon State University, 104 Wilkinson Hall, Corvallis, OR 97331, United States
* Brook, E J brooke@geo.oregonstate.edu, Department of Geosciences, Oregon State University, 104 Wilkinson Hall, Corvallis, OR 97331, United States

High resolution CO2 data from the Law Dome ice core document an abrupt ~10 ppm drop in CO2 at about 1600 AD (MacFarling Meure et al., Geophys. Res Lett., v. 33, L14810), which has been attributed to changes in human activities. CO2 measurements in ice cores are difficult, however, making verification of this feature an important task. We are undertaking a high-resolution study of CO2 between 1400 and 1800 AD in the WAIS Divide (Antarctica) ice core with a new dry extraction technique. The need for a dry extraction technique as opposed to a melt-refreeze technique in studies of CO2 from ice cores arises because of the well-documented artifacts in CO2 imposed by the presence of liquid water. Three dry-extraction methods have been employed by previous workers to measure CO2: needle-crushing method, ball-bearings method, and cheese-grater method (B. Stauffer, in: Encyclopedia of Quaternary Science, p. 1181, Elsevier 2007). Each has limitations, and we propose a simpler dry extraction technique, based on a large-capacity refrigerated centrifuge (the "centrifuge technique"), which eliminates the need to employ cryogenic temperatures to collect extracted gas and is more compatible with high sample throughput. The technique is now being tested on ~25-gram WAIS Divide samples in conjunction with CO2 measurements with a gas chromatograph. The technique employs a Beckman J- 6B centrifuge, in which evacuated stainless steel flask is placed: the flask has a weight inside positioned directly over a tall-standing piece of ice whose cross-section is small compared to that of the flask. Upon acceleration to 3000 rpm the weight moves down and presses the ice sample into a thin tablet covering flask's bottom, yielding the air extraction efficiency of ~80%. Preliminary tests suggest that precision and accuracy can be achieved at the level of ~1 ppm once the system is fine-tuned.