PP12A-01 INVITED
Mineral Dust in Polar Ice Cores as a Paleoclimatic Proxy
Ice cores from polar regions provide a unique archive for paleo climate since a variety of proxies can be obtained from the same sample. One of these are mineral dust particles eroded from arid areas which have been entrained into the atmosphere, transported over large distances and deposited on the ice caps. Those dust particles are stored in an extremely clean environment. Due to the thickness of the polar ice sheets and an extremely low snow accumulation rate the longest record obtained from Antarctica up to now spans more than 800 000 years (Lambert, 2008), records in Greenland reach up to 120 000 years B.P. (NGRIP- community-members, 2004). On the other hand very high resolution records of seasonal cycles can be obtained up to the last glacial period, which allows studies ranging from annual to glacial-interglacial variability. Dust concentration changes in ice cores mirror changing conditions in the source and during the atmospheric transport of the dust. Measurements of concentration and size at the same time offer the possibility to distinguish between influences by transport and by the source of the dust. The dust provenance can be obtained by comparison of the chemical composition of the ice core dust with that in the potential source areas. An overview of the dust records from Antarctic and Greenland ice cores is presented with emphasis on dust concentration variability on different timescales and studies of provenance and transport changes.
PP12A-02 INVITED
Late Quaternary climate variability in the Sahel: inferences from marine dust records offshore Senegal
Societies and ecosystems in northern Africa are strongly affected by the availability of water. As a
consequence, long-term absence of rainfall has very dear effects on the ecosystems, as was dramatically
shown in the 70'ies and 80'ies of the 20th century.
Recent high-resolution reconstructions of Sahel palaeoclimate allow for new insights into these drastic
climate variations and to disentangle the effects of the different components of the climate system on African
climate change.
In this study we extend the instrumental record of climate variability using marine sediment cores that were
retrieved off the coast of Senegal, northwest Africa. The sediment records contain continuous high-resolution
records of dust sedimentation ranging from about 4,000 to about 57,000 years.
A 210Pb age model for the youngest sediments allows for a matching of the proxy rainfall record with
instrumental precipitation data. Specifically, variations in the grain-size distributions of the terrigenous
sediment fraction, deconvolved with an end-member modelling algorithm (Weltje, 1997) are used to
reconstruct rainfall variability on land throughout the late Quaternary.
http://www.marum.de/jbstuut.html
PP12A-03
Toward the Interpretation of Geological Dust Records: The Role of Winds in Modulating Dust Particle Size Distributions
Geological records of dust from ice cores, deep sea sediments and loess deposits are often used as an indicator for changes in environmental conditions. The observed dust variability in these records may result from changes in a variety of processes involving emission, transport and deposition. It has be hypothesized that variations in the average dust particle size observed in sediment and ice cores is due to changes in wind strength in the source regions or transport wind strength whereby an increase in wind speed is thought to be associated with an increase in the average particle size. On the otherhand, wind tunnel studies in source regions suggest that stronger winds can generate more small particles, as the stronger kinetic energy causes more abrasion. However, no observational proof is available for dust involved in longrange atmospheric transport towards remote ocean basins or the Earth's polar regions where dust particle size is usually less than 5 μm. In this study, we address this problem by using observations of particle size distributions and other dust properties derived from 10 Robotic Aerosol Network (AERONET) sites located in the desert regions of North Africa and the Middle East. AERONET observations of dust properties are based on the whole atmospheric column. Multiple measurements are available during day time covering periods of several years within the last decade. The AERONET data are compared with contemporaneous surface wind speed from meteorological stations close to the AERONET sites. At most AERONET sites the median effective radius is in the range of 1-2 μm, in contrast to many marine core measurements that show average particle diameters in the range of 4-8 μm, making interpretation difficult. We show that the median particle size significantly increases with wind speed at most stations which suggests that the paleoclimate winds vs. size hypothesis is usually valid. However, at a few sites the increase in particle size is not significant or a decrease is observed (e.g. Saih Salam). The observed particle size distributions are compared with results from simulations of the global dust cycle using the NCAR Community Climate System Model (NCAR/CCSM). The results suggest that winds play an important role in modulating dust size distributions and that wind strength is indeed a potential factor which can explain observed particle size variations in paleorecords of dust.
PP12A-04
Landscape-Scale Interpretation of the Loess Record
Intensive OSL dating and a growing array of paleoecological and paleoclimatic proxies hold great promise for extracting new insights from the loess record, but are by necessity applied to a limited number of sections. These well-studied sections need to be placed within the broader context of the complex, landscape-scale geomorphic, ecological, and pedological processes that interact to produce the loess record. Source- proximal locations where loess accumulation can be exceptionally rapid potentially offer high-resolution paleoenvironmental records; however, observations in the Great Plains, the Upper Mississippi Valley, and the Chinese Loess Plateau demonstrate that geomorphic complexity is also greatest in these proximal settings. Large variations in loess thickness over short distances, together with streamlined hills and fluted landscapes, are most plausibly interpreted as the result of spatially variable remobilization of loess by aeolian as well as hillslope processes. An intriguing hypothesis is that some striking wind-aligned landforms result from catastrophic wind erosion of coarse noncohesive proximal loess. Remobilization by any means can clearly lead to incomplete records, but less easily detected effects could also be important. Loess grain size and deposition rate decrease rapidly with distance from the proximal edge of the loess deposit. If that edge shifts over time through large-scale wind erosion or stabilization of loess in previously unstable areas, nearby locations could experience large changes in grain size or accumulation rate not directly linked to regional climate. These changes could in turn affect proxies that respond to local vegetation or to the effectiveness of pedogenic processes. For example, the stable carbon isotope composition of organic matter records a shift toward greater C3 plant abundance at some sites in the central Great Plains, during Holocene episodes of accelerated dust influx from nearby active dunefields, possibly analogous with vegetation change during the 1930s Dust Bowl. In the Great Plains, and possibly the Loess Plateau, this landscape-scale geomorphic complexity diminishes with increasing distance downwind from loess sources, though the potential resolution of the loess record also decreases along that gradient. Furthermore, there is evidence for widespread geomorphic stability even in source-proximal regions, at times of low loess influx and/or high effective moisture, such as the Pleistocene-Holocene transition on the central Great Plains.
PP12A-05
MIS-13 Climate, astronomical and ice sheets forcing
During the interglacial MIS-13 (about 500 ka ago), the loess in China recorded an exceptional East Asian summer monsoon (EASM), the strongest over the whole Quaternary. Other exceptional climate events have also been found worldwide, in particular over India and Africa. This is very surprising, because the oxygen isotopic composition of the deep-sea sediments and the deuterium temperature record of the EPICA ice core show that globally MIS-13 appears to be the most glaciated and cooler interglacial of the last one million years. To understand the impacts of the astronomical forcing and of the reconstructed ice sheets on the MIS-13 climate, in particular the monsoons, different sensitivity experiments have been made. First, the Greenland and Antarctica ice sheets are assumed to keep their present-day volume and the Eurasian and North American ice sheets are assumed to exist with different ice volumes ranging from the largest ones (amounting respectively 11.9 and 24.2 106 km3) up to 0. Equilibrium experiments show that the astronomical forcing is the most important but also that the Eurasian ice sheet has a tendency to reinforce the EASM except for its Last Glacial Maximum size when Northern Hemisphere summer occurs at aphelion. Transient experiments are made where Greenland and Antarctica ice sheets are interactively coupled to the atmosphere and oceans, in order to analyze the response of the climate system to fixed MIS-13 insolation and greenhouse gases forcings. A similar experiment using the long term variations of insolation is done. All these experiments are expected to provide a better understanding of the impact of the insolation and ice sheets on the MIS-13 climate and therefore a better interpretation of the related proxy records.
PP12A-06
An Environmental Magnetic Record of Quaternary Moisture Change From the Loess/Loessoid Deposits of the Central Eastern Pampas of Buenos Aires, Argentina
The paucity of long terrestrial climate records from the southern hemisphere has left a considerable gap in our understanding of land-sea and basin-to-basin interactions with respect to global climate change. The loess record of Argentina provides an important opportunity to bridge this gap, however, its complexity has resulted in a poorly understood climatological model of depositional and pedogenic processes. Here we present one of the longest and most continuous loess/loessoid records from the central eastern Pampas of Argentina. Based on changes in the amount of ultrafine-grained ferrimagnetic minerals and the relative abundance of goethite, we suggest a magnetoclimatological model for moisture transport such that the position of the South Atlantic Anticyclone (SAA), controlled by South Atlantic summer sea-surface temperatures (SSSTs), controls moisture distribution to subtropical and tropical South America as well as the equatorial Pacific. Cold interglacial SSSTs prior to the mid-Pleistocene Transition (MPT; ~1.0 Ma) resulted in a northward-displaced SAA and decreased (increased) moisture transport to the central eastern Pampas (Amazon Basin and the equatorial Pacific). Following the MPT, warmer interglacial SSSTs resulted in a southward-displaced SAA and increased (decreased) moisture transport to the central eastern Pampas (Amazon Basin and equatorial Pacific). Thus, the model shows that subtropical South America was significantly drier during the early to mid-Pleistocene (1.9-1.1 Ma) than it was during the late Pleistocene and Holocene (1.1 Ma-present). Our model not only provides the link between the South Atlantic and climate of tropical and subtropical South America, but it suggests the ocean-atmosphere teleconnection between the South Atlantic and the equatorial Pacific.
PP12A-07
North-Atlantic millennial-timescale variability imprint on Western European loess deposits: a modeling study
We test if the dust cycle response to the North-Atlantic abrupt climate changes, the Dansgaard-Oeschger (DO) and Heinrich (H) events, could have led to the recording of these events in the Western European loess deposits. The LMDZ AGCM is used to simulate a reference glacial state "DOS", assimilated to a DO stadial, a cold perturbation "HE", resembling a H event, and a warm perturbation "DOI", assimilated to a DO interstadial. The reference state corresponds to the 40-kyr BP context, in the middle of the typical glacial period (75-15 kyr BP). The two perturbations are obtained by applying cold, respectively warm anomalies of up to 2°C to the North-Atlantic surface temperatures in the latitudinal band 30°- 63°N. The simulated climates are compared from the point of view of dust emission, with a focus on the English Channel and the south of the North Sea (ECSNS), important dust sources for the Western European loess deposits. When only considering the changes in wind, precipitation, soil moisture and snow cover over the ECSNS and loess deposit areas, the differences of dust emission flux between the simulations are small, in contradiction with the observed stadial-interstadial loess sedimentation rate variations. However, when including the vegetation effect of inhibiting the eolian erosion, the three climates are clearly differentiated, the dust flux for DOI becoming less than half of that for DOS and HE. This shows that vegetation changes have played a key role in the stadial-interstadial dust cycle variations.
PP12A-08 INVITED
'DIRTMAP2': Dust and Palaeoclimate.
The influence of dust on climate, through changes in the radiative properties of the atmosphere and/or the
CO2 content of the oceans and atmosphere (through iron fertilisation of high nutrient, low chlorophyll, HNLC,
regions of the world's oceans), remains a poorly quantified and actively changing element of the Earth's
climate system. Dust-cycle models presently employ a relatively simple representation of dust properties;
these simplifications may severely limit the realism of simulations of the impact of changes in dust loading on
either or both radiative forcing and biogeochemical cycling. Further, whilst state-of-the-art models achieve
reasonable estimates of dust deposition in the far-field (i.e. at ocean locations), they under-estimate - by an
order of magnitude - levels of dust deposition over the continents, unless glacigenic dust production is
explicitly and spatially represented. The 'DIRTMAP2' working group aims to address these problems directly,
through a series of explicitly interacting contributions from the international modelling and palaeo-data
communities. A key aim of the project is to produce an updated version of the DIRTMAP database
('DIRTMAP2'), incorporating (a) records and age models newly available since ~ 2001, (b) longer records,
and especially high-resolution records, that will target time windows also focused on by other international
research programs (e.g. DO8/9, MIS5), (c) metadata to allow quality-control issues to be dealt with
objectively, (d) information on mineralogy and isotopes relevant to provenancing, radiative forcing and iron
bioavailability, and (e) enhanced characterisation of the aeolian component of existing records. This update
will be coordinated with work (led by Karen Kohfeld) to expand the DIRTMAP database to incorporate
information on marine productivity and improved sedimentation rate estimation techniques. It will also build
upon a recently-developed dust model evaluation tool for current climate (e.g. Miller et al. 2006) to enable
application of this and other evaluative models to palaeoclimate simulations. We invite colleagues to
contribute to this update; the DIRTMAP2 database will shortly be accessible from the University of Lancaster
website.
http://www.lec.lancs.ac.uk/research/LU_themes/inqua_working_group.php