Quaternary Research 53, 62-69 (2000)
Copyright 2000 by the University of Washington.

Christian M. Zdanowicz, Gregory A. Zielinski, and Cameron P. Wake
Climate Change Research Center, EOS, University of New Hampshire, Durham, New Hampshire 03824

and

David A. Fisher and Roy M. Koerner
Terrain Sciences, Geological Survey of Canada, 601 Booth Street, Ottawa, Ontario KJA OE8 Canada

Received January 21, 1999

An > 11,550-yr-long record of atmospheric dust deposition was developed from an ice core (P95) drilled through the Penny ice cap, Baffin Island. The P95 record documents environmental changes that affected the production and transport of dust in the eastern Canadian Arctic during the late Pleistocene and Holocene. Dust deposition on the Penny ice cap was greatest in late-glacial time when the climate was dry and windy and comparatively low in the Holocene. Microparticles deposited during late-glacial time are finer than in Greenland cores, suggesting distinct dust sources and transport trajectories to each region. Dust deposition at the P95 site increased after ca. 7800 yr ago as the Penny ice cap receded and distance from local dust sources was reduced. Deflation of newly exposed marine sediments on southwestern Baffin Island may have contributed to this dust increase. The P95 and GISP2 (Greenland) dust records show diverging trends in the middle to late Holocene, reflecting the growing influence of regional environmental conditions (e.g., dust source area, snow cover extent) on atmospheric dust deposition. This study further demonstrates how valuable records of regional-scale paleoenvironmental changes can be developed from small circumArctic ice caps, even those affected by considerable melt. copyright 2000 University of Washington

Key Words: ice cores; paleoclimate; eollan dust; Holocene; Arctic.

INTRODUCTION

Proxy climate records from ice cores drilled in the Canadian' Norwegian, and Russian Arctic document regional-scale climatic variability in the circumpolar north (e.g., Koerner and Fisher, 1990; Fujii et al., 1990; Kotlyakov et al., 1991) and complement the long, high-resolution records developed from central Greenland cores. Among other proxies, the composition, concentration, and size distribution of insoluble microparticles in ice cores provide information about environmental changes that affected the sources, transport, and deposition of dust (e.g., Petit et al., 198 1; Thompson and Mosley-Thompson, 1981; Steffensen, 1997; Zielinski and Mershon, 1997). Studies of dust in polar ice cores have been, with a few exceptions (e.g., Koerner, 1977), mostly restricted to Greenland and Antarctica. Here we report on a new ice core record of dust deposition from the Penny ice cap (67o14'N, 65o43'W) on southern Baffin Island. The record includes high-resolution measurements of microparticle number, mass, and grain-size distribution. This multiparameter approach may allow discrimination between time-varying physical factors such as dust supply or transport distance (Zielinski and Mershon, 1997). We discuss the significance of the Penny ice cap dust record in relation to the glacial and environmental history of the Canadian Arctic. Comparisons are made to the Greenland Ice Sheet Project 2 (GISP2) dust record in order to differentiate between regional and hemispheric influences on atmospheric dust deposition.

METHODS

The 334-m P95 ice core was drilled in April-May 1995 from a site along the main divide of the Penny ice cap (Fig.1). Radar soundings indicate that the drill penetrated to within a few centimeters of the bed. The core was measured for solid electrical conductivity and sampled continuously for d180, major ion chemistry, and microparticles. Stringently clean sampling techniques were used to prevent contamination (Grumet, 1997). All samples remained frozen prior to analysis. The concentration and sizes of microparticles between 0.8 and 12 Am diameter were measured on an Elzone 280PC multichannel particle counter (Zdanowicz et al., 1998). Major ions (Na, NH4, K, Mg, Ca, Cl, NO3, and S04) were measured on a Dionex 4000i ion chromatograph with a precision of 8% for K and <5% for all other species (Buck et al., 1992; Grumet et al., 1998). Oxygen isotopes were measured at the Department of Geophysics, University of Copenhagen (Fisher et al., 1998). A modem accumulation rate of 0.37 m/yr (ice equivalent) was calculated for the P95 site and the percentage of ice formed by refreezing of meltwater was estimated to have averaged 50% (mass) in the last 100 years (Grumet, 1997). Despite this substantial melt, residual glaciochernical and dust signals are preserved in the P95 core, providing paleoclimatic information at multiannual or lower (decadal to centennial) resolution (Grumet 1997; Zdanowicz et al., 1998).

The P95 core was dated back to 7900 yr ago (319 m) by spectral analysis of the electrical conductivity record, which tracks seasonal chemical variations (Fisher et al., 1998; Grumet et al., 1998). Additional time control was provided by conductivity and S04 peaks related to the Laki (A.D. 1783), Katmai (A.D. 1912), and unnamed A.D. 1259 and 50 B.C. volcanic eruptions. Age estimates for these peaks fell within ± 15% of expected values. Beyond 7900 yr ago, the depth-age curve was adjusted to match the end of the Younger DryasHolocene transition (326 m) dated at 11,550 ± 70 yr ago in Greenland ice core records (Johnsen et al., 1992; Alley et al., 1993). Comparison of the P95 core d18O record with that of an adjacent core (P96) drilled 16 km away (Fig.1) showed evidence of stratigraphic disturbance below 327 m (Fisher et al., 1998). Thus, although microparticle measurements were performed on the entire P95 core, we mostly discuss the undisturbed part of the dust record (Fig.2).

To maximize the paleoclimatic information available from the P95 dust record, we calculated the number (N) and mass (M) of microparticles in each sample. The Elzone counter measures the volume of dust particles and assigns them a diameter (d) by assuming they are spherical. M was calculated from the total volume, V, following M = pV and using the mean density of crustal material (p = 2.6 g/cm^3). To compensate for the decreasing sampling resolution with depth and to improve the statistical significance of our data, time series of N and M were averaged over 100-yr intervals from 11,550 yr ago to the present (Fig.3)and (Table1). For comparison, we analyzed the mean size distribution (by volume) of microparticles in the P95 and GISP2 cores for late-glacial and selected Holocene time intervals (Fig.4). Each distribution was produced by stacking data from individual samples. In most cases, the stacked distributions could be conveniently characterized by fitting them with log-normal curves of the form (Formula), where V is the total dust volume (the area under the curve), dv the mode, and ag the standard deviation. This distribution is commonly used to represent soil-derived aerosols or ice core microparticles (e.g., Patterson and Gillette, 1977; Steffensen, 1997). Parameters (dv ag,) defining the log-normal curves used in our comparisons are given in (Table2).

THE P95 ICE CORE DUST RECORD

The dust profile in the P95 core shows concentrations varying between 3000 and 40,000/ml (18 to 800 ug/kg) from the surface down to a depth of 326 m (Fig.2). Between 326 and 327 m, the number of microparticles (N) increases abruptly to 165,000/ml, concurrent with a -12.5% shift in the d180 profile. The dust mass (M) shows a comparatively small increase (250 ug/kg), implying that the high microparticle numbers between 326 and 327 m are due to relatively small (submicrometer) dust particles that contribute little mass. Based on similarities with other Canadian Arctic and Greenland ice cores, we interpret the 326- to 327-m portion of the P95 core to represent the transition from the last ice age (Wisconsin Glaciation) into the Holocene (Fisher et al., 1998). Microparticles in ice cores tend to collect at melting surfaces (Koerner, 1977; Steffensen, 1985). Hence, the dust peak between 326 and 327 m in the P95 core could represent an uncomformity between glacial-age ice and overlying Holocene ice. However, the resemblance between the P95 d18O profile and other, continuous Northern Hemisphere ice core records (e.g., Greenland) argues against a discontinuity in the P95 record (Johnsen et al., 1992; Fisher et al., 1998). High dust concentrations were found in late-glacial ice cores from both hemispheres and were attributed to the dry, windy late-glacial climate that led to an expansion of continental dust sources, increased eolian deflation, and reduced precipitation scavenging during long-range transport (Petit et al., 1981; Thompson and Mosley-Thompson, 1981; Hansson, 1994). Likewise, the P95 dust record suggests that the atmospheric dust load in the eastern Canadian Arctic was considerably greater in lateglacial time than at present and decreased rapidly following the transition into the Holocene. The finer modal size of dust deposited on the Penny ice cap during late-glacial time (dv approx.= 1um) compared to that in modem snow (dv approx.= 2 um) suggests that late-glacial dust originated from sources farther from the ice cap than at present (Fig.4). This was probably due to the extensive ice and snow cover in the Canadian Arctic at the time and also due to the fact that late-glacial ice in the P95 core originated from the Foxe Dome of the Laurentide Ice Sheet, 500 krn west of Cumberland Peninsula (Dyke and Prest, 1987; Fisher et al., 1998). Furthermore, a comparison of microparticles in the P95 and GISP2 ice cores reveals that more and larger dust particles were deposited at the GISP2 site during late-glacial time than on the Penny ice cap (Fig.4). These differences suggest that dust was brought to the two sites from different sources or along different trajectories. In support of this hypothesis, some climate simulations indicate that dust deposited in Greenland during the last glacial maximum was mostly from sources north of 47N, whereas dust carried over the Baffin Bay region originated from more distant sources at lower latitudes (Andersen et al., 1998; Mahowald et al., 1999).

The portion of the P95 ice core below 327 m is affected by ice flow discontinuities, and its interpretation is therefore problematic (Fisher et al., 1998). For example, the high microparticle concentrations between 333 and 334 m may not be of eolian origin, but rather sediment entrained from the icebedrock interface (Fig.2). Accordingly, we focus the remainder of our discussion on the Holocene portion of the P95 core.

The P95 Holocene dust record can be divided into three periods (Fig.3): an early Holocene interval from 11,550 to ca. 7500 yr ago characterized by consistently low dust concentrations (1); a transition period between 7500 and 5000 yr ago with gradually increasing dust levels (II); and a middle to late Holocene interval from ca. 5000 yr ago to the present (111) with markedly higher and more variable dust levels (Table 1). Dust deposited on the Penny ice cap between 10,500 and 7500 yr ago was characterized by a single log-normal mode near 2 to 3 um, similar to what is observed in modern ice and snow (Zdanowicz et al., 1998) (Fig.4) and (Table2). In contrast, dust deposited between 5500 and 2500 yr ago has a second, coarser mode around 6 to 8 um. The coarse dust mode probably accounts for the early to middle Holocene dust mass increase in the P95 record because the mass of dust increases geometrically with particle size (M[alpha]d^3). The modal size of finer particles is close to that of soil-derived aerosols transported over long distances (10^2_10^3 km), suggesting that these fine dusts may be, in part, from sources very distant from the Penny ice cap (Zdanowicz et al., 1998). The coarser dust mode is probably from sources relatively close to the ice cap, possibly within a radius of 100 km. The lack of a comparable coarse dust mode in the GISP2 core supports this inference (Fig.4)(Fig. 4).

To investigate how changing dust sources affected the P95 Holocene dust record we analyzed time-series of terrestrial Ca (Fig.3)(Fig. 3). Calcium is a common element in wind-blown dust, with concentrations ranging from 0.1 to >12% (weight) depending on the source (Pye, 1987). In Greenland ice the proportion of Ca in dust is typically < 10% and is primarily associated with silicate minerals (feldspars, micas, clay minerals) or carbonates (calcite, dolomite) (Laj et al., 1997; Maggi et al., 1997). To estimate the terrestrial fraction of Ca in the P95 core, we apportioned the total Ca into sea salt and non sea salt (nss) components using the procedure of O'Brien et al. (1995). Non sea salt sources, presumably crustal, account on average for 88% of Ca deposited on the Penny ice cap during the Holocene. NssCa in the P95 core declined markedly from late-glacial time to ca. 10,000 yr ago and remained essentially constant thereafter (Fig.3). These findings indicate that the coarse dust deposited on the Penny ice cap in the middle to late Holocene was not derived from calcium-rich sources like calcareous sediments. Instead, it may have been deflated from glacial sediments derived from crystalline rocks that comprise most of the substrate on southern Baffin Island (Dyke et al., 1982). Possible dust sources may include outwash trains and sandurs in the valleys around the ice cap or raised marine sediments on the Great Plain of the Koukdjuak, west of Cumberland Peninsula (Andrews, 1989).

Few highly resolved proxy records of eolian activity exist that may be compared with the P95 record. Koerner (1977) presented a profile of dust concentration in a 299-m core from the Devon ice cap, Canadian High Arctic. This profile shows no significant changes in dust fallout over the last 10,000 years (Koerner, 1977; (Fig.1)). However, only the number of particles >= 1 um was measured in the Devon core, making comparison with the P95 record of limited value. A more direct comparison was made between time-series of dust (N, M) and nssCa developed from the P95 and GISP2 cores using identical methods (Fig.3)(Fig. 3). Because it is from a high-latitude site (72N), the GISP2 microparticle record is assumed to reflect the atmospheric dust load on a broad, possibly hemispheric scale (Zielinski and Mershon, 1997). Microparticle data between 13303350 and 8060-9430 yr ago in the GISP2 core are inconsistent with the remainder of the record due to sampling problems and are not considered. Nevertheless, the GISP2 and P95 records differ remarkably. Although levels and trends of dust and nssCa in both cores are comparable in the early Holocene, they diverge markedly ca. 7000 to 6000 yr ago as dust in the P95 core increases and nssCa in the GISP2 core decreases. The divergence is more evident in the time-series of dust mass (M) and is therefore probably linked to the added coarse dust mode in the P95 core seen in (Fig.4). These differences suggest that regional-scale environmental changes in the middle to late Holocene affected the source and/or transport of dust and nssCa to the Penny ice cap. Possibly the most influential of these changes was recession of the Penny ice cap, which reduced the transport distance from dust sources to the P95 site. Ice core evidence suggests that the Penny ice cap began thinning substantially ca. 8000 yr ago as it gradually separated from the Foxe Dome of the Laurentide Ice Sheet (Fisher et al., 1998). This is inferred from the diverging d180 profiles in the P95 and P96 ice cores which reflect the onset of distinct regimes of snow accumulation and ablation at the two sites. Moreover, geological evidence indicates that the northern and southern margins of the Penny ice cap retreated to or behind their present positions ca. 7750 yr ago (7000 14C yr B.P.), although final separation from the Laurentide Ice Sheet probably postdated 5000 yr ago (4500 14C yr B.P.; Dyke et al., 1982). As the ice cap shrank, katabatic winds decreased, thereby increasing the probability of upslope dust transport to the P95 site. In addition, the deglaciation and postglacial emergence of southern Baffin Island expanded the area of terrain exposed to wind erosion, particularly west of Cumberland Peninsula from ca. 8000 to 6000 yr ago (7000-5000 14C yr B.P.; Andrews, 1989). Soil-derived aerosols produced by wind speeds far exceeding the threshold for erosion are commonly characterized by a volume distribution with modes of < 1, ~~ 2, and >10 Am (Patterson and Gillette, 1977; Gomes et al., 1990). If allowance is made for loss of coarse particles during transport, this size distribution is very similar to that of dust deposited at the P95 site between 5500 and 2500 yr ago. Furthennore, a thick succession of eolian sand began accumulating on Cumberland Peninsula southeast of the Penny ice cap ca. 5000 yr ago (4500 14C yr B.P.), suggesting drier and/or windier conditions at that time (Dyke et al., 1982). Polar desert soils are highly susceptible to wind erosion due to limited vegetation cover and frost action. Although limited, modem field observations indicate that eolian activity in Arctic polar deserts is greatest in the fall and winter, when the absence of surface water, combined with gusty winds (up to 40 m/s) and cold, abrasive blowing snow grains, facilitates deflation of soils and sediments (McKenna-Neuman, 1993). Indeed, microparticle size data from modem snow suggest that deposition of dust derived from proximal sources on the Penny ice cap occurs in late summer and fall, when ground snow cover is at a minimum (Zdanowicz et al., 1998). Hence, windier-thanpresent fall and winter conditions may have contributed to the increasing dust deposition at the P95 site in the middle Holocene. However, larger and more proximal wind-erodible terrain alone could account for the dust increase.

In addition to the regional factors already discussed, Holocene dust deposition on the Penny ice cap may have been affected by changes in the general atmospheric circulation and dust transport from midlatitude sources. Starting ca. 7000 yr ago, increasing amounts of arboreal (subarctic) pollen were deposited on ice caps of the Canadian High Arctic and in lakes of southwestern Greenland (Fredskild, 1984; McAndrews, 1984; Bourgeois, 1986). The increase was attributed to northward migration of the summer polar front following retreat of the Laurentide Ice Sheet, which allowed for more frequent penetration of southerly air carrying arboreal pollen into the Arctic (Fredskild, 1984; McAndrews, 1984; Bourgeois, 1986). The midcontinent regions of Eurasia and western Asia were as wet or wetter than present during the early and middle Holocene, and dust export from these regions was probably comparable to today's (Van Campo and Gasse, 1993; Harrison et al., 1996). However, sand dunes and lake sediments in the Great Plains and north-central United States attest to increased eolian activity between ca. 8500 and 5500 yr ago (8000-5000 4C yr B.P.), presumably caused by a strong westerly circulation that brought dry air over the midcontinent region (Bradbury et al., 1993; Forman et al., 1995). Hence, a greater midlatitude dust export from these regions combined with enhanced poleward transport may have contributed to the early to middle Holocene dust increase in the P95 record. However, correlation of the P95 dust profile with continental eolian records remains equivocal due to possible dating errors in the P95 early to middle Holocene record and also due to the fact that continental eolian records may be affected by local environmental controls, such as vegetation cover, that limit dust mobility (e.g., Forman et al., 1995). In light of available evidence, the P95 Holocene dust record can be explained solely by changes in the size of the Penny ice cap or the availability of regional dust sources.

CONCLUSIONS

The record of atmospheric dust deposition developed for the Penny ice cap provides valuable indications about environmental changes that affected the production and transport of windblown dust in the eastern Canadian Arctic during the lateglacial period and the Holocene. Maximum dust concentration occurs in the late-glacial portion of the P95 core, in agreement with worldwide ice core evidence of a dustier atmosphere than present at this time. The dust apparently came from sources more distant than those found in the GISP2, central Greenland, ice core. Different transport trajectories related to the lateglacial general circulation could account for the different modal size of dust in the P95 and GISP2 cores. Comparing the mineralogy and geochemistry of dust samples from the P95 and GISP2 cores should help clarify this issue in the near future.

The rate of dust deposition on the Penny ice cap decreased sharply in the early Holocene and remained low until ca. 7500 yr ago. There followed a gradual increase leading to a middle to late Holocene interval (ca. 5000 yr to present) of markedly higher and more variable dust deposition. This increase was primarily due to recession and thinning of the Penny ice cap after ca. 8000 yr ago, which reduced the distance from dust sources and facilitated upslope transport of dust to the P95 site. Other factors that may have contributed to this dust increase include the emergence of new sources during deglaciation of southern Baffin Island or a transition to drier or windier conditions in the middle and late Holocene, thereby increasing eolian deflation and transport of dust to the Penny ice cap. Mean dust and non sea salt calcium (nssCa) levels in the P95 core were stable in the late Holocene, but increased markedly in the GISP2 core as a result of intensified atmospheric circulation (O'Brien et al., 1995; Kreutz et al., 1997). The lack of a comparable increase in the P95 dust record is presumably due to expansion of the Penny ice cap and greater snow cover during the late Holocene, which increased the distance between dust sources and the P95 site and limited eolian deflation and dust transport at that time.

The P95 ice core record provides valuable information on changes in atmospheric dust loading, windiness, and aerosol transport paths in the eastern Canadian Arctic during the late Pleistocene and Holocene. It also provides supportive evidence for the postglacial history of the Penny ice cap and northeastern Laurentide Ice Sheet, as postulated from geological and ice core d18O data (Dyke et al., 1982; Fisher et al., 1998). Comparison between dust and non sea salt calcium (nssCa) trends in the P95 and GISP2 cores allow regional influences to be distinguished from those that may be hemispheric in character. The divergence between dust and nssCa trends in the P95 and GISP2 cores after ca. 6000 yr ago supports the view that the climate became increasingly affected by regional environmental conditions (e.g., snow and ice cover) during the middle to late Holocene (O'Brien et al., 1995). Findings from this study further demonstrate how valuable records of regionalscale paleoenvironmental changes can be developed from small circum-Arctic ice caps. Development of additional, comparable records will contribute to a better understanding of the variability of paleoclimate in the Arctic, which is needed to assess the sensitivity of high-latitude regions to future climatic change.

ACKNOWLEDGMENTS

Support in the field was provided by the Polar Continental Shelf Project, Parks Canada, and the communities of Iqaluit and Pangnirtung. Ice core drilling was conducted with assistance from Erik Blake and Mike Gerasimoff of Icefield Instruments Inc., Whitehorse, Yukon. Michelle Day and Mike Leo helped with microparticle measurements, and Sallie Whitlow and Nancy Grumet performed the major ion analyses. The manuscript benefited from reviews by Arthur Dyke, Lonnie Thompson, and an anonymous referee. This research was supported by the Office of Polar Programs, National Science Foundation. It is a contribution to the Ice-Core Circum-Arctic Paleoclimate Program.

REFERENCES

Alley, R. B., Meese, D. A., Shuman, C. A., Gow, A. J., Taylor, K. C., Grootes, P. M., White, J. W. C., Ram, M., Waddington, E. D., Mayewski, P. A., and Zielinski, G. A. (1993). Abrupt increase in Greenland snow accumulation at the end of the Younger Dryas event. Nature 362, 527-529.

Andersen, K. K., Armengaud, A., and Genthon, C. (1998). Atmospheric dust under glacial and interglacial conditions. Geophysical Research Letters 25, 2281-2284.

Andrews, J. T. (1989). Quaternary geology of the northeastern Canadian Shield. In "Quaternary Geology of Canada and Greenland" (R. J. Fulton, Ed.), pp. 277-317. Geological Survey of Canada, Geology of Canada No. 1, Ottawa.

Andrews, J. T., Davis, P. T., and Wright, C. (1976). Little Ice Age permanent snow cover in the eastern Canadian Arctic: Extent mapped from Landsat-I satellite imagery. Geografiska Annaler 58, 7 1- 8 1.

Bourgeois, J. C. (1986). A pollen record from the Agassiz Ice Cap, northern Ellesmere Island, Canada. Boreas 15, 345-354.

Bradbury, J. P., Dean, W. E., and Anderson, R. Y. (1993). Holocene climatic history of the north-central United States as recorded in the varved sediments of Elk Lake, Minnesota: A synthesis. In "Elk Lake, Minnesota: Evidence for Rapid Climate Change in the North-Central United States" (J. P. Bradbury and W. E, Dean, Eds.), pp. 309-328. Geological Society of America Special Paper 276, Boulder, CO.

Bradley, R. S. (1990). Holocene paleoclimatology of the Queen Elizabeth Islands, Canadian High Arctic. Quaternary Science Reviews 9, 365-384.

Buck, C. F., Mayewski, P. A., Spencer, M. J., Whitlow, S. L, Twickler, M. S., and Barrett, D. (1992). Determination of major ions in snow and ice cores by ion chromatography. Journal of Chromatography 594, 225-228.

Davis, P. T. (1985). Neoglcial moraines on Baffin Island. In "Quaternary Environments: Eastern Canadian Arctic, Baffin Bay and Western GreenIand" (J. T. Andrews, Ed.), pp. 682-718. Allen and Unwin, Boston.

Dyke, A. S., and Prest, V. K. (1987). Late Wisconsinan and Holocene history of the Laurentide Ice Sheet. Giographie Physique et Quatemaire 41, 237-263.

Dyke, A. S., Andrews, J. T., and Miller, G. H. (1982). "Quaternary Geology of Cumberland Peninsula, Baffin Island, District of Franklin." Memoir 403, Geological Survey of Canada, Ottawa, 32 pp.

Fisher, D. A., Koerner, R. M., Bourgeois, J. C., Zielinski, G. A., Wake, C. P., Hammer, C. U., Clausen, H. B., Gundestrup, N., Johnsen, S., Goto-Azuma, K., Hondoh, T., Blake, E., and Gerasimoff, M. (1998). Penny ice cap cores, Baffin Island, Canada, and the Wisconsinan Foxe Dome connection: Two states of Hudson Bay ice cover. Science 279, 692-695.

Forman, S. L., Oglesby, R., Markgraf, V., and Stafford, T. (1995). Paleoclimatic significance of Late Quaternary eolian deposition on the Piedmont and High Plains, central United States. Global and Planetary Change 11, 35-55.

Fredskild, B. (1984). Holocene palaeo-winds and climatic changes in west Greenland as indicated by long-distance transported and local pollen in lake sediments. In "Climatic Changes on a Yearly to Millennial Basis" (N. A. M6mer and W. Karl6n, Eds.), pp. 161-171. Reidel, Dordrecht.

Fujii, Y., Kamiyama, K., Kawamura, T., Kameda, T., Izumi, K., Satow, K., Enomoto, H., Nakamura, T., Hagen, J. 0., Gjessing, Y., and Watanabe, 0. (1990). 6000-year climate records in an ice core from Hoghetta ice dome in northern Spitsbergen. Annals of Glaciology 14, 85-89.

Gomes, L., Bergametti, G., Coud.6-Gaussen, G., and Rognon, P. (1990). Submicron desert dusts: A sandblasting process. Journal of Geophysical Research 95, 13,927-13,935.

Grumet, N. S. (1997). "Glaciochemical Investigations of the Last 1000 Years from the Penny Ice Cap, Baffin Island. M.Sc. thesis, Univ. of New Hampshire, Durham, 90 pp.

Grumet, N. S., Wake, C. P., Zielinski, G. A., Fisher, D. A., and Jacobs, J. D.(1998). Preservation of glaciochemical time-series in snow and ice records from the percolation zone of the Penny Ice Cap, Baffin Island. Geophysical Research Letters 25, 357-360.

Hansson, M. E. (1994). The Renland ice core: A northern hemisphere record of aerosol composition over 120,000 years. Tellus 46B, 390-418.

Harrison, S. P., Yu, G., and Tarasov, P. E. (1996). Late Quaternary lake-level record from northern Eurasia. Quaternary Research 45, 138-159.

Johnsen, S. J., Clausen, H. B., Dansgaard, W., Furher, K., Gundestrup, N., Hammer, C. U., Iversen, P., Jouzel, J., Stauffer, B., and Steffensen, J. P. (1992). Irregular glacial interstadials recorded in a new Greenland ice core. Nature 359, 311-313.

Koerner, R. M. (1977). Distribution of microparticles in a 299-m core through the Devon Island ice cap, Northwest Territories, Canada. In "Isotopes and Impurities in Snow and Ice," pp. 371-376. International Association of Hydrological Sciences, Grenoble.

Koerner, R. M. (1980). The problem of lichen-free zones in Arctic Canada. Arctic and Alpine Research 12, 87-94.

Koerner, R. M., and Fisher, D. A. (1990). A record of Holocene summer climate from a Canadian high-Arctic ice core. Nature 343, 630-631.

Kotlyakov, V. M., Nikolayev, V. M., Korotkov, 1. M., and KJementyev, 0. L. (1991). Climate-strat~igraphy of Severnaya Zemlya ice dome in the Holocene (in russian). In "Stratigraphy and Correlation of Quaternary Deposits of East Asia and the Pacific Region," pp. 100 -112. Nauka, Moscow.

Kreutz, K. J., Mayewski, P. A., Meeker, L. D., Twickler, M. S., Whitlow, S. I., and Pittalawala, 1. 1. (1997). Bipolar changes in atmospheric circulatior during the Little Ice Age. Science 277, 1294-1296.

LaJ, P., Ghermandi, G., Cecchi, R., Maggi, W., Riontino, C., Hong, S., Candelone, J.-P., and Boution, C. (1997). Distribution of Ca, Fe, K and S between soluble and insoluble material in the Greenland Ice Core Project ice core. Journal of Geophysical Research 102, 26,625-26,623.

Locke, C. W., and Locke, W. W. (1977). Little Ice Age snow-cover extent and paleoglaciation thresholds: North-central Baffin Island, N.W.T., Canada. Arctic and Alpine Research 9, 291-300.

Maggi, V. (1997). Mineralogy of atmospheric microparticles deposited along the Greenland Ice Core Project ice core. J. Geophys. Res. 102, 26,72526,734.

Mahowald, N., Kohfeld, K., Hansson, M., Balkanski, Y., Harrison, S. P., Prentice, 1. C., Shiilz, M., and Rodhe, H. (1999). Dust sources and deposition during the last glacial maximum and current climate: A comparison of model results with paleodata from ice cores and marine sediments. Journal of Geophysical Research 104, 15,895-15,916.

McAndrews, J. H. (1984). Pollen analysis of the 1973 ice core from Devon Island glacier, Canada. Quaternary Research 22, 68-76.

McKenna-Neuman, C. (1993). A review of aeolian transport in cold environments. Progress in Physical Geography 17, 137-155.

O'Brien, S. R., Mayewski, P. A., Meeker, L. D., Meese, D. A., Twickler. M. S., and Whitlow, S. 1. (1995). Complexity of Holocene climate as reconstructed from a Greenland ice core. Science 270, 1962-1964.

Overpeck, J., Hughen, K., Hardy, D., Bradley, R., Case, R., Douglas, M.. Finney, B., GaJewski, K., Jacoby, G., Jennings, A., Lamoureux, S., Lasca. A., MacDonald, G., Moore, J., Retelle, M., Smith, S., Wolfe, A., and

Zielinski, G. (1997). Arctic environmental change of the last four centuries. Science 278, 1251-1256.

Patterson, E. M., and Gillette, D. A. (1977). Commonalties in measured size distributions for aerosols having a soil-derived component. Journal of Geophysical Research 82, 2074-2082.

Petit, J. R., Briat, M., and Royer, A. (1981). Ice age aerosol from East Antarctic ice core samples and past wind strength. Nature 293, 391-394.

Pye, K. (1987) "Aeolian Dust and Dust Deposits." Academic Press, San Diego, 334 pp.

Steffensen, J. P. (1985). Microparticles in snow from the Greenland ice sheet. Tellus 37B, 286-295.

Steffensen, J. P. (1997). The size distribution of microparticles from selected segments of the Greenland Ice Core Project ice core representing different time periods. Journal of Geophysical Research 102, 26,755-26,763.

Thompson, L. G., and Mosley-Thompson, E. (1981). Microparticle variations linked with climatic change: evidence from polar ice cores. Science 212, 812-815.

Van Campo, E., and Gasse, F. (1993). Pollen- and diatom-inferred climatic and hydrological changes in Sumxi Co basin (western Tibet) since 13,000 yr B.P. Quaternary Research 39, 300-313.

Zdanowicz, C. M., Zielinski, G. A., and Wake, C. P. (1998). Characteristics of modem atmospheric dust deposition in snow on the Penny ice cap, Baffin Island, Arctic Canada. Tellus 50B, 506-520.

Zielinski, G. A., and Mershon, G. R. (1997). Paleoenvironmental implications of the insoluble microparticle record in the GISP2 (Greenland) ice core during the rapidly changing climate of the Pleistocene-Holocene transition. Geological Society of Arnerica Bulletin 109, 547-559.

'To whom correspondence should be addressed at Glaciology, Terrain Sciences Division, Geological Survey of Canada, 601 Booth Street, Ottawa, Ontario KIA OE8 Canada.