This material is based upon work supported by the National Science Foundation under Cooperative Agreement Number OPP-9617009 to the Climate Change Research Center, Institute for the Study of Earth, Oceans, and Space, University of New Hampshire. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
In 1986 an ad hoc committee of the Polar Research Board issued a series of recommendations intended to stimulate U.S. ice coring activities (National Research Council, 1986) in response to the growing need for high quality paleoclimate records that could be used to understand climate change. Their report reviewed the then current status of ice core research in the U.S. and the elements necessary to implement a strong program of ice core collection and analysis. Now, in 1998, the vision set forth for a strong U.S. ice coring program has been achieved. The U.S. ice coring community has, in concert with its international partners, made remarkable strides to advance in our understanding of global change. This new document, assembled by the Ice Core Working Group (ICWG), provides a synthesis of the global change lessons learned thus far, and the requirements and plans for solving new global change questions utilizing future ice coring activities.
The ICWG is sponsored by the Office of Polar Programs, National Science Foundation, in response to a recommendation made by the Polar Research Board (National Research Council, 1986). It provides a forum for the discussion of issues related to, and future directions for, the U.S. ice coring community. Previous planning documents developed by this group (e.g., ICWG 1987, 1988, 1989) have led to the initiation of major deep drilling efforts (GISP2, Siple Dome) for the U.S. community. The current members of the ICWG* and authors of this document are:
Paul A. Mayewski* (University of New Hampshire), Chair
John Bolzan (Ohio State University)
Debra Meese* (Cold Regions Research and Engineering Laboratory)
Todd Sowers* (Pennsylvania State University)
Mark Twickler (University of New Hampshire)
Gregory Zielinski* (University of New Hampshire)
Roger Bales* (University of Arizona)
Gordon Hamilton* (Ohio State University)
David Morse* (University of Texas at Austin)
Michael Ram* (State University of New York at Buffalo)
Eric Steig* (University of Colorado)
Cameron Wake* (University of New Hampshire)
Since ancient times humans have modified their environment, but only since the beginning of the Industrial Revolution has human activity had a dramatic effect at global scales. Without question, human activities have affected the composition of the global atmosphere, and the magnitude of human disturbance to biogeochemical cycles may now be approaching a critical level. Over the past decades concentrations of atmospheric carbon dioxide (CO2) and methane (CH4) have moved into a range unprecedented during the last few million years. This increase has produced serious concern regarding the heat balance of the entire planet. Greenhouse gases are, however, only part of the human-induced problem. Sulfur aerosols, ozone and dusts, to mention only a few, can either reinforce or counteract greenhouse gas effects on local to regional scales.
While intense efforts are underway to determine the history and significance of human influences on climate, including atmospheric chemistry, and on resource depletion, our understanding of climate change is still hampered by lack of knowledge of the natural controls on climate. Mounting evidence points to the significance of a variety of natural climate forcing agents such as solar variability, planetary orbital cycles, volcanic activity, changes in size and flow of large ice sheets, and changes in thermohaline circulation of the ocean. From records of climate variability in the past we know that these controls have produced significant changes in climate that have, at times, occurred very rapidly (over a few years).
Our ability to understand climate change, to decipher the influence of human activity, and to predict future climate depends on a coupled investigation of both modern and past climate. The modern era of climate is probably the most difficult to understand because of the combined influences of natural and human-induced activities; hence climate understanding and prediction now poses an immense challenge to science.
Records from a variety of sources (e.g., instrumental records, historical documents, deep-sea and continental sediments, tree rings, and ice cores) provide the basic boundary conditions (e.g., sea surface temperature, precipitation and atmospheric circulation patterns) necessary for robust environmental reconstructions. However, ice core records provide the most direct, detailed, and complete measure of past climate change; as a consequence, considerable national and international attention has been paid to the unique perspective they provide. Much of this attention has been focused on the deep ice cores recovered from Summit, Greenland and from Vostok in East Antarctica (Figure 1).
Ice core records provide detailed descriptions of climate change that are extremely valuable for comparison with modern observations. Further, they document not only a wide range of environmental parameters that are both measures of and responses to climate change (e.g., atmospheric chemistry and circulation, temperature, precipitation) but also many of the causes of climate change (e.g., solar variability, volcanic activity, greenhouse gases). Because of their high resolution (sub-annual), long time span (several glacial cycles) and precise dating (annual), they also provide a framework for interpreting other records of past climate.
While the following overview is by no means an exhaustive compilation of the lessons learned from ice core research over the last decade, it is intended to highlight many of the major scientific accomplishments of this period.
The Greenhouse Gas/Temperature Relationship
The climate records from the Vostok ice core are the longest continuous record of Antarctic climate available to date. Geochemical measurements have been conducted to a depth of 3350 meters, and the results indicate that the age of the ice at this level is approximately 420,000 years, covering more than three full glacial/interglacial cycles (Petit et al., 1997). Data from the upper 2100 meters (corresponding to ~160,000 years) are included in Figure 2. Of prime importance is the temperature record, which has been reconstructed from data on the isotopic composition of paleoprecipitation (Lorius et al., 1985; Jouzel et al., 1987). The last interglacial period (centered at 130,000 years ago) had characteristic temperatures that were slightly higher than today. Beginning about 130,000 years ago, Antarctic temperatures dropped in a step-wise fashion until they reached full glacial values of at least -6oC, relative to today, between 25,000 and 18,000 years ago. The last glacial/interglacial transition began around 18,000 years ago and ended around 10,000 years ago with a slight cooling trend apparent throughout the last 10,000 years.
The most oft-cited finding from the Vostok ice core is the close correspondence between temperature and concentrations of two radiatively active gases, viz., CO2 and CH4. During glacial periods, atmospheric CO2 and CH4 levels were 30% and 50% lower respectively than interglacial levels, (Barnola et al., 1987; Chappellaz et al., 1990). Because CO2 and CH4 absorb a portion of the outgoing long wavelength radiation and re-emit it back to the Earth's surface (a process better known as the greenhouse effect), higher concentrations of these gases in the atmosphere will tend to raise global surface temperatures (with all other factors being equal). The fact that atmospheric CO2 and CH4 levels were lower during the glacial period helps explain the lower temperatures and increased continental ice volume during these periods. In a study employing a general circulation model to test the sensitivity of global climate to CO2 and CH4 variations, Lorius et al. (1990) estimated that as much as 50% of the temperature variations over the last full climate cycle could be related to observed CO2 and CH4 variations. The strong degree of covariation between the concentration of greenhouse gases and climate throughout the last 160,000 years has been used to predict future climate changes related to the buildup of CO2 and CH4 in the atmosphere as a result of anthropogenic activities.
Rapid Climate Change Events
One of the most dramatic recent contributions to our understanding of paleoclimate during the last glacial cycle has come in the millennial scale range of climate variability. Unprecedented swings in the Earth's climate over sub-decadal to millenial time scales have now been recorded in two ice cores from central Greenland, instigating new, higher resolution investigations of land and marine paleoclimate records.
In 1993 the Greenland Ice Sheet Project Two (GISP2) successfully completed drilling to the base of the Greenland Ice Sheet in central Greenland. In so doing, GISP2, along with its European companion project GRIP (Greenland Ice Core Program), developed the longest (110,000 years ) high resolution paleoenvironmental record available from the northern hemisphere. Based on the comparison of electrical conductivity and oxygen isotope series between the two cores (Grootes et al., 1993; Taylor et al., 1993), at least the upper 90% of these cores display extremely similar, if not absolutely equivalent records. The current best estimate of the age at this depth (~2800 meters) is ~110,000 years, based on a combination of multi-parameter annual layer counting (Alley et al., 1993; Meese et al., 1994a, 1997) and measurements of the oxygen isotope ratios (d18O) of atmospheric O2 calibrated with the Vostok ice core (Bender et al., 1994). Maximum error estimates in the dating are quite remarkable: 1% for the last 3,500 years, 2% to 40,000 years , 10% to 57,000 years and 20% at 110,000 years (Alley et al., 1993; Sowers, et al, 1993; Meese et al., 1994a, 1994b, 1997). Agreement between the GISP2 and GRIP ice cores (separated by 30 km or ~10 ice thicknesses) over the record period of the last ~110,000 years provides strong support for the climatic origin of even the minor features of these records and implies that investigations of subtle environmental signals can be rigorously pursued.
The millennial scale events recorded in the upper 110,000 years of the two central Greenland ice cores are unequivocally climate events. They represent large climate deviations (massive reorganizations of the ocean-atmosphere system) that occur over decades or less, during which ice-age temperatures in central Greenland may have been slightly more than 20oC colder than today (Cuffey et al., 1995; Figure 3). These climate deviations seem to have been largest during the cooling to the glacial maximum and the warming from the glacial maximum, with smaller amplitudes during the glacial maximum as well as during the interglacial maximum.
The GISP2 record of insoluble dust, which complements the dissolved calcium record, provides another insight into the abruptly variable global atmosphere and climate. It has long been known that the atmosphere was much dustier during cold periods than warm periods but the exact cause is not certain. Recent results from GISP2 using mineralogical and isotopic tracer characteristics (Biscaye et al., 1997) and continuous dust content by laser-light scattering (Ram and Koenig, 1997) have provided new insights into how source area conditions and dust production affects atmospheric turbidity. Using mineralogical and isotopic tracer characteristics, it was determined that dust from just prior to the last glacial maximum originated from the deserts of eastern Asia. Since these tracer characteristics were virtually constant throughout this period, the source areas could not have changed substantially in location or size, and the observed dust variations must have been due to changes in the intensity of atmospheric circulation (Biscaye et al., 1997). Examination of the GISP2 dust profile measured by laser-light scattering has revealed that the dust concentration in the core is modulated by an 11-year period back to 100,000 years before present. These modulations are believed to be a result of solar energy flux changes associated with the 11-year solar cycle which affects the aridity of dust source regions. The effect is very strong, particularly during the Wisconsinan, where fivefold changes in dust concentration occur over a single 11-year cycle (Ram et al., 1997; Figure 4).
Examination of a subsequent abrupt event, the Younger Dryas (a return to near-glacial conditions during the last deglaciation, previously identified in a variety of paleoclimate records), demonstrates the importance of conducting multi-parameter, high resolution paleoclimate investigations on well-dated records. During this event, lowered temperatures (15oC +/- 3oC relative to today; Severinghaus et al., 1998) were accompanied by a 50% drop in snow accumulation (Alley et al., 1993), order-of-magnitude increases in the amount of wind-blown dust and sea-salt in the atmosphere (Mayewski et al., 1993a; Zielinski and Mershon, 1997) and large changes in methane concentration (Brook et al., 1996), synonymous with cold, dry, and dusty conditions (Figure 5). Annually resolved sampling over early and late stages of the Younger Dryas indicates that this ~1300-year-duration event had an onset and termination each accomplished in 5-20 years (Alley et al., 1993, Mayewski et al., 1993a, Taylor et al., 1993, 1997; Figure 6).
The identification of rapid climate change events in the GRIP CH4 record (Chappellaz et al., 1993; Brook et al., 1996; Figure 7) prompted considerable interest in the identification of such events in other regions since the source areas for CH4 during the last glaciation may have been in the middle to lower latitudes. Recent work in correlating ice cores from Greenland and Antarctica has shown that the millennial scale "Dansgaard/Oeschger" events, which are so prominent in the North Atlantic, are also found in Antarctica when the event lasts more than 2,000 years (Bender et al., 1994; Figure 8). Bipolar similarities in climate events are now known to extend to events less than 2000 years in duration (Mayewski et al., 1996) based on comparison between GISP2 and Taylor Dome ice cores (Figure 1). Unfortunately, it has not yet been possible to unambiguously establish the precise temporal relationship between the Greenland and Antarctic climate variations. The ambiguity arises partly from the uncertainty in the age difference between the solid ice and the gas trapped within it.
While the temporal relationship between the Antarctic and Greenland ice core climate records has yet to be established for rapid climate change events, timing of events leading to the last deglaciation has been suggested. The initial phase of the deglaciation in Antarctica is believed to begin with a warming at around 18,000 years ago (Sowers and Bender, 1995) that is coincident with a decrease in the flux of dust to Antarctica (Jouzel et al., 1992), increases in atmospheric CO2 and CH4 concentrations, and the initial rise in sea level as deduced from the U/Th-dated Barbados coral record (Bard et al., 1990). In the North Atlantic, deglaciation has been linked to the northward migration of the polar front and the consequent "turn-on" of deep water formation by 14,700 years ago. These results suggest that the North Atlantic warming may not have responded to the initial forcing that caused the southern hemisphere to warm, ice sheets to begin melting, and accumulation of CO2 and CH4 to increase in the atmosphere.
Ice core records provide a framework for the interpretations of records recovered from other natural archives. Paleoclimate records from North Atlantic marine sediment cores also contain notable millennial scale variability (e.g., Heinrich, 1988; Broecker et al., 1992; Bond et al., 1992; Andrews and Tedesco, 1992; Lehman and Keigwin, 1992), although the exact timing of these events is known less precisely than for the Greenland ice cores. Several of the marine cores reveal evidence that the formation of NADW (North Atlantic Deep Water-warm, saline, nutrient-depleted deep return flow water), and thus the oceanic thermohaline circulation, fluctuated dramatically in the past (e.g., Ruddiman and McIntyre, 1981; Boyle and Keigwin, 1987). NADW diminished greatly during the last glaciation and was relatively strong during the interglacials. Recent studies confirm that NADW fluctuates on millennial time scales and correlates with sea surface and atmospheric temperature (e.g., Oppo and Lehman, 1995).
Changes in the flux of ice-rafted detritus, d18O of foraminifera shells, and the abundance of climate-sensitive foraminifera, as recorded in deep sea sediments, indicate that during the last glaciation the North Atlantic was punctuated by iceberg discharge events, termed "Heinrich events" (Heinrich, 1988), potentially produced in response to changes in ice sheet dynamics (MacAyeal, 1993). The larger of these Heinrich events has a characteristic recurrence in the marine record of 5000-10,000 years. The Dansgaard/Oeschger rapid climate change events, which were first idenified in ice cores, are also observed in marine record (Bond et al., 1993; Bond and Lotti, 1995; Cortijo et al., 1995).
Evidence for the presence of millennial scale climate fluctuations has been extended beyond the North Atlantic and polar regions. Marine cores from the Santa Barbara Basin reveal that perturbations in the ocean circulation patterns of the East Pacific region (Kennett and Ingram, 1995; Kotilainen and Shackleton, 1995; Behl and Kennett, 1996;) correlate with ice-rafted debris events in the North Pacific and with the Greenland ice core records. Abrupt changes in atmospheric circulation patterns and precipitation regime are recorded over eastern Asia in a thick sequence of wind-deposited loess from central China (Porter and An, 1995). Records of alpine glacier fluctuations, mountain snowlines and paleo-vegetation in the Andes reveal climate fluctuations that are similar to events in the Greenland ice cores (Lowell et al., 1995).
New advances in paleoclimate reconstruction also come from the tropics. For example, a 30,000-year-long paleotemperature record from lowland Brazil suggests a cooling of as much as 5oC (Stute et al., 1995), which contrast with earlier estimates from marine cores which limit cooling to < 3oC (CLIMAP Members, 1981). Further, an ice core from the Andes suggests reduced water vapor content during the Younger Dryas (Thompson et al., 1995). These new findings have stimulated examination of other tropical paleoclimate records and renewed investigations into climate forcing related to the hydrological cycle that is tied to changes in the tropics.
While the exact timing and causal mechanisms for glacial age climate fluctuations are not fully understood, important and new guidance is provided by high resolution, well-dated, continuous records like GISP2. For example, evidence of regularity in the timing of some climate events is building. Studies ranging from the North Atlantic (GISP2) to the subtropics demonstrate 1450-to 1800-year periodicities for rapid climate change events (Cortijo et al., 1995; Sirocko et al., 1996; Mayewski et al., 1997). In addition, the cumulative effect of multiple climate forcings can now be demonstrated. As an example, ~90% of the variance in the GISP2 paleoatmospheric circulation series is produced from a combination of the following: insolation changes induced by earth's orbital cycles, ice sheet dynamics, thermohaline ocean circulation and solar variability (Mayewski et al., 1997; Figure 9).
Natural Climate Variability During the Holocene
Annually resolved, continuous paleoclimate records from the GISP2 ice core demonstrate that Holocene climate is characterized by annual to millennial scale variability and that Holocene climate is significantly more complex than glacial age climate (e.g., Meese et al., 1994b; O'Brien et al., 1995). Time series of major ion concentrations in the ice, used as tracers for major atmospheric circulation systems (Mayewski et al., 1994, 1997) reveal a strong association between expansions of northern hemisphere polar atmospheric circulation systems and a variety of discontinuous paleoclimate records (Denton and Karlen, 1973; Harvey, 1980) that record worldwide coolings (O'Brien et al., 1995; Figure 10).
Complexities in Holocene climate are noted in a comparison of several environmental parameters. For example, a major reorganization in the climate system ~7800-8400 years ago has recently been documented using a variety of northern hemisphere paleoclimate records plus the GISP2 ice core (Alley et al., 1997) and in a comparison of lake sediment records from tropical Africa with ice core records from Greenland and Antarctica (Stager and Mayewski, 1997). Based on the GISP2 record the climate system response during this major event operated similarly to pre-Holocene cooling events (Figure 4). Namely, cooler temperatures (more negative d18O), reduced CH4, reduced accumulation rate and intensification of polar atmospheric circulation (expanded PCI (polar circulation index)) vary together. However, the coherence between these variables weakens as the events get younger, and is particularly poor during periods between these events, suggesting increased regionalization of climate from early to late Holocene. This progressive regionalization of climate may be the manifestation of the varying influences of a variety of climate forcings such as changes in total and season-to-season insolation, ice sheet and sea ice extent, solar variability and volcanism (O'Brien et al., 1995).
Major reorganizations in Holocene climate as well as finer scale climate fluctuations may be explained by a combination of climate forcings. A variety of paleorecords are available from ice core studies for testing the impact of these forcing mechanisms, including, for example, 10Be time-series (Beer, 1990; Steig et al., 1996; Finkel and Nishiizumi, 1997); CO2 time-series (Barnola et al., 1987; Wahlen et al., 1991; Etheridge et al., 1996); CH4 time-series (Chappellaz et al., 1993; Blunier et al., 1995; Brook et al., 1996) and volcanic sulfate time-series (Zielinski et al., 1994, 1996a). Particularly exciting results from high-resolution ice cores include the observation that many geochemical parameters show strong spectral power at frequencies close to or identical to those observed in the sun. Worldwide coolings during the Holocene have a quasi-periodicity of 2600 years in phase with previously defined ~2500 year variations in d14C (e.g., Denton and Karlen, 1973; Stuiver and Braziunas, 1989, 1993; O'Brien et al., 1995). Also, d18O in the GISP2 core is coherent with both the ice-core 10Be time-series and with the tree-ring record of atmospheric 14C (Stuiver et al, 1995; Figure 11). Remarkably, the series are coherent not only in phase but also in amplitude, providing what is probably the best evidence to date for the elusive sun-climate relationship, a subject of debate for more than a century.
The detection of the solar cycle in dust concentrations and geochemical data may not only provide information about the sun-climate relationship, but can be used to improve ice core chronologies. It is possible to detect the 11-year cycle in 10Be at some low-accumulation-rate sites where annual stratigraphy is not preserved (Steig et al., 1998). Measurement of the "11-year" layer thickness can be used as an independent check on flow-model estimates of layer thickness and to estimate past accumulation rates, or to provide a direct, layer-counted chronology.
Extreme Events-Volcanic Events and Biomass Burning
Perhaps no single naturally occurring phenomenon better illustrates the dynamic nature of the Earth and its individual systems than does an explosive volcanic eruption. Ice core records provide the best means available to the scientific community for determining the potential atmospheric impact of explosive volcanic activity with several key findings of global significance stemming from these studies. For example, volcanic events recorded in Greenland ice core sulfate series correlate with annual changes in atmospheric temperature, providing evidence for sulfate aerosol shielding (Lyons et al., 1990; Stuiver et al., 1995; White et al., 1997; Figure 12) and providing the most reliable means for evaluating past variability within the volcanism-climate system. Estimates of the stratospheric mass loading of an eruption with the subsequent calculation of the stratospheric optical depth (Zielinski, 1995), as derived from volcanic sulfate flux values, are the important parameters needed to hindcast and forecast the climatic impact of an eruption. These parameters are particularly useful to climate modelers.
Ice core records from both polar regions have clearly shown that the number of eruptions capable of perturbing climate over the last 110,000 years is far higher than previously thought as based on geological and historical records (e.g., Langway et al., 1995; Zielinski et al., 1996a; Figure 13). Ice core records have shown that in some cases large historical eruptions have been poorly studied and that these eruptions undoubtedly had a major impact on the atmosphere (e.g., Dai et al., 1991; Delmas et al., 1992; Zielinski, 1995). Consequently, future activity from these same volcanoes should be monitored closely. Findings from the GISP2 ice core suggest that the Toba eruption (largest eruption of the last 500,000 years, occurring about 71,000 years ago) may have been a driving force leading to several centuries of cold climatic conditions. Should such an eruption occur today it would have a tremendous impact on human populations (Zielinski et al., 1996b). Interestingly, the results from the GISP2 ice core also yielded information that supports a converse hypothesis, namely, that deglaciation can actually lead to increased volcanic activity, particularly in the northern hemisphere (Zielinski et al., 1997).
An additional phenomenon that has an impact on the chemistry of the atmosphere and biogeochemical cycles as a whole is widespread biomass burning. As for the volcanic records in ice cores, it has recently been determined that changes in chemical species such as ammonium, potassium and nitrate (NO3) recorded in ice cores characterize the deposition of chemical compounds associated with plumes from biomass burning events (e.g., Legrand et al., 1992; Dibb et al., 1996). As of now, the available records reflect variability in biomass burning events in the northern hemisphere and particularly in subarctic regions of Canada and possibly into Alaska (Whitlow et al., 1994; Taylor et al., 1996). Establishing the past record of these burn events enables the scientific community to evaluate the relationship between hot and dry climatic conditions and the frequency of large fires (such as the Yellowstone fires of 1988). Burn records from several northern hemisphere ice cores have shown the impact of fire management in North America over the last few decades (Whitlow et al., 1994).
Recent Climate Change (last 2000 years) as a Precursor to Modern Climate
Although the exact timing and geographic distribution of Holocene climate change events is complex, the last 1000-2000 years offer important opportunities for unraveling the climate variability, on centennial scales and finer, that influences modern climate. There is general agreement that glaciers around the world and, notably, Arctic sea ice expanded during at least parts of the 13th to 19th centuries (Grove, 1988), a period called the Little Ice Age (LIA), and that warming occurred for several centuries prior to this period, at least in some regions, during what is controversially called the Medieval Warm Period (MWP). Previous research summarized by Lamb (1995) demonstrates changes in climate such as increased severity of winter storms and sea ice extent plus accompanying changes in food harvests during the LIA and contrasting milder conditions during the MWP.
The LIA event appears to play an important role in our understanding of modern climate. Based on the GISP2 paleo-atmospheric circulation record (Figure 10) the LIA had the most abrupt onset (AD 1400-1430) of any of the Holocene rapid climate change events (O'Brien et al., 1995). This extends previous findings from a 1500-year-long ice core record in the Andes (Thompson and Mosley-Thompson, 1987) and a 1300-year-long ice core from central Greenland (Mayewski et al., 1993b) both of which suggested that entrance into and out of the LIA was abrupt.
A bipolar comparison of annually resolved ice cores from Greenland (GISP2) and West Antarctica (Siple Dome, Figure 1) demonstrates the near synchronous onset of increased intensity of marine storminess in the North Atlantic and South Pacific, that characterized the onset of the LIA (AD 1400-1420) (Figure 14). Interestingly, marine storminess in these regions has yet to return to MWP levels, and some combination of changes in earth orbit-induced insolation, solar output, greenhouse gases, and perhaps volcanism and other factors must be invoked to explain the timing and duration of the LIA (Kreutz et al., 1997).
Hindcasting Instrumental Records
The ice core record offers great potential as a tool for deciphering regional climate variability. Where relationships can be developed between instrumental and paleoclimate records, the latter can be used to hindcast the former. Several examples follow.
GISP2/GRIP d18O and dD (deuterium) series have significant correlations with the North Atlantic Oscillation (NAO, pressure contrast between Icelandic Low and Azores High) instrumental series, demonstrating that the former can be used in hindcasting this series (Barlow et al., 1993; White et al., 1997; Figure 15).
Proxy ENSO(El Nino Southern Oscillation) series are available from polar ice cores (Legrand and Feniet-Saigne, 1991) and tropical ice cores (Thompson et al., 1992). For example, over the past 60 years in south polar snow and firn, high methylsulfonic acid (MSA) concentrations are correlated with the occurrence of El Nino years, possibly reflecting more efficient air-sea exchange of dimethyl sulfide (DMS) due to higher winds (Legrand and Feniet-Saigne, 1991). Proxies for Antarctic sea ice extent are also available (Welch et al., 1993; Steig et al., in press).
Anthropogenic Impact on Atmospheric Chemistry
Over the last 200 years, world population has increased by more than 500% (McEvedy and Jones, 1978). One consequence of the population explosion and industrialization is increased CO2, CH4 and nitrous oxide (N2O) emissions over a very short time period. Because the removal of these species from the atmosphere cannot keep pace with the elevated emissions, the concentration of these trace species has been increasing. However, the first direct measurements of these trace gases were not made until after 1957 AD which constitute the best archive of atmosphere prior to 1957. The record of the concentration of CO2, N2O, and CH4 in the atmosphere over the last 200 years has been reconstructed from measurements of the trapped gases in ice (Figure 16). The results show a doubling of CH4, a 25% increase in CO2, and a 10% increase in atmospheric N2O concentrations. While there is still some doubt as to the exact portion of the increase which can be related to human activities, there is no doubt that the majority of the increase in all three species is related to the world population increase and accompanying industrial and land use stresses over the last two centuries.
High resolution time series for sulfate and nitrate from a south Greenland ice core covering the last two centuries show dramatic increases in the concentration of these anions over the past one hundred years and clearly demonstrate the difference between natural, pre-1900 levels of these two acidic species versus post-1900 values (Mayewski et al., 1990; Figure 17). In the pre-anthropogenic atmosphere over Greenland, nitrate levels were approximately twice the sulfate levels. Increased levels of sulfate during the anthropogenic era mask volcanic sulfate levels, indicating the large rise over natural background during this period. An observed increase in excess chloride (portion of chloride not at seasalt ratio) at GISP2 (Mayewski et al., 1993b) as of the 1940s is believed to be a byproduct of the increased levels of anthropogenically derived nitric (HNO3 ) and sulfuric (H2SO4) acids, which are believed to aid in the volatilization of hydrochloric acid (HCl) from seasalt aerosol (Eriksson, 1959). Confirmation of the role that anthropogenic pollutants may have on perturbing the chemistry of the atmosphere comes from the coincidence of increased sulfate levels and depression of North Atlantic temperatures between ~1940-1970 (Wigley, 1990; Charlson et al., 1992), which has been demonstrated by a comparison of GISP2, south Greenland and Yukon Territory ice cores with temperature change records (Mayewski et al., 1993c).
Documentation of the distribution of the radioactive fallout related to the Chernobyl nuclear accident has been aided by analysis of ice cores throughout the polar regions. Remarkably, this event is even detected at the South Pole (Dibb et al., 1990).
Human Response to Climate Change As Deduced From the Paleorecord
Although the issue of human response to climate change is controversial, several recent studies find close correlations in timing between climate change and changes in civilization. These studies have focused on changes in temperature in relation to high latitude societies and changes in moisture availability for mid to low latitude societies.
In ice marginal regions, events such as the disappearance of the Norse colonies in Greenland during the mid to late 14th century appear to be chronologically correlated at some sites with the occurrence of a few extremely cold winters at the onset of the LIA and recorded in Greenland ice cores (Stuiver et al., 1995; Buckland et al., 1996).
Ice cores recovered from the Peruvian Andes have provided evidence connecting the rise and fall of coastal and highland human activity to regional climatic changes. Annual accumulation records reconstructed from the Quelccaya ice cores (Thompson et al., 1985) indicate that an extended period of relatively high accumulation from about 750 to 1050 AD roughly corresponds to when southern highland cultures flourished (Thompson et al., 1994). Low accumulation intervals before and after this period in the ice core are synchronous with the flourishing of Peruvian coastal cultures, which declined during highland wet periods. Similarly, the highland cultures declined during the low accumulation intervals seen in the ice core. An explanation may be related to ENSO events, during which precipitation today in coastal Peru and Ecuador is out of phase with that in the southern Peruvian highlands. The Quelccaya records suggest that this seesaw relation may have been a persistent feature of the regional climate, extending over at least a millennium.
Pronounced dust peaks from the Quelccaya record lasting about 130 years and centered at about 600 and 900 AD seem to reflect local agricultural activity, rather than climatic or volcanic events (Thompson et al., 1987). Analysis of the dust indicates that it is primarily of local origin with few volcanic shards, while the periods of enhanced dust concentration do not correspond to pronounced isotopic variations. These ice core records document an intimate connection between climate and human activity in this region and enhance our understanding of the evolution of human society.
The key scientific questions facing paleoclimate researchers have been articulated in a series of international projects including: PAGES (Past Global Changes), IGBP; CLIVAR (Climate Variability), WCRP; and GLOCHANT (Global Changes in the Antarctic), SCAR. Through the integration of ice, ocean and terrestrial paleorecords, these international efforts seek to develop a basis for understanding the characteristics of natural global environmental change, notably climate change. These paleodata are essential for assessing human influence on the global environment and for the evaluation of predictive climate models.
Several questions and issues have evolved as foci for the paleo community. These consensus views have been expressed in several documents, notably PAGES PANASH (1995) and the PAGES/CLIVAR Intersection (1994). These documents and the lessons learned from previous ice core studies form the basis for the following key scientific aims developed by the Ice Core Working Group.
In order for U.S. ice core research to maintain its cutting edge in the field of paleoclimate, the Ice Core Working Group suggests that the following requirements be vigorously addressed:
The Ice Core Working Group has developed a plan for future U.S. ice core research that outlines ice coring activities in three broad geographic regions (Antarctica, mid-low latitudes and the Arctic).
The vast continent of Antarctica has been a major focus of scientific exploration for relatively few decades when compared to most areas on Earth. Yet what is already known about Antarctica conclusively demonstrates that despite its remote location it plays a significant role in the global system. Encircled by the world's most biologically productive oceanic regions, Antarctica is the largest storehouse of fresh water on the planet, a major site for the production of the cold deep water that drives ocean circulation, a major player in Earth's albedo dynamics, and an important driving component for atmospheric circulation. Thus, Antarctica plays a critical role in the dynamic linkages that couple the spatially and temporally complex components of the Earth's system (atmosphere, biosphere, anthrosphere, hydrosphere, cryosphere and lithosphere).
While several ice cores have been recovered from Antarctica (Figure 18), surprisingly few of them are well dated or provide continuous, multivariate records. As a consequence, understanding of even the modern spatial distribution of some primary ice core parameters is quite limited. Based on modern understanding of the regional differences in climate, the sparse distribution of well documented records does not provide sufficient temporal or spatial coverage. As a consequence, the U.S. and several other nations have embarked on a series of new ice coring activities. Over the next few years U.S. ice coring activities will primarily focus on two community projects in West Antarctica, WAISCORES and U.S. ITASE.
WAISCORES (Siple and Western Divide)
The main goals of the WAISCORES program are to investigate the stability of the West Antarctic Ice Sheet and to investigate the cause of rapid changes in climate. The program is part of the larger West Antarctic Ice Sheet (WAIS) Program already underway. The program will analyze two ice cores, one from Siple Dome, and one from an inland site near the ice flow divide in West Antarctica. The two sites have different characteristics.
Siple Dome, at a lower elevation (680 m) near the edge of the ice sheet, is surrounded by ice streams. The Western Divide (or Inland) site will be at a higher elevation (~2000m) that is more representative of the center of the WAIS. Both sites are expected to provide records of the last 80,000+ years, with annually resolved records to ~11,000 years ago at Siple Dome and ~30,000 years ago at the Western Divide site (Nereson et al., 1996). Site selection activities are underway for the Western Divide site, where drilling is expected to begin around 2002. Drilling started at Siple Dome in 1996/97, and a 1000 m core is expected to be recovered by January 1999.
There are now 14 science groups funded by NSF to work on the Siple Dome core. More projects may be added. In conjunction with other ongoing programs, the Siple Dome Project will address many aspects of climate change and ice sheet stability, including:
For further information about WAISCORES, refer to the Ice Core Working Group document "Science plan for WAISCORES deep ice coring in West Antarctica" or the WAISCORES web site at http://www.maxey.dri.edu/WRC/waiscores, where the science plan and other information is available.
ITASE and U.S. ITASE
The broad aim of the multi-national ITASE (International Trans Antarctic Scientific Expedition) effort is to establish how the recent atmospheric environment (climate and atmospheric composition) is represented in the upper layers of the Antarctic ice sheet. This activity has been formally accepted at the international level by both PAGES (IGBP) and GLOCHANT (SCAR). Primary emphasis is placed on the last ~200 years of the record. This time period was chosen since observational records for the Antarctic are sparse in time and space and because it covers the onset of major anthropogenic involvement in the atmosphere. Four U.S. ITASE traverses are planned as a complement to WAISCORES (Figure 19). The following six major U.S. ITASE scientific objectives have been formulated in order to understand environmental change in West Antarctica:
In fulfilling these objectives, U.S. ITASE in conjunction with ITASE will produce continental-scale "environmental maps," elucidate transfer functions between components of the atmosphere and snow/ice, validate atmospheric models, and interpolate spatial time-series by satellite remote sensing.
For further information concerning U.S. ITASE, refer to the "Science and Implementation Plan for a U.S. Contribution to the International Trans Antarctic Scientific Expedition: 200 Years of Past Antarctic Climate and Environmental Change" (1996).
Proposed Schedule (Antarctic)
Science proposals for ITASE (first attempt)
98/99 field season
Deep drilling at Siple Dome Rock coring at Siple Dome Install 1-3 AWS (Automatic Weather Stations) units around proposed Inland site
Science proposals for ITASE (second attempt)
99/00 field season
Clean up Siple Dome ITASE traverse: Surface evaluation of proposed Inland sites includingsnowpits, shallow cores, shallow radar Install several more AWS units for Inland WAISCORES
Propose preliminary hot water coring at Inland WAISCORES site
Complete maps and models of Inland WAISCORES site and identify proposed sites
00/01 field season
Submit science proposals for Inland WAISCORES site (first attempt) Submit Inland WAISCORES science coordination/lead proposal
ITASE traverse Preliminary hot water coring at Inland WAISCORES site
Submit Inland WASICORES science proposals (second attempt) Submit proposals for preliminary work (site selection, geophysical reconaissance) for future intermediate to deep drilling sites (e.g., South Pole, Titan Dome, Hercules Dome, Transantarctic Mountains, Antarctic Peninsula, Siple Station)
Select Inland WAISCORE site
ITASE traverse Move drilling equipment to Inland WAISCORES site, shallow drilling, set casing for deep drilling
Deep drilling at Inland WAISCORE site
Deep drilling at Inland WAISCORE site
Deep drilling at Inland WAISCORE site
Clean-up Inland WAISCORE site
Mid to Low Latitude Sites
While the longest and most detailed ice core records of global change have been recovered from the polar regions, the vast majority of the Earth's population lives in the mid to low latitudes. It is in these regions where future climate change will have the greatest impact on people's lives. In order to extend the findings from the poles, and to document the regional scale variability associated with global change in the tropical and temperate regions, it is imperative that we continue to develop multi-parameter, high-resolution ice core records from mid to low latitude glaciers that can be compared directly with those available from the Greenland and Antarctic ice sheets. In addition, the most direct paleorecords of mid to low latitude circulation systems (e.g., Asian monsoon and ENSO) are likely to come from ice cores recovered from the Himalaya/Tibetan Plateau and the Andes.
The widespread occurrence of high elevation glaciers in the mountains of central Asia provides a variety of sites from which ice core records can be recovered (Figure 20). Well-dated, high-resolution (i.e., annual to decadal) ice core records offer a means of extending the climate record back in time in this region where the Asian monsoon impacts almost half of the world's human population. In addition, the Tibetan Plateau exerts a strong influence on global climate, making the study of climatic change in this region globally significant. In this way, we hope to better understand the natural variability of the Asian monsoon climate system and identify processes and forcing factors that contribute to this variability. Ultimately, this improved understanding should lead to the development of models that can explicitly forecast future variability of climate in monsoon Asia. The collection and analysis of ice core records from central Asian glaciers, and comparison with instrumental and other paleorecords available from the highlands of central Asia (e.g., tree rings, lake sediments, loess, glacier variations, peat deposits), are the central objectives of the Himalayan Interdisciplinary Paleoclimate Program (HIPP) (Wake and Mayewski, 1996).
The current extent of surface-to-bedrock ice cores from central Asian glaciers that have already been recovered (i.e., Dunde (Thompson et al., 1989) and Guliya (Thompson et al., 1997) ice caps) provide very valuable but limited spatial coverage. Given the spatially variable influence of circulation patterns, there is clearly a need to develop detailed ice core records throughout central Asia if we are to document and determine the causes of climate variability over this broad region. U.S. programs are currently recovering and analyzing ice cores from the Mt. Everest and Mt. Xixabangma regions of the eastern Himalaya, which should provide high resolution records that can be directly tied to monsoon circulation. Future ice coring programs should focus on the recovery of surface-to-bedrock cores from suitable high elevation glaciers in the other mountain ranges that are spread throughout central Asia, including the western Himalaya, Karakoram, the Pamirs, the Tien Shan, the Tanggula Shan and southeastern Tibet (Figure 20).
Ice core records from Huascaran in the Peruvian Andes (Figure 21) have provided a record of low latitude climate change extending back through the past glacial-interglacial transition (Thompson et al., 1995). A good correlation is found between the Huascaran isotopic record and a marine isotopic record recovered off the coast of Portugal, indicating a close climatic connection between the North Atlantic and the Amazon Basin.
Unfortunately, many of the low latitude, high elevation glaciers, such as those in the Andes and Central Asia, are currently retreating. While complete wastage may not occur for some time, increased melting is destroying their structure and compromising the valuable information these unique archives contain. Thus there is an urgent need to obtain a spatial network of high-quality ice cores from these glaciers and ice caps as soon as possible.
Models of future climatic change from anthropogenic emissions and the records produced from the GISP2 and GRIP ice cores have played major roles in placing the importance of the Arctic region into global perspective. The greatest warming from increased greenhouse gases, as would occur with a doubling of CO2, often is predicted to occur in the Arctic (ARCSS, 1993). Future warming could also result in melting of many permafrost regions in northern polar regions, with the subsequent release of additional greenhouse gases (i.e., CH4). Such a scenario would have a major impact on biogeochemical cycles initially in the Arctic and sub-Arctic and, eventually, globally. The rapid climatic shifts observed in the Greenland cores particularly emphasize the sensitive nature of the Arctic atmosphere. Changes in the global climate system are undoubtedly amplified in the Arctic. The ideal example is the much higher cooling observed in the high latitudes of the northern hemisphere compared to mid latitudes and equatorial regions following an explosive volcanic eruption in the tropics (Self et al., 1981). Once changes in the climate system occur in the Arctic, different hemispheric circulation patterns may develop as an adjustment to new pole-equator pressure gradients. Moreover, feedbacks and linkages among various components in the Arctic climate system (e.g., sea ice, meltwater, iceberg discharge, ocean circulation, variability in albedo) may then lead to further modification of global climate (e.g., Alley, 1995). Thus, understanding the natural variability in Arctic climate is of utmost importance, not only to the U.S. Global Change Program, but to the major population centers of the entire northern hemisphere.
As the U.S. ice core community attempts to decipher the natural variability in climate across the circum-Arctic, the complexity of the task is becoming evident. Complicating the undertaking is the existence of many modern climatic boundaries, both north-south and east-west, especially in the Canadian Arctic archipelago (Maxwell, 1980, 1982), the region with the greatest number of ice caps in the Arctic. As a result, any single ice core from the Arctic is unable to provide a complete picture of variability in past climate. This regionalization of Arctic climate is most pronounced under warmer climatic conditions (O'Brien et al., 1995).
To evaluate the regional nature of Arctic climate, it is necessary for the U.S. ice core community to obtain an array of ice cores across the circum-Arctic (Figure 22), including the Greenland Ice Sheet , smaller ice caps and mountain glaciers.Recent work in Greenland is being coordinated through the Program for Arctic Regional Climate Assessment (PARCA, 1997), an initiative directed to measuring and understanding the mass balance of the ice sheet. Special emphasis is placed on establishing long term histories of climate and accumulation rate into which results of short term observations, such as those from remote sensing platforms, can be placed. Four cores covering the last 300-900 years and eight cores covering the last 50-70 years have been obtained from sites well distributed over the ice sheet (Figure 22). The records are used to study interannual and decadal scale changes in accumulation rates, climate and atmospheric chemistry. Automatic weather stations installed at many of the core sites are being used to study the processes of snow accumulation and surface ablation. One of the expected products of the core analyses is an improved map of surface accumulation rates for the Greenland Ice Sheet. Work in future field seasons is expected to concentrate on recovering additional 50 to 70 year core records.
Work outside of Greenland has included the collection of several cores from the Canadian Arctic, primarily by the Geological Survey of Canada, although more extensive collaboration with U.S. investigators has begun.Several cores from the Russian Arctic have been collected by the international ice coring community, including U.S. investigators (e.g., Kupol Ventrenxy), thus providing a perspective on past change from the eastern Arctic (Figure 22).However, parameters measured in many of these cores have been limited to d18O, and occasionally electrical conductivity and melt layers (a proxy for summer temperature; e.g., Koerner and Fisher, 1990). Consequently, the multiparameter approach to ice core analysis developed for the GISP2 core is necessary to obtain the most complete paleoclimate record available. We encourage individuals intending to obtain future cores from the Arctic to take this approach.
Guidance for the collection of future cores by U.S. researchers is available from the Ice-Core Circum-Arctic Paleoclimate Program (ICAPP-PAGES). This recently developed international program will assist in synthesizing existing ice core records, especially for use with other proxy data (such as those used in CAPE, the Circum-Arctic Paleoenvironments program of PAGES), as well as in developing multinational programs to more thoroughly evaluate ice core records from the smaller ice caps of the Arctic.
The 1995 Penny ice core was the first core collected under the auspices of ICAPP. Presently, ICAPP is directly involved with plans to collect a surface-to-bedrock cores from Akademii Nauk and Severnaya Zemlya. This will be an international effort with the intention of producing the first high-resolution, multi-parameter record through the Holocene, and probably a discernible record to the last glacial maximum, from the Russian Arctic. Because ice coring on the smaller ice caps of the Arctic and the collection of shallow-depth cores from Greenland does not require the extensive planning needed for deep coring operations to bedrock on the Greenland and Antarctic ice sheets, a detailed schedule of upcoming projects is not presented in this document. Nevertheless, plans for work in the near future on the smaller ice caps of the Arctic continue to be developed (e.g., Devon Ice Cap, near Nord Station). Plans for the collection of high elevation ice cores from Alaska (e.g., Churchill-Bona, Denali) are also under development.We anticipate a continued effort in the Arctic by the U.S. ice core community in the more distant future.
We are grateful for the support and guidance of the Office of Polar Programs at the National Science Foundation; the helpful comments from the community, especially from R. Alley and E. Waddington; editorial comments from I. Pittallwala and F. Smith helped blend the writing styles of the various authors;the graphic and layout design of J. Fithian. Portions of this document were taken from Mayewski and Bender (1995). Much of the data presented in the figures accompanying this document are available to the public on the NGDC server ftp://ftp.ngdc.noaa.gov/paleo/ , the NOAA web site http://www.cmdl.noaa.gov/ftpdata.html and the NSIDC web site http://arcss.colorado.edu.