The geophysical metronome (Milankovitch components of the past surface temperature variations) and the isotope–temperature transfer function deduced from the borehole temperature profile at Vostok station, Antarctica, are applied to date the 2500 mdeep ice core from Dome Fuji station, Antarctica, and to reconstruct paleoclimatic conditions at the drilling site on the basis of the local δ18O isotope record. Special attention is paid to consistency of this depth–age relation with the mass-balance reconstruction and predictions of ice-flow modeling. the present-day ice mass-balance rate at Dome Fuji is estimated as 3.2 cm a–1. the ice age at the borehole bottom (590m above the bedrock) is around 335 ± 4.5 kyr and may reach 2000 kyr at about 3000 mdepth.The difference in the ice-sheet surface temperatures between Holocene optimum and Last Glacial Maximum is found to be 17.8˚C at the temporal isotope/temperature slope, about 30% lower than the modern geographical estimates. A good agreement between modeled and measured (preliminary data) borehole temperatures is obtained at the geothermal flux 0.059 Wm–2 and ice-fusion temperature (–2˚C) at the ice–rock interface with minimum (zero) melt rates.

Large-scale positive degree-day based melt parameterizations for the Greenland ice sheet are highly sensitive to their parameters (standard temperature deviation, snow and ice degree-day factors). In this paper, these parameters are simulated with a coupled atmosphere–snow regional climate model for southern Greenland during summer 1991, forced at the lateral boundaries with European Centre for Medium-Range Weather Forecasts re-analyses at a high horizontal resolution of 20 km. the calculated (from net ablation, i.e. melt minus refreezing) snow and ice positive degree-day factors vary considerably over the ice sheet. At low elevations, the modelled snow degree-day factor closely approaches the generally accepted value of 3 mm w.e. d–1 ˚C–1.Higher up the ice sheet, large values up to 15 mm w.e. d– 1 ˚C– 1 are simulated. for ice melt, maximum values of 40 mm w.e. d–1 ˚C– 1 are found. the snow and ice positive degree-day factor distributions peak, respectively, at 3 and 8mm w.e. d–1 ˚C–1. Refreezing is of small importance close to the ice-sheet margin. Higher up the ice sheet, refreezing considerably lowers the amount of net ablation. the monthly simulated 2 m air-temperature standard deviation exhibits a strong seasonal cycle, with the highest (3.0–5.0˚C) values in May and June. July shows the lowest temperature fluctuations, due to the melting of the surface.

In the search for abetter understanding of the dominant mechanisms of the Earth’s climate system, we present a study of rapid warm-climate transitions to Dansgaard– Oeschger events as seen in the ice cores from the Greenland ice sheet. We present a continuous δ18O record from the Greenland Icecore Project (GRIP) core with a resolution of 5 years until 50 kyr BP and of 20 years until 100 kyr BP. These data are compared with other high-resolution records, i.e. the Greenland Ice Sheet Project II (GISP2) chemistry record (25 years until 50 kyr BP) and the GRIP Ca2+ record (3 years until 100 kyr BP). All records have been transformed to the GISP2 Meese/Sowers time-scale. the high-resolution records are separated into interstadials and stadials, defined by the GRIP and GISP2 δ18O records. We examine in detail the transitions into the Dansgaard–Oeschger events, and propose a scenario for the changes that occur in the different ice-core records during the approximately 50 years the transition takes. Themain difference from previous studies is the much higher-resolution datasets available until 50 kyr BP; previous high-resolution studies by Taylor and others (1997) have focused on only the Younger Dryas/Preboreal transition. the data are compared to model simulations of the Dansgaard–Oeschger events performed with the CLIMBER-2 model of intermediate complexity (Petoukhov and others, 1998 ; Ganopolski and Rahmstorf, 2001).

The interannual variations in atmospheric transport patterns to Summit, Greenland, are studied using twice-daily, three-dimensional, 10 day backward trajectory data corresponding to the summers (1 June–31 August) of 1989–98. While previous trajectory climatology studies have been prepared for Summit, the present work considers both the horizontal and vertical components of transport. A three-dimensional residence-time methodology is employed to account for both horizontal and vertical components of transport. the vertical transport component is quantified by passing all trajectories through a three-dimensional grid and tracking the time spent (i.e. the residence time) in each gridcell. This method also allows inspection of trajectory altitude distributions corresponding to transport from upwind regions of interest. the three-dimensional residence-time methodology is shown to be a valuable tool for diagnosing the details of long-range atmospheric transport to remote locations. for Summit, we find that the frequent transport from North America tends to occur at low altitudes, whereas transport from Europe is highly variable. Mean summertime flow patterns are described, as are anomalous patterns during 1990,1996 and 1998.

We present results from a greenhouse warming experiment obtained from an atmosphere–ocean–sea-ice general circulation model that is interactively coupled with a three-dimensional model of the Greenland ice sheet. the experiment covers the period 1970–2099 and is driven by the mid-range Intergovernmental Panel on Climate Change SRESB2 scenario. the Greenland model is a thermomechanical high-resolution (20 km) model coupled with a viscoelastic bedrock model. the melt-and-runoff model is based on the positive degree-day method and includes meltwater retention in the snowpack and the formation of superimposed ice. the atmospheric–oceanic general circulation model (AOGCM) is a coarse-resolution model without flux correction based on the Laboratoire de Météorologie Dynamique (Paris) LMD 5.3 atmospheric model coupled with a primitive-equation, free-surface oceanic component incorporating sea ice (coupled large-scale ice–ocean (CLIO)). By 2100, average Greenland annual temperature is found to rise by about 4.5˚C and mean precipitation by about 35%. the total fresh-water flux approximately doubles over this period due to increased runoff from the ice sheet and the ice-free land, but the calving rate is found to decrease by 25%. the ice sheet shrinks equivalent to 4 cm of sea-level rise. the contribution from the background evolution is not more than 5% of the total predicted sea-level rise. We did not find significant changes in the patterns of climate change over the North Atlantic region compared with a climate-change run without Greenland fresh-water feedback.

The high accumulation rate, nearly complete preservation and detailed chronology of the Eclipse ice core, Yukon Territory, Canada, are well suited for comparison of the glaciochemical recordwith instrumental time series of temperature, precipitation and sea-level pressure. Results of cross-correlation analysis of instrumental temperature records with the Eclipse δ18O time series reveal a significant positive relationship between summertime δ18O at Eclipse and summer (April–September) temperatures in northwestern North America. the results indicate that the Eclipse δ18O time series provides a better proxy for regional temperature than does the δ18O time series from the Mount Logan ice-core record for which only negative correlations were found. Winter accumulation at Eclipse is significantly correlated with several sites in Alaska, but not with any sites in the Yukon. the δ18O, accumulation and glaciochemical time series also display significant correlations with the Northern Hemisphere sea-level pressure dataset, especially between wintertime sulfate and nitrate concentrations at Eclipse and the intensity of the wintertime Siberian high and Aleutian and Icelandic lows. These results suggest that year-to-year variability in the deposition of pollutants at Eclipse can be linked to changes in atmospheric circulation, while long-term trends can be explained by changes in source strength.

In this paper, we explore the climate signal contained in the annual snow-accumulation time series from a high-altitude ice core drilled on Mount Logan in the Saint Elias mountain range of western Canada. With the global meteorological fields from the U.S. National Centers for Environmental Prediction re-analysis, we construct composites of the atmospheric circulation and temperature patterns associated with anomalous snow accumulation at the Mount Logan site over the period 1948–87. These results confirm, with an independent method, previous work that identified the existence of a coherent upper-tropospheric circulation anomaly extending over much of the North Pacific Ocean and North America that is associated with snow accumulation at the site. This anomaly has a similar structure to that associated with the extratropical response to the El Niño–Southern Oscillation. Coherent structures consistent with this circulation pattern also exist in both air- and land-temperature fields. In particular, heavy (light) snow accumulation at the site is associated with warmer (colder) air and surface temperatures over the North Pacific Ocean and North America. Over the North Pacific, the sea-surface temperature anomaly associated with heavy snow accumulation at the site has a “horseshoe” pattern that is similar to that associated with the Pacific Decadal Oscillation.

An annually dated ice core recovered from South Pole (2850 m a.s.l.) in 1995, that covers the period 1487–1992, was analyzed for the marine biogenic sulfur species methanesulfonate (MS). Empirical orthogonal function analysis is used to calibrate the high-resolution MS series with associated environmental series for the period of overlap (1973–92). Utilizing this calibration we present a ~500 year long proxy record of the polar expression of the El Niño–Southern Oscillation (ENSO) and southeastern Pacific sea-ice extent variations. These records reveal short-term periods of increased (1800–50, 1900–40) and decreased sea-ice extent (1550–1610, 1660–1710, 1760–1800). In general, increased (decreased) sea-ice extent is associated with a higher (lower) frequency of El Niño events.

Three ice cores distributed across Dasuopu (DSP) glacier in the Himalaya were recovered. the annual net accumulation (An) record reconstructed from one of the cores reflects major precipitation variability for the central Himalaya. This record agrees well with precipitation records from Nepal and northeast India, while it does not compare well with the all-India precipitation record. Singular spectrum analysis applied to the 200 year long Dasuopu accumulation (DSP An) record indicates significant variability on interannual and inter-decadal time-scales. the record shows that the early 18th century was dry, wet conditions prevailed during the period 1820–1930, and after 1930 An decreased to its present value. Distinct secular trends were not found in the record. for the period 1950– 94, the variability of DSP An is found to correlate significantly with the thermal contrast between the Tibetan Plateau and the Indian Ocean surface temperature. Additionally, a significant relationship between DSP An and solar activity is detected.

In order to select glacier sites suitable for the reconstruction of short-term climate variability, chemical constituents of shallow cores from Chimborazo, Ecuador, Illimani, Bolivia, and Cerro Tapado, Chile, were analyzed to determine if they represent the climate and atmospheric conditions of the particular area. Pronounced variations were observed in the Chimborazo core, which were attributed to the seasonal occurrence of wet and dry periods. Using chemical tracers, the Pacific and the Amazon basin were identified as the two principal sources of atmospheric moisture. In the Illimani record, seasonality is obvious but is less regular than in the Chimborazo core. Chemical tracers point to the Amazon basin as a moisture source. Post-depositional sublimation masks a possible atmospheric signal in the glaciochemical records from Cerro Tapado. the arid conditions and a prevailing high condensation level also cause little variability in the stable-isotope content. Irregular, more negative δ18O values occur during the humid phases of El Niño events. Despite alterations due to sublimation, chemical tracers indicate the Pacific Ocean as a main moisture source for precipitation preserved on Glaciar CerroTapado.

Comparison of climate records from Antarctic and Greenland ice cores shows that the two regions respond asynchronously during millennial-scale climate changes. the apparent out-of-phase relationship between the records has been described as a climate ``seesaw’’ in which cooling in the Northern Hemisphere is balanced by warming in the Southern Hemisphere. the same relationship has also been attributed to the initiation of climate-change events in the Southern Hemisphere, rather than the North Atlantic as is conventionally assumed. A simple statistical approach−band-pass filtering combined with lag–correlation tests−used to examine the phase relationships in more detail shows that neither an anti-phase nor a phase-lag relationship adequately describes the observations. Whereas Antarctic and Greenland climate records do exhibit approximate anti-phase behavior about 50% of the time, they are generally in phase during cooling. A phase lead of Southern Hemisphere climate of 1000–1600 years is statistically indistinguishable from a lag of 400–800 years, whether for Dansgaard–Oeschger, Heinrich or longer-duration events. the ``seesaw’’ or ``Southern lead’’ appearance of the data arises from the fundamentally different characteristics of the climate time series, most importantly the absence of rapid warming events in Antarctica comparable to those in Greenland. to be consistent with the observations, climate models will need to capture these characteristics, in addition to reproducing the correct phase relationships.

Rapid climate changes during the last glacial period were first observed in ice-core records (Dansgaard and others, 1982). These shifts between interstadials, called Dansgaard–Oeschger (D-O) events, and stadials or deep glaciation were later seen in Atlantic sediment records (Bond and others, 1993), pointing to the ocean circulation as a strong component in the dynamics of these shifts (Wright and Stocker, 1991). the interstadial states are observed to have a characteristic ``sawtooth’’ shape, indicating a gradual drift of the stable interstadial state toward the stable stadial state. In order to contrast the two climate states, we have separated the δ18O signal from the Greenland Icecore Project ice core into periods corresponding to the two states. the climate variability in the two different climatic states is different (Johnsen and others, 1997). We find that the standard deviation is significantly larger in the stadial than in the interstadial state. Both states are found to have a larger standard deviation than the Holocene part of the record. the correlation times in the different states are difficult to obtain because of limited data resolution and diffusion of the isotopic signal. However, using a statistical technique, we have estimated the correlation times. We do not find significant differences in the correlation times, which are of the order of months, in the different climatic states. These findings are interpreted in the context of a simple linear stochastic model which provides information about the relative roles of the climatic forcing and the stability of the climate state governing the climate variability.

Snow-pit and shallow firn-core records of oxygen isotope ratios (δ18O) and trace ion species were generated at a high-accumulation site on Law Dome, East Antarctica. Concordance between accumulation events identified in records up to 7.7 km a part confirms that the observed glaciochemical variations are the result of regional rather than local surface effects. This allows calibration of the snow-pit records with measured meteorological parameters. Net accumulation periods that comprise the snow-pit record are identified using hourly snow-accumulation measurements from a co-located automatic weather station (AWS). Particular focus is given to three net accumulation periods preserved during austral summer 1999/2000 that correspond to the top 0.5 m of the snow pit. Local meteorological conditions recorded during the summer accumulation periods by the AWS are combined with regional and synoptic-scale meteorology derived from Casey station (110 km away) and Advanced Very High Resolution Radiometer satellite imagery to identify potential source regions for chemical signals preserved in summer snow at Law Dome.

To determine annual layers for reconstructing the past environment at annual resolution from ice cores, we employed snow-stake data back to 1972, tritium content, solid electrical conductivity measurements (ECM) and stratigraphic properties for the 73m ice core at the H72 site, east Dronning Maud Land, Antarctica. the average annual surface mass balance at H72 is 307 mma–1w.e. during the last 27 years from continuous accumulation data, 317 mma–1 w.e. according to the densification model and 311 mma–1 w.e. according to the average surface mass balance for 167 years based on annual-layer counting. the ECM age is closely coincident with tritium age, and corresponds with the snow-stake record back to AD 1972 from the surface to 15 m depth. the H72 ice core is dated as AD 1831by ECMat 73.16 mdepth.The time series of yearly surface mass balance at H72 shows an almost constant 311 mm a–1 w.e. for the last 167 years. the oxygen-isotope records indicate a significant trend to lower values, with negative gradient of 1.7% (100 years)–1.

Conditions for passive folding near ice-sheet centers are derived treating the ice as an anisotropic viscous medium. Vertical uniaxial compression at a dome, or pure shear stress near a ridge divide, both tend to stretch and flatten folds, while horizontal simple shear deformation tends to overturn folds. Overturned folding is likely in a given vicinity in a steady-state flow field, if the initial slope of a layer disturbance exceeds the ratio of compressive/extensive deformation to shear deformation. Analytical equations for particle tracks in steady state allow us to model the evolution of layers with initial slope disturbances (``wrinkles’’). the effects of anisotropy are explored using an analytical solution for the strain rate as a function of a vertically symmetric c-axis orientation distribution, called a cone fabric. Stronger anisotropy (small cone angle) makes the material softer in horizontal shear, and facilitates folding, i.e. elements with smaller slopes can be overturned for the same stress. the relation between anisotropy and folding is complicated by the fact that, for a range of cone angles, the material is also softer in compression, which opposes folding. Simulating a layer with spatially variable tilt of cone symmetry axes, and accounting for fabric development, demonstrates that variations in the fabric cause localized flow variations that could create the initial perturbations.

A one-dimensional model problem for computation of the age field in ice sheets, which is of great importance for dating deep ice cores, is considered. the corresponding partial differential equation (PDE) is of purely advective (hyperbolic) type, which is notoriously difficult to solve numerically. By integrating the PDE over a space–time element in the sense of a finite-volume approach, a general difference equation is constructed from which a hierarchy of solution schemes can be derived. Iteration rules are given explicitly for central differences, first-, second- and third-order (QUICK) upstreaming as well as modified TVD Lax–Friedrichs schemes (TVDLFs). the performance of these schemes in terms of convergence and accuracy is discussed. Second-order upstreaming, themodifiedTVDLF scheme with Minmod slope limiter and, with limitations of the accuracy directly at the base, first-order upstreaming prove to be the most suitable for numerical age computations in ice-sheet models.

The Comninou–Dundurs effect (C-D effect) has yet to be considered in slip processes at the base of ice sheets and glaciers. This effect exists whenever slip occurs between two solids of differing mechanical properties and at least one of the solids is compressible. It gives rise to an additional normal traction stress across a non-uniformly slipping interface, a stress which otherwise would not exist if the two solids had identical properties. the C-D effect should exist to some degree, therefore, wherever ice slips over glacial till. It will be enhanced where large slip gradients exist, such as at boundaries of subglacial lakes. In this paper it is shown that, because of the C-D effect, subglacial lakes that lie at a basal till/ice interface may migrate in the direction of increasing ice thickness.

A local two-dimensional flow model which accounts for the anisotropic behaviour of polar ice and the evolution of its strain-induced anisotropy is briefly reviewed. Due to its complexity, it is not yet possible to use this model to simulate the flow of a whole ice sheet, and its potential applications are presently restricted to limited spatial domains around existing drilling sites. In order to calculate the local flow of ice, boundary conditions must be applied on the lateral edges of the studied domain. Since these limits correspond to fictitious sections of the ice sheet, the type of boundary condition to adopt is not obvious. In the present paper, different kinds of boundary conditions of the Dirichlet type, applied at the lateral boundary of an idealized ice sheet of simplified geometry, are discussed. This will serve as a first step towards the coupling of the local flow model with a global ice-sheet flow model.

Air–snow transfer processes impact both ice-core interpretation and exchange affecting atmospheric chemistry. An understanding of seasonal differences in the character of the surface snow will facilitate evaluation of possible preferential seasonal exchange of reactive chemical species. Both diffusive processes and advective (ventilation) processes can serve to alter the physical, chemical and isotopic character of snow and firn. In this paper, we examine measurements of surface roughness over the course of a year at Summit, Greenland, and the implications for snow and firn ventilation. At Summit, during the winter-over experiment, summer and fall sastrugi amplitudes were approximately 5 cm and had smoothly curved profiles. the average amplitudes experienced mild increases in January, but by the end of February through March the amplitude increased to approximately 20 cm, and the profiles exhibited more abrupt geometries. Calculations are performed to show the potential impact of the changing roughness on interstitial ventilation rates in the snow, assuming that the permeability profile does not change in time. Under high winds, ventilation velocities in the near-surface snow can be up to 3 cms–1 in the winter, compared to 1 cms–1 in the summer. the frequency of 12 ms–1 winds in the summer, however, is less than in the winter. Under low-wind conditions, the summer roughness causes ventilation rates that are comparable to diffusion rates. However, in winter even 5 ms–1 wind conditions can cause the interstitial airflow due to ventilation to exceed the diffusion rates.

A high-resolution study of c-axis fabrics has been performed on the NorthGRIP deep ice core from North Greenland. A newly developed automatic ice-fabric analyzer has been used to measure c-axis orientations on vertical thin sections at 142 different depths between 100 and 2920 m in the core. Detailed comparison studies show that this new method produces results that are in full accordance with those obtained from conventional manual measurements. Fabric development in ice of the NorthGRIP core can be explained by rotation of c axes, controlled by the prevailing stress systems. In the upper 800 mthe c axes appear to distribute uniformly, but in the depth interval 900–2500m they tend to cluster around a vertical plane. This vertical girdle pattern is strengthened with depth, forming the first clear evidence of this fabric type in a Greenland deep ice core. Such fabric development has previously been observed in the Vostok ice core from East Antarctica, where converging flow is believed to occur. the most likely interpretation of the NorthGRIP girdle fabric is that the c axes are rotating away from a horizontal-tension axis across the main ice divide, which runs north-northwest–south-southeast through the NorthGRIP drilling site. This is supported by information available from surface velocity measurements. Within the girdle fabric, increased concentration of c axes parallel to the core axis is observed with increasing depth, indicating combined effects of vertical compression and horizontal tension on the fabric development. from about 2500 m depth, the girdle-type fabric starts to give way to a strong vertical single-maximum fabric, which persists to 2920m depth, where drilling was terminated 150 mabovebedrock in the year 2000. the single maximum seems to suggest that bed-parallel simple shear is exerting a strong influence on the fabric in the lowest part of the ice sheet down to 2920 m.