Deuterium analysis of ice core GISP2

The Greenland Ice Sheet Project 2 (GISP2) core can enhance our understanding of the relationship between parameters measured in the ice in central Greenland and variability in the ocean, atmosphere, and cryosphere of the North Atlantic Ocean and adjacent land masses. Seasonal (summer, winter) to annual responses of dD and deuterium excess isotopic signals in the GISP2 core to the seesaw in winter temperatures between West Greenland and northern Europe from A.D. 1840 to 1970 are investigated. This seesaw represents extreme modes of the North Atlantic Oscillation, which also influences sea surface temperatures (SSTs), atmospheric pressures, geostrophic wind strength, and sea ice extents beyond the winter season. Temperature excursions inferred from the dD record during seesaw/extreme NAO mode years move in the same direction as the West Greenland side of the seesaw. Symmetry with the West Greenland side of the seesaw suggests a possible mechanism for damping in the ice core record of the lowest decadal temperatures experienced in Europe from A.D. 1500 to 1700. Seasonal and annual deuterium excess excursions during seesaw years show negative correlation with dD. This suggests an isotopic response to a SST/ land temperature seesaw. The isotopic record from GISP2 may therefore give information on both ice sheet and sea surface temperature variability. Cross-plots of dD and d show a tendency for data to be grouped according to the prevailing mode of the seesaw, but do not provide unambiguous identification of individual seesaw years. A combination of ice core and tree ring data sets may allow more confident identification of GA and GB (extreme NAO mode) years prior to 1840.

(Table 1) Deuterium excess record during different time periods in Lomonosovfonna ice core

In this paper we present a deuterium excess (d) record from an ice core drilled on a small ice cap in Svalbard in 1997. The core site is located at Lomonosovfonna at 1255 m asl, and the analyzed time series spans the period 1400-1990 A.D. The record shows pronounced multidecadal to centennial-scale variations coherent with sea surface temperature changes registered in the subtropical to southern middle-latitude North Atlantic during the instrumental period. We interpret the negative trend in the deuterium excess during the 1400s and 1500s as an indication of cooling in the North Atlantic associated with the onset of the Little Ice Age. Consistently positive anomalies of d after 1900, peaking at about 1950, correspond with well-documented contemporary warming. Yet the maximum values of deuterium excess during 1900-1990 are not as high as in the early part of the record (pre-1550). This suggests that the sea surface temperatures during this earlier period of time in the North Atlantic to the south of approximately 45°N were at least comparable with those registered in the 20th century before the end of the 1980s. We examine the potential for a cold bias to exist in the deuterium excess record due to increased evaporation from the local colder sources of moisture having isotopically cold signature. It is argued that despite a recent oceanic warming, the contribution from this local moisture to the Lomonosovfonna precipitation budget is still insufficient to interfere with the isotopic signal from the primary moisture region in the midlatitude North Atlantic.

Accumulation rates for western Dronning Maud Land from firn cores and snow pits

As a result of intensive field activities carried out by several nations over the past 15 years, a set of accumulation measurements for western Dronning Maud Land, Antarctica, was collected, based on firn-core drilling and snow-pit sampling. This new information was supplemented by earlier data taken from the literature, resulting in 111 accumulation values. Using Geographical Information Systems software, a first region-wide mean annual snow-accumulation field was derived. In order to define suitable interpolation criteria, the accumulation records were analyzed with respect to their spatial autocorrelation and statistical properties. The resulting accumulation pattern resembles well known characteristics such as a relatively wet coastal area with a sharp transition to the dry interior, but also reveals complex topographic effects. Furthermore, this work identifies new high-return shallow drilling sites by uncovering areas of insufficient sampling density.

Seasonally resolved chemical records of six ice cores from the Dronning Maud Land, Antarctica

The ammonium, calcium, and sodium concentrations from three intermediate depth ice cores drilled in the area of Dronning Maud Land, East Antarctica, have been investigated. Since all measurements were performed by a high-resolution Continuous Flow Analysis system, for the first time seasonal signals of chemical trace species could be obtained from the interior of central Antarctica over a period of approximately 2 millennia. Although the elevation as well as the accumulation rate differ between the drilling sites, similar values were obtained by comparing mean concentrations spanning the last 900 years. However, a distinct lack of intersite correlation was found on decadal timescales. Despite a noticeable accumulation change, apparent in one core, no significant concentration change of all three species has occurred. All the measured ions show clear seasonal signals over the whole records. While the sea-salt-related component sodium peaks simultaneously with calcium, the maximum ammonium concentration occurs in the snow with a time lag of 2 months after the sea-salt peak. More than 60% of the calcium concentration can be attributed to an ocean source. Elevated sodium concentrations were found within this millennium compared to mean values of the whole records, but the spatially varying shape of the increase suggests that a possible climatic signal is biased by local deposition effects.

Match points used for synchronization of NGRIP, GRIP, and GISP2

We here present a synchronization of the NGRIP, GRIP, and GISP2 ice cores based mainly on volcanic events over the period 14.9-32.45 ka b2k (before AD 2000), corresponding to Marine Isotope Stage 2 (MIS 2) and the end of MIS 3. The matching provides a basis for applying the recent NGRIP-based Greenland Ice Core Chronology 2005 (GICC05) time scale to the GRIP and GISP2 ice cores, thereby making it possible to compare the synchronized palaeoclimate profiles of the cores in detail and to identify relative accumulation differences between the cores. Based on the matching, a period of anomalous high accumulation rates in the GISP2 ice core is detected within the period 16.5-18.3 ka b2k. The d18O and [Ca2+] profiles of the three cores are presented on the common GICC05 time scale and generally show excellent agreement across the stadial-interstadial transitions and across the two characteristic dust events in Greenland Stadial 3. However, large differences between the d18O and [Ca2+] profiles of the three cores are seen in the same period as the 7-9% increase in the GISP2 accumulation rate. We conclude that changes of the atmospheric circulation are likely to have occurred in this period, altering the spatial gradients in Greenland and resulting in larger variations between the records.

Age models of five ice cores from the Dronning Maud Land, Antarctica

For the first time, annually resolved accumulation rates have been determined in central Antarctica by means of counting seasonal signals of ammonium, calcium, and sodium. All records, obtained from three intermediate depth ice cores from Dronning Maud Land, East Antarctica, show rather constant accumulation rates throughout the last 9 centuries with mean values of 63, 61, and 44 mm H2O yr**-1 and a typical year-to-year variation of about 30%. For the last few decades, no trend was detected accounting for the high natural variability of all records. A significant weak intersite correlation is apparent only between two cores when the high-frequency part with periods less than 30 years is removed. By analyzing the records in the frequency domain, no persistent periods were found. This suggests that the snow accumulation in this area is mainly influenced by local deposition patterns and may be additionally masked by redistribution of snow due to wind. By comparing accumulation rates over the last 2 millennia a distinct change in the layer thickness in one of the three cores was found, which might be attributed either to an area upstream of the drilling site with lower accumulation rates, or to deposition processes influenced by surface undulations. The missing of a clear correlation between the accumulation rate histories at the three locations is also important for the interpretation of small, short time variations of past precipitation records obtained from deep ice cores.