The primary objective of our work is to develop an accurate estimate of the annual accumulation at various sites on the Greenland ice sheet from chemical analysis of 20-m to 150-m deep ice cores. A second objective is to develop high-resolution, spatially distributed ice-core chemical data for modeling changes in atmospheric chemistry from pre-industrial time through the present. The main atmospheric chemistry question that we are focusing on concerns the historical change in atmospheric oxidation capacity, mainly concentrations of hydroxyl radical (OH).

Work Done in 1996

During the 1996 field season we had two people (Martin Anklin and B. Snider) in the field to participate in the drilling and chemical analysis of the cores at GITS (Camp Century). Field analyses for the firn part of the core were necessary because one species of interest, formaldehyde, is changed during shipment. Subsequently, Betty Snider went to Ohio State and cut a longitudinal section from each piece of core from both GITS and Tunu and shipped the ice to our laboratory in Tucson. Chemical analyses for the GITS cores should be completed in September, 1996; analyses on the Tunu cores should be completed in January, 1997.

We completed analyses for the Nasa-U, Humboldt and Crawford Point cores that were drilled in spring 1995. The NASA-U annual layer thickness data are complete, and differences with Ohio State (E. Mosley-Thompson) analyses reconciled. For Humboldt the annual layer thickness data for the 20-m cores and the top 20-m of the main core are complete, and differences with Ohio State analyses reconciled. Our analyses of the Crawford Point data are complete, but lacking density versus depth measurements, estimated layer thicknesses are less accurate than for other sites. Also, we would like to compare our results with delta-18 O analyses that will be done by Colorado (J. White).

Results to Date

The NASA-U core (73.84 N, 49.49 W) and Humboldt Glacier core (78.53 N, 56.83 W) gave records extending back 351 and 848 years (1645-1995 and 1168-1995), respectively. Annual layers were identified in the cores by using concentrations of hydrogen peroxide, calcium, ammonium, and nitrate, which complimented the dust analyses (Humboldt and NASA-U) and delta-18 O analyses (Nasa-U) being done by Ohio State. We used this multi-parameter approach in an effort to get maximum accuracy in estimating depths of annual layers and thus accumulation rates. Each parameter gives a different estimate of annual thickness, as can be seen in the 8-m section of the NASA-U core shown on Figure 1. This occurs in part because concentration peaks and troughs for each parameter occur at different times of the year, and in part because of multiple or missing peaks for some parameters in some years. Also, some parameters are better indicators in different parts of the core. For example, hydrogen peroxide is an excellent indicator down to about 60 m in the NASA-U core, but is less distinct due to diffusion in the next few tens of m. Deeper in the core it is again a good indicator, due in part to decay at points where dust is high. Nitrate is not a good indicator in recent decades due to anthropogenic influences, but does show distinct peaks deeper in the core. Hydrogen peroxide and delta-18 O are useful annual indicators only in the very top part of the Humboldt cores because of diffusional smoothing at this lower-accumulation site. The 1783 AD Laki volcanic eruption, located by E. Mosley-Thompson's sulfate analy ses, provides a reference date lower in the core.

Chemical analyses were done using a continuous melter. Resolution is about 0.02 m, and depends on dispersion in the system. Replicate analyses using the melter showed very good reproducibility (Figure 2).

Our overall annual layer thicknesses are based on a comparison of all of the species that show an annual signal. However, each parameter does give an independent estimate of layer thickness. We illustrate the difference between results using different parameters by comparing calcium and hydrogen peroxide (Figure 3). Taking 5-yr averages greatly reduces the considerable scatter in annual values, and the scatter is absent in the 25-yr averages.

Annual layer thicknesses during the current century and the periods of record at the NASA-U and Humboldt Glacier sites averaged 0.36 and 0.15 m ice, respectively (Figure 4). Cores from both sites showed a slightly increasing layer thicknesses during the last 300 years of about 1-3 percent per century.

Two 20-m firn cores drilled near the main 150-m NASA-U core showed that interannual variability of annual layer thickness exceeded spatial variability (Figure 5). One 20-m core was within 50 m of the main core, and the other was 2 km away. However, at the NASA-U site, four 20-m cores drilled 25 km from the 150-m core showed about equal spatial and temporal variability. That is, the spatial variability in the set of Humboldt cores is greater than in the set of NASA-U cores. Temporal variability is also slightly higher at Humboldt as compared to NASA-U.

The thicknesses at both sites showed a low frequency variation of about 100 yr, which can be seen more clearly on the 25-yr averaged data (Figure 6). The Humboldt core showed higher annual thicknesses rates around 1800 AD, possibly indicating a changed circulation pattern between 1760 to 1810 AD for the northern part of Greenland. However, the other site, NASA-U, located south-west from Humboldt, did not show such an anomaly, indicating that a changed circulation pattern influenced only the more northern part of the Greenland ice sheet. Also note the contrast between Humboldt and Summit (GISP2, Figure 6) during this time period. Data from the older Camp Century core (Figure 6) show some of this 1760-1810 anomaly, but it is not apparent in the older Milcent (Figure 6) core from Central Greenland.

The layer thicknesses at NASA-U and Summit are in phase 1650-1820; and then in phase again 1860 to present, but with an apparent 10-yr lag. Thicknesses at Humboldt and Summit are weakly in phase approximately 1840-1920, and more-strong\-ly in phase 1450-1650 (data not shown). Thus there are some periods when the part of the ice sheet represented by these cores may have experienced gains and losses over a large area, and other times when within-region variability offset overall ice sheet gains and losses.

Figure 7 shows the hydrogen peroxide concentrations for Crawford Point, which provide a clear delineation of annual layers. Using the NASA-U density values for this site gives an average layer thickness of 0.47 m ice per year. This is probably a low estimate, since density could be higher at Crawford Point than at NASA-U.

Figure 2. Replicate analyses on adjacent pieces cut from a 1-m section of NASA-U core showing reproducibility of melter analyses.

Figure 3. Comparison of annual layer thicknesses from NASA-U core 1 using calcium versus hydrogen peroxide. Upper panels are comparison for annual, 5-yr and 25-yr averages; lower panels show histograms of differences between thicknesses based on calcium versus hydrogen peroxide for the same data.

Figure 4. Annual layer thicknesses based on analysis of all parameters for NASA-U and all parameters except dust for Humboldt.

Figure 5. Histogram of spatial and temporal differences for 3 NASA-U cores (28 yr) and 5 Humboldt cores (64 yr). Spatial variability values calculated as difference between annual value for individual core and 3-core (NASA-U) or 5-core (Humboldt) mean for each year. Temporal variability values calculated as difference between annual value for individual core and 28-yr (NASA-U) or 64-yr (Humboldt) mean for that core. Horizontal lines on each bar graph show coefficient of variation (c.v.).

Figure 6. Comparison of 25-yr running averages of annual thicknesses for NASA-U and Humboldt cores, with other published data. GISP2 based on Meese et al., [1994], Camp Century and Milcent based on Claussen et al., [1988].

Figure 7. Hydrogen peroxide data for Crawford Point, showing 13 annual cycles. Vertical lines mark annual layers.

D. A. Meese, A. J. Gow, P. Grootes, P. A. Mayewski, M. Ram, M. Stuiver, K. C. Taylor, E. D. Waddington, and G. A. Zeilinski. The accumulation record from the GISP2 core as an indicator of climate change throughout the holocene. Science, 266:1,680-1,682, 1994.

H. B. Clausen, N. S. Gundestrup, S. J. Johnsen, R. Bindschadler, and J. Zwally. Glaciological Invetigations in the Crete Area, Ceneral Greenland: A Search for a New Deep-Drlling Site. Annals of Glaciology, 10, 1988.