R. C. Bales, J. F. Burkhart, J. R. McConnell, B. Snider

Dept. of Hydrology and Water Resources,

University of Arizona, Tucson, AZ 85721

Grant: NAG5-6779

The objective of the 1998 SHAllow COring (SHACO) field campaign was to elucidate questions regarding spatial variability of accumulation rates throughout the Greenland ice sheet. In the 1998 field season were primarily concerned with an accumulation trend along the central-western portion of the ice sheet. In addition, data were needed to fill a large gap of information in the central and southern portions of the ice sheet. In order to achieve a spatially extensive data collection program depth (temporal variation) was limited to 20-m so that a greater number of cores could be efficiently collected. Cores developed under PARCA use multi-species chemical analyses of ice cores to develop accurate estimates of annual accumulation at various sites. The ice core data are also analyzed for the interannual variability of snow-accumulation rates, which is a major cause of short-period variability of ice sheet elevation, and must be understood before we can infer long-term trends in ice sheet volume from observed surface elevation changes. This year's results will offer a spatially distributed chemical record for the purposes of developing a high-resolution accumulation map of the Greenland ice sheet.

We have also analyzed cores collected in 1996 at GITS (77.1ºN, 61.0ºW) and TUNU (78.1°N, 34.0°W). Average accumulation (still preliminary) is 0.36m water at GITS and 0.11m water at the main TUNU site (paper in preparation).

In 1997, shallow firn cores were collected at eight sites (Figure 1) around the perimeter of the Greenland ice sheet at approximately the 2500-m elevation contour. At those shallow coring sites that were co-located with automatic weather stations, more than one core was collected in order to investigate issues of short-scale spatial variability in snow accumulation. Average accumulation at two sites in northwest Greenland (75-76°N) had average annual accumulation values of 0.32-0.36 m water equivalent, and two sites in west central Greenland (71-72°N) had values of 0.40-0.42 m water equivalent. All four sites had values that were only 70-80% of that estimated from prior data and call into question the accumulation "ridge" in western Greenland that is apparent in older data. Figure 2 gives a comparison of preliminary accumulation values from our 1995-97 cores with values from the accumulation map done by Ohmura and Reeh [1991]. Note that at about half of our sites the standard deviation encompasses the Ohmura and Reeh [1991] interpolated value, and at about half their value lies outside our range.

Figure 1. Locations of PARCA ice cores 1995-1998

Figure 2. Comparison of recent and historical data (PARCA data preliminary)

Field Work: During the months of April and May, eighteen shallow cores were collected on the Greenland ice sheet. The locations of the cores were selected to cover the suspect accumulation "ridge" and to fill data voids in the central and southern portion of the ice sheet. The primary focus of the 1998 field season was to collect cores in the region from 60-65^o North and 35-45^o West (Figure 1). Cores were collected by 2-3 person teams using the re-designed "Sidewinder II" 4-inch core drill, developed by Jay Kyne of PICO.

The results from the 1997 field season demonstrated the effectiveness of "commuter" coring. The 1998 field season was planned using this flight mode. Commuter coring uses a Twin-Otter to its fullest extent. All flights departed from Kangerlussuaq fully loaded with coring equipment, fuel, and emergency gear. Core drilling, weighing, and packing takes no more than 4 hours at each site. Upon return the plane is once again full with the weight of the ice replacing the weight of the fuel. In general, cores up to 20-m were collected from two separate locations in a single day. A total of 233-m of core was collected with only ten flights. All cores from the 1998 field season were stored in aluminum bags in order to prevent contamination and allow better preservation of reactive chemical species.

Table 1. NASA 1998 Shallow Core Information, as of Sept. 30, 1998.

Laboratory Analyses: The shallow cores were sent from Kangerlussuaq, Greenland to the Byrd Polar Research Center at Ohio State University. In August the 4-inch cores were cut, and segments sent to the University of Arizona. Along with the SHACO cores, we are also analyzing a 20-m and a 120-m core from the Raven (66.38^oN, 46.18^oW) site that was collected by Ellen Mosely-Thompson of Ohio State.

From a 3.5x3.5 cm longitudinal cut, a 2-cm diameter core is continuously melted and analyzed using a flow-through analytical system. Our current analytes include Ca⁺², NO₃, NH₄, H₂O₂, and HCHO. Currently our system is configured such that the vertical resolution is 0.02 m, with some variation resulting from dispersion in the system.

At present we have cores that will provide accurate accumulation estimates for 29 locations on the ice sheet. Combined with the 36 points from the German traverse in North Greenland plus recent cores (e.g. GISP2), there are over 65 points with accurate accumulation estimates for the past 1-3 decades. From the Ohmura and Reeh compilation (about 250 points on the ice sheet), only about 50 points are based on a record that is 10 years or longer (Figure 3). To get an idea of the uncertainty associated with short records, we sampled different record lengths from the most recent 200 years of the NASA-U and Humboldt cores. For a record length of 1 year, the range is ±50-120% of the mean (Figure 4). For a record length of 10 years this drops to ±15-22%. The corresponding standard deviations for 1 and 10 years are ±25-30 and ±7%, respectively. For the PARCA cores we have aimed for record lengths of at least 10 years, and preferably at least 20 years. Note, however that the standard deviation drops off much more slowly after 10 years as compared to 20 years. It thus appears that the combined PARCA and German data, together with a subset of the historical data used by Ohmura and Reeh, should provide a much more accurate accumulation map that has been produced in the past.

Based just on the 1995-98 cores, our preliminary data call into question the accumulation ridge in the western part of the ice sheet that appears on the Ohmura and Reeh accumulation map. Note our means on Figure 2 versus Ohmura and Reeh values. Especially striking is the Northwest Greenland site with 0.28-m accumulation (Figure 2), which lies near a reported 0.63-m accumulation point on the 1912 Koch-Wegner traverse. That is, our preliminary data suggest that Ohmura and Reeh's map overestimates accumulation in this portion of the ice sheet by 20-30%. The result is 3-5% less accumulation on the ice sheet as a whole as compared to Omura and Reeh's map. Our plan is to work with the AWI group to develop an updated accumulation map that combines PARCA and AWI data, pending completion of analysis of the 1998 shallow cores this fall.

Figure 4. Mean and Standard Deviation information for NASA-U and Humboldt Cores

Our work involves a combination of 10-20 year cores to establish the recent accumulation on the ice sheet, and 200+ year cores to establish the long-term fluctuations. It is now apparent that gaps in the historical accumulation data compiled by Ohmura and Reeh (1991) are larger than has been assumed, bringing up the need to continue the shallow 10-20 year cores.

For cores of approximately 150-m depth, along with 20-m cores in the vicinity to estimate spatial variability, candidate locations are:

• South Greenland (1 core): South Dome or 63.8^oN, 45^oW
• West-central Greenland (1-2 cores): 70^oN, 45^oW and 72^oN, 46^oW vicinity and possibly 68.5^oN, 41^oW vicinity

Candidate locations for shallow (20-m) cores to fill in questions concerning accumulation high points are:

• Northwest Greenland (3-5 cores): 76.5^oN, 55^oW vicinity
• West Greenland (2-3 cores): 72^oN, 46^oW to 73^oN, 47^oW vicinity
• Southeast Greenland (1-3 cores): 70^oN, 35^oW vicinity
• South Greenland (1-3 cores): South Dome (63^oN, 45^oW) and Saddle (66^oN, 44.5^oW) vicinity

Final site selection will depend on the results from the in-progress analyses of the '98 cores.

--M. Anklin, R. C. Bales, E. Mosley-Thompson and K. Steffen. Annual accumulation at two sites in northwest Greenland during recent centuries. Journal of Geophysical Research, in press.

-- M. Anklin and R. C. Bales and E. Mosley-Thompson and K. Steffen. Variability of accumulation rates in two Greenland firn and ice cores using multiple independent parameters. Eos, Transactions, American Geophysical Union, 77:17:S150 (1996)
-- R. C. Bales. Comparison of accumulation trends in recent centuries in northwest Greenland. Eos, Transactions, American Geophysical Union, 78:46:F8 (1997)

-- R. C. Bales and J.R. McConnell. Reducing uncertainty in accumulation trends from shallow cores on the Greenland Ice Sheet. Eos, Transactions, American Geophysical Union, 79:17:S8 (1998)