Embedded Files
Photo: Ice Waterfall. Svalbard, Norway. 2014. © Paul Nicklen Photography, Inc.
The purpose of this project is to analyze data variations over multiple deglaciation dating techniques
Figure 1: Laurentide Ice Sheet Retreat modeled by Radiocarbon dating of bulk sediments and macrofossils, and Cosmogenic Nucleotide Dating of terminal moraines via Be-10 in quartz. Data further seperated by sources, Halstead, Corbett.
As an ice sheet retreats, it leaves behind clues as to when it passed. Carbon is left in plant and animal macrofossils, or is incorporated into lake sediment deposits. These samples can be dated by measuring the concentration of the C-14 isotope. Error ranges from the chemical distribution of the sample, but also from lag time delays in the appearance of macrofossils in the post-glacial environment, and from carbon contamination or carbon recycling in bulk sediments. Measuring the ratio of stable to unstable isotopes created from cosmogenic nuclides will also yield an exposure date of a rock sample. Error here ranges from erosion and shielding (under estimate) to undecayed prior exposure isotopes (over estimate).
Using data provided by Halstead and Corbett, Figure 1 plots Radiocarbon dating of bulk sediments and radiocarbon dating of macrofossils along both the furthest south and north terminal moraines of the Laurentide Ice Sheet (LIS) in New England, as well as Be-10 cosmogenic nuclide dating of many similar points. Spreadsheet of Laurentide Ice Sheet Retreat Data and Calculations
Figure 2. Sites of sample dating, labeled with respective date, color coded by method.
Figure 3: Inverted map of LIS retreat data for simple linear regression (Date BP = -0964.7(°L)+55991
Figure 4: Inverted map of LIS retreat data for outlier analysis on two Cosmogenic terminal moraine dates. (Date BP = -716.3(°L)+44983
Figure 2 maps these sample sites by date on a map of New England. From Figures 1 &2, it is clear there is a discrepancy between methods as to the exact date, and therefore rates of retreat.
Measuring the distance between sites along three transects (Hudson-Champlain, Conneticuit River, and Cape Cod) allows for 1st approximation rate of retreat estimates in these locations, as detailed in the "Lat Long Retreat Calc" sheet in Spreadsheet of Laurentide Ice Sheet Retreat Data and Calculations. Further, as the LIS retreated mostly north since the Last Glacial Maximum (LGM), a secondary set of calculations was performed between sites using (1° Latitude ~ 110567m). Averaging the two rates by dating technique found 127.78 m/yr by radiocarbon bulk sediments, 100.4386 m/yr by radiocarbon macrofossils, and 55.3652 m/yr by cosmogenic nuclides (excluding the very slow retreat around Martha's Vinyard/Cape Cod, which is addressed later), or a total average of 94.53 m/yr.
This discrepancy is caused by our low amount of data points for cosmogenics, but also by two data that are out of agreement with other measurements nearby.
To analyze the these outliers and check the sensitivity of calculated retreat rate, I first inverted my axes before performing a simple linear regression, both with and without these potential outliers (see Figures 3&4).
Correlation with the data is low, considering only two primary latitudes were sampled for analysis. This is further compounded by the generally faster retreat further away from the ocean (as seen in Spreadsheet Calculations) however, we are looking for sensitivity of a rate for these outliers. So we continue.
Using the regression equations and solving for °Lat, and applying the conversion 1°Lat ~ 110567 m, we find that when outliers are excluded, we see an average retreat rate of -154.352 m/yr, while with them included we find -114.658 m/yr. This 35% difference by two points is large, and may need further consideration.
Looking further, Lambeck (2014) shows 120m SLE over 9.5 k yrs, Or ~12.63 mm/yr (Figure 5).
Given the current measure of 362.5 Gt of ice loss per 1 mm sea level rise, we find ourselves with -4350 gt/yr
The LIS is generally thought to have contributed some 70 m of this 120m SLE (70/120).
Dr. Michaela King found that total Greenland Ice Sheet discharge related to cumulative km retreat of outlet glaciers on the island Total Discharge = 14 Gt (total km) + 312 Gt (King, 2020). Using this relation as a proxy, we find 288.429 km retreat for total glaciers per perimeter. If the LIS lost mass on its main front from Great Lakes to New York to Nova Scotia we can estimate this perimeter to about 3000 km yielding ~96.1429 m average annual retreat per outlet glacier.
Which has a 1.7% difference from my calculated mean that included all three techniques for dating. This leads me to believe the radio carbon numbers are a tad young and show a preferentially fast retreat. Cosmogenic data seems a bit old showing slower retreat. This could be rationalized by fluctuations in the ice during retreat from acceleration and thinning, the cosmogenic sample sites all near the coastal margin, or from inherited isotopes from the previous interglacial period.
Figure 5: Sea Level Rise Result modified from Lambeck (2014)
Page updated
Google Sites
Report abuse