In order to calculate the retreat of the Laurentide Ice Sheet, I began by creating a Web App using ArcGIS in order to determine the distances between points. The map shown below currently displays all points.

The points are categorized by dating method and symbolized by years from present, with earlier years in red and later years in orange and yellow. Looking at the point distribution on both the map and the graphs, it is difficult to see a clear trend in glacial retreat. The older points should be closer to the bottom while the younger points should be closer to the top, symbolizing glacial retreat.

We see such a big difference between these methods partially because a majority of the southernmost points are thousands of years older than the southern points that were dated using bulk sediments or macrofossils. While there were some older points scattered throughout the bulk sediment and macrofossil data, they were farther north than younger, southern points and were not strongly considered for the calculations because they indicated that the glacier was located farther south later on. Additionally, many of the cosmogenic nuclide data points were 5,000-10,000 years older than macrofossil or bulk sediments points located at a similar latitude in the same area. This suggests that radiocarbon-based data may yield significantly different results than cosmogenic nuclide data. Looking at the northern data points, the cosmogenic nuclide data suggests that the extent of the retreat reached northern Vermont and New Hampshire more recently than what the radiocarbon data suggests. This would extend the number of years that it took for the ice sheet to retreat, which contributes to the significantly slower retreat rate seen in the cosmogenic data.

If the cosmogenic nuclide and radiocarbon data yield drastically different results, how do we know which method is more accurate? According to one study, bulk sediment ages may be inaccurate due to the recycling of old carbon in bulk sediments, and macrofossils in VT tend to be hundreds of years younger than bulk sediment ages (Corbett et al 2018). Cosmogenic nuclides, on the other hand, accumulate in known rates but use several assumptions including that no erosion occurred following the retreat of the glacier (Corbett et al 2018).

We also know that glaciation occurred at different rates, and these rates tended to be rapid at lower, warmer elevations where the retreat rate exceeded the advance rate (Corbett et al 2018). The rates calculated above were an average, and did not account the rapid retreat at lower elevations or the slower, more uncertain retreat rates at higher elevations. It is also important to consider the type of glacier in this analyis. The Laurentide Ice Sheet was likely a combination of cold and warm-based, being frozen to the substrate in some areas but not others. It is possible that the glacier was cold-based in higher elevations where more snow accumulated and warm-based in lower elevations where it gradually became warmer as the ice sheet was retreating.

Ultimately, it is most likely that the cosmogenic data was more accurate than the radiocarbon data. The retreat rate and spacing of the points sounds more probable and the timeline in general makes more sense. Additionally, the sea level rise graph from the Lambeck et al paper would have displayed different results if it were following the path of the radiocarbon data.