Area, elevation and mass changes of the two southernmost ice caps of the Canadian Arctic Archipelago between 1952 and 2014

Abstract

Abstract. Grinnell and Terra Nivea Ice Caps are located on the southern Baffin Island, Nunavut, in the Canadian Arctic Archipelago. These relatively small ice caps have received little attention compared to the much larger ice masses further north. Their evolution can, however, give valuable information about the impact of the recent Arctic warming at lower latitudes (i.e. ~ 62.5° N). In this paper, we measure or estimate historical and recent changes of area, elevation and mass of both ice caps using in situ, airborne and spaceborne data sets, including imagery from the Pléiades satellites. The area of Terra Nivea Ice Cap has decreased by 34 % since the late 1950s, while that of Grinnell Ice Cap has decreased by 20 % since 1952. For both ice caps, the areal reduction accelerated at the beginning of the 21st century. The estimated glacier-wide mass balance was −0.37 ± 0.21 m a−1 water equivalent (w.e.) over Grinnell Ice Cap for the 1952–2014 period, and −0.47 ± 0.16 m a−1 w.e. over Terra Nivea Ice Cap for the 1958/59–2014 period. Terra Nivea Ice Cap has experienced an accelerated rate of mass loss of −1.77 ± 0.36 m a−1 w.e. between 2007 and 2014. This rate is 5.9 times as negative when compared to the 1958/59–2007 period (−0.30 ± 0.19 m a−1 w.e.) and 2 times as negative when compared to the mass balance of other glaciers in the southern parts of Baffin Island over the 2003–2009 period. A similar acceleration in mass loss is suspected for the Grinnell Ice Cap, given the calculated elevation changes and the proximity to Terra Nivea Ice Cap. The recent increase in mass loss rates for these two ice caps is linked to a strong near-surface regional warming and a lengthening of the melt season into the autumn that may be indirectly strengthened by a later freezing of sea ice in the Hudson Strait sector. On a methodological level, our study illustrates the strong potential of Pléiades satellite data to unlock the under-exploited archive of old aerial photographs.

Recent changes in area and thickness of Torngat Mountain glaciers (northern Labrador, Canada)

“…The large thinning rates measured at Abraham, Hidden, and Minaret glaciers be- (1981, 1982, and 1983: −0.24, −0.21, and −0.81 m yr −1 w.e., respectively) and is the most negative balance yet recorded at this or any other glacier in Labrador. This strongly negative post-2005 mass balance is similar in magnitude to that measured at nearby Terra Nivea ice cap on southern Baffin Island (Papasodoro et al, 2015). While these 6 years of discontinuous measurements represent the only geodetic mass balance observations yet recorded at Labrador glaciers, it is important to note that they may mask considerable interannual variability in accumulation and ablation at the sampled glaciers and within the wider regional population.…”

“…Total volume changes were then calculated by multiplying δh by the pixel area and summing across the glacier surface (e.g. Barrand et al, 2009;Nuth et al, 2010). Volume changes were converted to geodetic mass balances in water equivalent units (m yr −1 w.e.)…”

“…Individual centreline elevation changes (δh) were averaged into 10 m elevation bins and applied to all pixels in each bin across the glacier surface (e.g. Nuth et al, 2010). In the event that no δh measurements were available (due to safe-access issues during surveying or culling of low-accuracy points; see Sect.…”

“…Total elevation change errors, E Z , were then calculated by combining E PT(δh) in quadrature with extrapolation errors (E EXT ), reducing by the square root of the total number of independent measurements (e.g. Arendt et al, 2006;Nuth et al, 2010), considered to be one measurement per 10 m elevation bin, given the presence of weakly positive spatial autocorrelation. E EXT quantifies the uncertainty in applying centreline δh measurements to entire elevation bins.…”

“…E EXT quantifies the uncertainty in applying centreline δh measurements to entire elevation bins. Following Nuth et al (2010), E EXT was approximated by the standard deviation of the averaged δh in each elevation bin. Where bins had fewer than five δh measurements, E EXT was set to twice the glacier average, while bins with extrapolated measurements were assigned 3 times the average.…”