Ice dynamics of the Haupapa/Tasman Glacier measured at high spatial and temporal resolution, Aoraki/Mount Cook, New Zealand

Abstract

Glaciers are among the clearest of signals for anthropogenic climate change and their retreat is considered symptomatic of the observed warming since the start of the 20th century from anthropogenic sources (Mann et al., 2004). New Zealand has 3,100 mountain glaciers, with those in the Southern Alps experiencing losses of 34% since 1977 and a decline in volume of 51 km3 in 1994 to 41 km3 in 2010 (NIWA, 2011). The direct impact of increasing atmospheric temperatures on glaciers is well understood (Chinn, 2012) through its effects on the melt and accumulation rates (Kirkbride, 2010; Purdie, 2011; Chinn, 1997; Oerlemans, 2001). However lake calving glaciers such as the Tasman Glacier exhibit different behaviour and are suggested to be at least partially decoupled from climate forcing (Benn et al., 2007).

Here, I present a temporally and spatially complete study of Haupapa/Tasman Glacier, Aoraki/Mt. Cook over three years to investigate the ice dynamics at the terminus. I used oblique photogrammetry at high resolution for data acquisition and adapted computer vision algorithms for correcting this oblique view to a real-world geometry. This technique has been rarely used (Murray et al., 2015; Messerli and Grinsted, 2015; Ahn and Box, 2010; Harrison et al., 1986 and Flotron, 1973) but owing to its cost-effectiveness and high data yields, it is becoming an increasingly powerful methodology favoured by glaciologists.

During the 3 year study period, Tasman Glacier terminus retreat rate Ur was 116 ± 19 m a⁻¹ (2013-2014), 83 ± 18 m a⁻¹ (2014-2015) and 204 ± 20 (2015-2016). A strong seasonal pattern was evident in the calving events. Three major calving events occurred over the study, one occurring in the summer of 2013 and two in the summer of 2016. The latter two events are responsible for the elevated Ur in 2015-2016. These events were characterised as distinct large-magnitude calving (usually as a large tabular iceberg) which continued to drift and break up in the lake for weeks to months. Three large calving events accounted for 47% of the total surface area loss for the 38 month study period with the remaining surface area loss from 2nd order calving including notching at the waterline and the spalling of lamallae of ice from surface fractures, and ice-cliff melt. During the spring/summer months of 2014 and 2015 there was no large buoyancy driven calving event such as those seen in 2013 and 2016, but there were many smaller-magnitude calving events. Smaller-magnitude events were less frequent in winter months as compared to summer months. Ice flow in winter has been shown to be less than in summer (Horgan et al, 2015). While seasonal temperatures and changes to the basal water pressure are linked to these observations, it is also likely that the relatively faster ice flow in summer/autumn could be influencing the rate of 1st and 2nd order calving mechanisms. Overall, the calving rates were calculated as 171 ± 18 m a⁻¹ (2013-2014), 136 ± 17 m a⁻¹ (2014-2015) and accelerated to 256 ± 20 m a⁻¹ in the last year (2015-2016). My results show that almost half of the ice loss at the terminus comes from large, infrequent calving events and that retreat rates for 2015-2016 were high compared to the historic record but the area loss is lower than it has been because of the relatively narrow terminus.