Geochemical signatures of pingo ice and its origin in Grøndalen, west Spitsbergen

Abstract

Pingos are common features in permafrost regions that form by subsurface massive-ice aggradation and create hill-like landforms. Pingos on Spitsbergen have been previously studied to explore their structure, formation timing and connection to springs as well as their role in postglacial landform evolution. However, detailed hydrochemical and stableisotope studies of massive-ice samples recovered by drilling have yet to be used to study the origin and freezing conditions in pingos. Our core record of 20.7 m thick massive pingo ice from Grøndalen is differentiated into four units: two characterised by decreasing δ 18 O and δD and increasing d (units I and III) and two others showing the opposite trend (units II and IV). These delineate changes between episodes of closed-system freezing with only slight recharge inversions of the water reservoir and more complicated episodes of groundwater freezing under semi-closed conditions when the reservoir was recharged. The water source for pingo formation shows similarity to spring water data from the valley with prevalent Na + and HCO − 3 ions. The sub-permafrost groundwater originates from subglacial meltwater that most probably followed the fault structures of Grøndalen and Bøhmdalen. The presence of permafrost below the pingo ice body suggests that the talik is frozen, and the water supply and pingo growth are terminated. The maximum thaw depth of the active layer reaching the top of the massive ice leads to its successive melt with crater development and makes the pingo extremely sensitive to further warming.

Numerical modelling of permafrost spring discharge and open-system pingo formation induced by basal permafrost aggradation

“…Generically though, they relate more to closed-system pingos, because the causal mechanism of hydraulic pressures is essentially similar, although operating over much longer time-scales. We emphasise that this conceptual model for pingo formation represents an end-member of pingo-forming processes, which is not exclusive, but may act in combination with others, such as those reported nearby by Demidov et al (2019).…”

“…by Demidov et al, 2019), we would not expect to observe such a systematic trend along the valley axis, as this would be unlikely in a system dependent upon localised areas for infiltration. We therefore suggest that the unexpected landwards increase in Clis difficult to explain without the sequence of events illustrated in Fig.…”

“…The hydrogeological mechanisms causing the sustained flow of sub-permafrost groundwater to surface springs remain elusive (Scheidegger et al, 2012). Earlier it was proposed that subglacial meltwater from underneath warm-based ice sheets or glaciers would sufficiently recharge a sub-permafrost aquifer (Demidov et al, 2019;Liestøl, 1977;Scheidegger et al, 2012;Scheidegger and Bense, 2014). However, in regions of continuous permafrost lacking warm-based glaciers or other groundwater recharge pathways, such models do not seem applicable (Ballantyne, 2018;Grasby et al, 2014;Woo, 2012).…”

Numerical modelling of permafrost spring discharge and open-system pingo formation induced by basal permafrost aggradation

“…For S s , we used a matrix compressibility of 7 × 10 −10 to 7 × 10 −8 Pa −1 (based on common estimates for fractured rocks; e.g. Domenico and Mifflin, 1965;Domenico and Schwartz, 1998;Fitts, 2002) yielding a S s of 7×10 −6 to 7×10 −4 m −1 (in line with literature values; cf. Singhal and Gupta, 2010).…”

“…The hydrogeological mechanisms causing the sustained flow of sub-permafrost groundwater to surface springs remain elusive (Scheidegger et al, 2012). Earlier it was proposed that subglacial meltwater from underneath warmbased ice sheets or glaciers would sufficiently recharge a sub-permafrost aquifer (Demidov et al, 2019;Liestøl, 1977;Scheidegger et al, 2012;Scheidegger and Bense, 2014). However, in regions of continuous permafrost lacking warmbased glaciers or other groundwater recharge pathways, such models do not seem applicable (Ballantyne, 2018;Grasby et al, 2014;Woo, 2012).…”