Palaeo-environmental significance of evaporative calcite crusts in the Untersee Oasis, East Antarctica

Abstract

Secondary carbonate precipitates on the surface of clasts have rarely been reported from Antarctica. Here, we infer the origin, age and palaeo-environmental significance of the calcite crusts in the Untersee Oasis, East Antarctica. The distribution of calcite crusts, which are up to 1 mm thick, is limited to locations with residual snow patches, and they have some of the highest δ¹⁸O (up to +17.4‰ Vienna Standard Mean Ocean Water (VSMOW)) and δ¹³C (up to +14.6‰ Vienna Pee Dee Belemnite (VPDB)) compositions of any carbonate deposits in terrestrial polar environments. Their δ¹⁸O and δ¹³C values are substantially enriched with respect to the isotopic values expected from equilibrium precipitation from the δ¹⁸O and δ¹³C_DIC (DIC = dissolved inorganic carbon) of snow meltwater. The formation of the calcite crusts is ascribed to the evaporation of residual snow meltwater and the low relative humidity and strong winds, favouring a kinetic isotope effect. The ¹⁴C age distribution of the calcite crusts (1550 cal yr bp to modern) provides a minimum age for ice retreat and drainage of the palaeo-lake in Aurkjosen Cirque. However, in this polar desert environment in which surface melting is limited, the calcite crusts require sufficient snow accumulation and air temperatures warm enough to generate meltwater, and their age distribution corresponds to the late Holocene warm-wet climate period.

Introduction

Secondary carbonate deposits in terrestrial polar regions are common in limestone environments. They typically occur on the surface or underside of clasts (Hanshaw & Hallet 1978, Hillaire-Marcel et al. 1979, Souchez & Lemmens 1985, Courty et al. 1994), within fissures in bedrock outcrops (Lauriol & Clark 1999) and on the surface of ice in caves and icings (Lauriol et al. 1991, Clark & Lauriol 1997, Omelon et al. 2001, Lacelle et al. 2006). They are less prominent in non-limestone settings, but they have still been observed in crystalline and granitic bedrock environments on the surface of clasts (Blake 1999, Lacelle 2007). If the carbonate deposits precipitate under equilibrium conditions, their δ¹⁸O composition can provide palaeo-climatic information (e.g. Clark et al. 2004, Lacelle 2007); alternatively, if they precipitated from waters in equilibrium with atmospheric CO₂, they can be used as palaeo-environmental proxies. For example, the ¹⁴C ages of carbonate crusts on the surface of moraines were used to infer the timing of deglaciation in east-central Ellesmere Island (high Arctic Canada; Blake 2005), southern Baffin Island (low Arctic Canada; Lacelle et al. 2007) and northern Pakistan (Waragai 2005), and the chronologies matched well with those derived from other methods.

Reports of carbonate precipitates on clasts or in soils have been spatially limited in Antarctica. In the McMurdo Dry Valleys, paedogenic carbonates were observed in the soils of multiple valleys (Claridge & Campbell 1977, Campbell & Claridge 2009, Campbell et al. 2013, Lyons et al. 2020), in soils near Shackleton Glacier (Diaz et al. 2020) and in carbonate crusts near the shore of Lake Vanda (Nakai et al. 1975). The paedogenic calcites were used to inform about chemical weathering (i.e. sources of calcium and carbon in the soils). Here, we report the discovery of carbonate crusts on the surface of clasts in the polar desert of the Untersee Oasis in East Antarctica. The carbonate crusts occur on morainic material around Lake Untersee and in Aurkjosen Cirque, including within a former lake basin. This study aims to determine the origin and age of the carbonate crusts in the Untersee Oasis to provide information regarding the timing of glacial retreat, drainage of a proglacial lake and climate conditions that would allow for their formation. This objective is reached by determining the mineralogy of the carbonates, their δ¹⁸O and δ¹³C composition to infer their process of formation and the ¹⁴C ages to infer the timing of their growth. Residual snow patches were also sampled and analysed for major ions to determine their saturation indices with respect to carbonate minerals and δ¹⁸O to assess whether the carbonates precipitated under equilibrium or kinetic conditions.

Study area

Untersee Oasis (71.3°S, 13.5°E) is located within the Gruber Mountains of Queen Maud Land, ~150 km from the coast and ~90 km south-east of the Schirmacher Oasis (Fig. 1). The Untersee Oasis includes three main ice-free regions: 1) Lake Untersee valley, an 11 km-long and 4 km-wide north-south trending valley, 2) Aurkjosen Cirque, a ~3.5 km-long and 2 km-wide east-west trending valley and 3) Pritzker Valley (informal name), a 2.5 km-long a 0.5-km wide north-east-south-west trending valley with a small ice patch near the head of the valley.

Figure 1. Map of the Untersee Oasis in Queen Maud Land, East Antarctica. (Top) The locations of sampling sites of the calcite crusts and inferred glacial limits (from Schwab 1998). Ages of the calcite crusts are provided as cal yr bp. (Bottom) Recessional glacial limits in the Untersee Oasis and palaeo-lake basin in Aurkjosen Cirque (from Schwab 1998).

The East Antarctic Ice Sheet covered the Untersee Oasis during the Late Pleistocene. Based on the ¹⁴C ages of sub-fossilized stomach oils from snow petrel nests, thinning of the ice sheet began at c. 35–30 kyr bp (40 160–34 450 cal yr bp; Hiller et al. 1988, 1995), which led to a reconfiguration of the local ice flow. According to Schwab (1998), proglacial Lake Untersee probably formed between 13 and 11 kyr bp (15 560–13 000 cal yr bp) when the Anuchin Glacier started to retreat northwards. During its retreat, a proglacial lake also occupied Aurkjosen Cirque, as evidenced by palaeo-shorelines. That lake drained and/or evaporated to dryness once Anuchin Glacier receded from that valley c. 7 kyr bp (7800 cal yr bp) along a meltwater channel into Lake Untersee (i.e. the current lake ice ablation rate is 0.4–0.6 m yr^-1; Faucher et al. 2019). The Untersee Oasis probably came to its current configuration at c. 6–4 kyr bp.

The local geology in the Oasis consists of Precambrian norite, anorthosite and anorthosite-norite alternation of the Eliseev massif complex (Kampf & Stakebrandt 1985, Bormann et al. 1986, Paech & Stackebrandt 1995). There are no units of carbonate rocks in the area. The surface sediments consist mainly of till and colluvium, often covered by a thin layer of aeolian sediments (Schwab 1998). Vegetation and lichens are absent, and the soils consist of poorly sorted sediments with very low organic content (Shamilishvili et al. 2020).

The region is characterized by a polar desert climate regime. Ten years of climate data (2008–2017) collected by an automated weather station along the shore of Lake Untersee (71.34°S, 13.45°E, 612 m above sea level) shows a mean annual air temperature (MAAT) of -9.5 ± 0.7°C, thawing degree-days ranging from 7 to 51 degree-days and a mean relative humidity of 42 ± 5% (Andersen et al. 2015, Faucher et al. 2019). Despite having a relatively warm MAAT for Antarctica, the climate in the Oasis is dominated by intense ablation, which limits surface melt features (no surface runoff) due to cooling associated with latent heat of sublimation and vaporization (e.g. van den Broeke et al. 2006; Hoffman et al. 2008).

Methods

White carbonate crusts were collected in November–December of 2015 and 2017 from the surface of clasts within the morainic material surrounding Lake Untersee and in Aurkjosen Cirque; in the latter, crusts were also collected on the surface of clasts within the palaeo-lake basin (Figs 1 & 2). These crusts are all found in the immediate vicinity of residual snow patches and are different from the carbonate-poor greyish lithificates that occur only along the palaeo-shorelines of Lake Untersee (Levitan et al. 2012). The carbonate crusts, up to 1 mm thick, can be found on clasts up to 1 m above the surface of the soils (probably reflecting the height of residual snow patches), and they were collected with a chisel and hammer and placed in sealed Falcon tubes.

Figure 2. Field photographs of calcite crusts in the Untersee Oasis, East Antarctica. The calcite crusts are up to 1 mm thick.

The mineral compositions of five carbonate crusts were determined in the field using Raman spectroscopy. The excitation source was a frequency-doubled, Q-switched neodymium-doped yttrium aluminium garnet (Nd:YAG) pulsed laser source, and the spectra were analysed and recorded with a Kaiser Optical System fitted with an intensified charged-coupled device camera.

The δ¹⁸O and δ¹³C compositions of 17 carbonate crusts were determined to infer their origin (process of formation). The measurements of ¹⁸O/¹⁶O and ¹³C/¹²C ratios were made on CO₂ gas produced by reacting the powdered crusts with 100% phosphoric acid in glass septum vials for 24 h at 25°C. The evolved CO₂ gas was analysed in continuous flow using a Gas Bench II interfaced with a Finnigan Mat Delta+ XP isotope mass spectrometer at the Jan Veizer Laboratory (University of Ottawa, Canada). Stable isotope data for O and C were expressed in δ-notation, where δ represents the part per thousand difference of ¹⁸O/¹⁶O and ¹³C/¹²C in the sample with respect to the Vienna Pee Dee Belemnite (VPDB) standard (Kim et al. 2015). To facilitate comparison with the residual snow patches, the δ¹⁸O composition of the carbonate crusts were converted to the Vienna Standard Mean Ocean Water (VSMOW) scale (δ¹⁸O_VSMOW = 1.0309 δ¹⁸O_VPDB + 30.92; Coplen et al. 2002). Analytical reproducibility is 0.15‰ for both isotopes.

Radiocarbon measurements of 13 carbonate crusts were made to infer their carbon source and ages. The measurements were made at the Andre Lalonde AMS Laboratory (University of Ottawa, Canada) using a 3MV tandem accelerator mass spectrometer (AMS). The ¹⁴C/¹²C ratios are expressed as fraction of modern carbon (F¹⁴C) and corrected for spectrometer and preparation fractionation using the AMS-measured ¹³C/¹²C ratio (Crann et al. 2017). Radiocarbon ages are calculated as -8033ln(F¹⁴C) and reported in ¹⁴C yr bp (bp = ce 1950; Stuiver & Polach 1977), and the ¹⁴C ages were calibrated to cal yr bp using OxCal and the IntCal20 calibration curve (Bronk Ramsey 2009, Reimer et al. 2020).

Samples of residual snow patches in the vicinity of the carbonate crusts were also collected in November–December 2015 and 2017 to determine their geochemical and δ¹⁸O compositions. The concentrations of the major cations (Ca²⁺, K⁺, Mg²⁺, Na⁺) and anions (SO₄^2-, Cl^-, NO₃^-) of the filtered snow meltwater were measured using an Agilent 4200 inductively coupled plasma atomic emission spectrometer and ion chromatography (Dionex ICS-2100), respectively. Prior to the analysis of cations, water samples were acidified to pH 2 with ultra-pure HNO₃ acid. Analytical reproducibility for solute analysis is ±5%. The ¹⁸O/¹⁶O and D/H ratios of the meltwater were determined using a Los Gatos Research liquid water analyser coupled to a CTC LC-PAL autosampler for simultaneous ¹⁸O/¹⁶O and D/H ratios measurements of H₂O and verified for spectral interference contamination. The results are presented using the δ-notation (δ¹⁸O and δD), where δ represents the parts per thousand differences for ¹⁸O/¹⁶O or D/H in a sample with respect to VSMOW. The analytical reproducibility values for δ¹⁸O and δD are ±0.3‰ and ±1‰, respectively.

Results

Based on Raman spectroscopy and the sharp Raman band at 1088 cm^-1, the carbonate crusts in the Untersee Oasis consist of calcite minerals (Fig. 3). The δ¹⁸O and δ¹³C of the majority calcite crusts range from +10.8 to +17.4‰ and +9.6 to +14.6‰, respectively; however, two of the 17 calcite crusts had negative δ¹⁸O and δ¹³C values (Fig. 4a & Table I). The calcite crusts have median bulk ¹⁴C ages in the 1589 cal yr bp to modern (2011 cal yr ce) range (Table II). The calcite crusts on the peninsula were dated from 1589 to 545 cal yr bp, those in Aurkjosen Cirque were dated from 683 cal yr bp to modern and those above the shoreline of Lake Untersee were the youngest (119 cal yr bp to modern; Fig. 1).

Figure 3. Raman spectra of five carbonate crusts in the Untersee Oasis, East Antarctica. The intense sharp Raman band at 1088 cm^-1 is indicative of the mineral calcite.

Figure 4. δ¹³C and δ¹⁸O compositions of the calcite crusts in the Untersee Oasis, East Antarctica. a. δ¹³C and δ¹⁸O compositions of the calcite crusts in the Untersee Oasis compared to others in Antarctica: lithificates along palaeo-shorelines of Lake Untersee (Levitan et al. 2012), evaporative calcite crusts surrounding Lake Vanda (Nakai et al. 1975) and paedogenic calcite in soils surrounding Shackleton Glacier (Diaz et al. 2020). b. Comparison of δ¹³C and δ¹⁸O compositions of the calcite crusts in the Untersee Oasis (red box) with other terrestrial polar environments (from Lacelle 2007). VPDB = Vienna Pee Dee Belemnite; VSMOW = Vienna Standard Mean Ocean Water.

Table I. δ¹³C and δ¹⁸O measurements of calcite crusts in the Untersee Oasis, East Antarctica. The δ¹⁸O composition of the carbonate crusts were converted to the Vienna Standard Mean Ocean Water (VSMOW) scale according to: δ¹⁸O_VSMOW = 1.0309 δ¹⁸O_VPDB + 30.92 (Coplen et al. 2002).

VPDB = Vienna Pee Dee Belemnite.

Table II. Radiocarbon results from calcite crusts collected in the Untersee Oasis, East Antarctica. The ¹⁴C ages were calibrated using OxCal and the IntCal20 calibration curve (Bronk Ramsey 2009, Reimer et al. 2020).

F¹⁴C = fraction of modern carbon.

The snow meltwater of residual snow patches has a Ca-Na-SO₄ geochemical facies with [Ca²⁺] = 6.0 mg l^-1, [Na⁺] = 5.0 mg l^-1 and [SO₄^2-] = 9.1 mg l^-1. This geochemical facies is similar to that of regional snow (Isaksson et al. 1996, Marsh et al. 2020) and of local ponds recharged by snowmelt (Faucher et al. 2021). Based on the Phreeqc hydrogeochemical program, and assuming that the meltwater is in equilibrium with atmospheric CO₂, the meltwater is undersaturated with respect to calcite (SIcal = -4.5). The snow meltwater has δ¹⁸O values averaging -33.5 ± 2.4‰.

Discussion

Origin of the calcite crusts

All except two of the calcite crusts collected from the surface of clasts in the Untersee Oasis have some of the highest reported δ¹⁸O and δ¹³C values of any type of calcite precipitates found in terrestrial polar environments (Fig. 4b). The δ¹⁸O and δ¹³C values are much higher than the carbonate-poor lithificates observed along the palaeo-shorelines of Lake Untersee (δ¹⁸Oavg = +4.4‰, δ¹³Cavg = -3.3‰; Levitan et al. 2012). However, they are in the range of the paedogenic carbonates in the McMurdo Dry Valleys (δ¹³C in the +1.3 to +11.0‰ range) and the crusts on the dried lakebed in the basin of Lake Vanda (δ¹³C values in the +7.8 to +17.6‰ range; Fig. 4b).

Under most circumstances, the δ¹⁸O and δ¹³C of calcites can be estimated from the δ¹⁸O and δ¹³C_DIC (DIC = dissolved inorganic carbon) of the water from which the calcite precipitated and the temperature at which the precipitation occurred (Lacelle 2007). The calcite crusts in the Untersee Oasis probably precipitated at a temperature near 0°C (maximum hourly summer air temperatures are 9°C). Under equilibrium conditions (ɛCaCO₃-H₂O = 33.6‰ at 0°C; Kim & O'Neil 1997), the δ¹⁸O values of the calcite crusts (+10.8 to +17.4‰ range) are enriched by 10–17‰ relative to the average δ¹⁸O composition of snow meltwater (–33.5 ± 2.4‰). Although the δ¹³C_DIC of snow meltwater was not analysed, it can be assumed that its composition in a plagioclase setting would be in equilibrium with atmospheric CO₂. Over the past 2000 years, the concentration of CO₂ increased from 270 to 410 ppm, and the δ¹³C of CO₂ evolved from -6 to -8‰ (Schmitt et al. 2012, Eggleston et al. 2016). Under equilibrium conditions, the pH of snow meltwater for these conditions for CO₂ would be 6.4–6.6, with a δ¹³C_DIC of -2.8 to -1.4‰ (εCO_2aq-CO_2g = -1.2‰; εHCO₃-CO_2aq = 10.9‰). This would produce calcite with δ¹³C of +0.8 to +2.1‰ (εCaCO₃-HCO₃ = 3.6‰), which is much lower than the δ¹³C value measured from the calcite crusts in the Untersee Oasis (+9.6 to +14.6‰).

There are only a few processes capable of generating calcite precipitates with enriched δ¹⁸O and δ¹³C compositions over those of the parent water in polar environments: 1) kinetic freezing or 2) kinetic evaporation. According to Clark & Lauriol (1992), kinetic freezing imparts an enrichment in the order of 31.2 ± 3.1‰ for δ¹³C (ε¹³C_KIE CaCO₃-CO₂; KIE = kinetic isotope enrichment) and of 36.7 ± 1.3‰ for δ¹⁸O (ε¹⁸O_KIE CaCO₃-H₂O). However, kinetic freezing has only been observed to develop cryogenic calcite powders, as observed in caves (Clark & Lauriol 1992) and icings (Lauriol et al. 1991, Lacelle et al. 2006). As a result, this process can be ruled out because in the Untersee Oasis the calcites developed crusts on the surface of clasts and not powders.

Kinetic evaporation is the only process capable of explaining the enrichment in both δ¹⁸O and δ¹³C in the calcite crusts in the Untersee Oasis. The effect of evaporation on the δ¹⁸O of the residual snow meltwater and the calcite saturation index in the Untersee Oasis was modelled using the IsoVap7 computer code (appendix A in Fisher et al. 2020). The open system aqueous solution evaporation model uses the Freezchem database and accounts for isotopic exchanges with the ambient moisture based on the model of Sofer & Gat (1975) that includes isotope salt effects. The model with nearly pure water chemistry is similar to that described in Criss (1999) and has been used in Antarctic studies (e.g. Lapalme et al. 2017, Faucher et al. 2021). Despite the relative humidity in the air being 42 ± 5%, the relative humidity at the ground surface is consistently higher than in the air (often > 70%; D. Lacelle et al., unpublished data). This is consistent with studies from the McMurdo Dry Valleys that reported ground surface relative humidities higher than in the air (e.g. Fisher et al. 2016, McKay et al. 2019, Marinova et al. 2022). In all numerical scenarios (Fig. 5), evaporation at ground surface relative humidity < 70% caused strong δ¹⁸O enrichment in the residual water, with enrichment in the order of 30–40‰ when most of the water had evaporated. However, for ground surface relative humidity > 85%, the equilibrium exchange between snowmelt and atmospheric moisture approached a steady-state value that is dependent on the δ¹⁸O of the moisture and relative humidity. If we assume that the snow meltwater from which the calcite crusts formed had initial δ¹⁸O near -33‰ and the δ¹⁸O water vapour was in the -44 to -37‰ range, then at least 25% of the meltwater must evaporate to reach the values measured in the calcites (Fig. 5). During evaporation of snowmelt in equilibrium with P _CO2 of -3.5, calcite saturation also progressively increases as the solutes are concentrated in the residual water. Given the chemistry of the snow, calcite saturation is expected when 85% of the water has evaporated. As such, it is more likely that ground surface relative humidity was > 70% during the evaporation of snowmelt (Fig. 5). This would allow for the δ¹⁸O of the residual snow meltwater to reach a steady-state value, whereas continued evaporation would allow the solution to reach calcite saturation. For the effect of evaporation on the δ¹³C composition of calcites, caliche that developed on the surface of basalts in the Arizona volcanic field had δ¹³C values in the 4–12‰ range, and Knauth et al. (2003) proposed that their elevated values could be produced from the degassing of CO₂ and an unknown isotope effect during evaporation. Subsequently, Lacelle et al. (2007) experimentally measured the ¹³C enrichment during kinetic evaporation. It was found that the kinetic isotope effect between CO₂ and calcite ranged between 20 and 40‰ (with the kinetic isotope effect between HCO₃ and calcite ranging between 11.7 and 32.1‰). If we assume that the δ¹³C_DIC of snow meltwater was in equilibrium with atmospheric CO₂ (δ¹³C_DIC in the -2.8 to -1.4‰ range), then the elevated δ¹³C of the calcite crusts can be explained by kinetic evaporation once ~85% of the snow meltwater has evaporated.

Figure 5. Evolution of δ¹⁸O of evaporating water (snow meltwater chemistry) for a range of relative humidities (RHs) using the Criss (1999) and Sofer & Gat (1975) models. The initial δ¹⁸O composition of snow was set at -33‰ (average of the measurements) and that of the atmospheric moisture above the evaporating surface was set at -44‰ (equilibrium with meltwater) and -37‰, and the RH at the ground surface ranged from 50 to 95% (values near 50% represent atmospheric humidity, those > 70% probably represent those at the ground surface; i.e. Fisher et al. 2016). VSMOW = Vienna Standard Mean Ocean Water.

In the Untersee Oasis, the calcite crusts are present on the surface of clasts (gravels to large boulders) in the immediate vicinity of residual snow patches within morainic material (Fig. 2). There, a portion of the residual snow patches can melt during the summer and interact with the plagioclase rocks, and, as it wets the surfaces of clasts, the meltwater can precipitate calcite following evaporation. The snow patches contain little dissolved organic carbon (DOC) that could increase P _CO2 and decrease the δ¹³C_DIC in the meltwater (DOC concentration in nearby ponds recharged by snowmelt range from < 0.3 to 1 ppm; Faucher et al. 2021). The Untersee Oasis, with its low relative humidity and strong winds, is dominated by intense ablation, and these conditions would favour an increase in the rate of evaporation and create a kinetic isotope effect during the precipitation of the calcite crusts. This effect is caused by the rapid degassing during evaporation of a thin film of meltwater, such that there is preferential outgassing of CO₂ as Ca and HCO₃ concentrate prior to and during calcite precipitation. These are different from the lithificates found solely below the palaeo-shorelines of Lake Untersee (Levitan et al. 2012) and from the subglacially precipitated calcites found in formerly glaciated regions (e.g. Hillaire-Marcel et al. 1979, Souchez & Lemmens 1985). The latter tend to develop on the leeside of bedrock obstacles following regelation, are often striated and tend to be in isotopic equilibrium with the δ¹⁸O and δ¹³C_DIC of the basal meltwater from which they precipitated.

Palaeo-environmental significance

Based on the ¹⁴C ages of sub-fossilized stomach oils from snow petrel nests, thinning of the ice sheet in the Untersee Oasis region began at c. 40–35 cal yr bp (Hiller et al. 1988, 1995). This led to a reconfiguration of the local ice flow in the Untersee Oasis and the formation of proglacial Lake Untersee c. 15 560–13 000 cal yr bp (Fig. 1). During the deglaciation period in the Untersee Oasis, a proglacial lake also occupied Aurkjosen Cirque, as evidenced by palaeo-shorelines. The timing of the disappearance of this lake in Aurkjosen Cirque is unknown (drainage and/or evaporation to dryness once Anuchin Glacier receded from that valley c. 7800 cal yr bp). Our ¹⁴C ages of calcite crusts can be used to infer the minimum age of ice retreat and lake disappearance in Aurkjosen Cirque (Fig. 6). Considering that some of the bulk ¹⁴C ages of the calcite crusts were modern (Table II), we can assume that their ¹⁴C content was nearly 100% at the time of their precipitation and can provide a minimum age of carbonate precipitation. In Aurkjosen Cirque, the calcite crusts have ages ranging from 683 cal yr bp to modern. This would suggest that the palaeo-lake basin in Aurkjosen Cirque came into existence at least 750 years ago. However, in this polar desert environment, the calcite crusts require sufficient snow accumulation and temperatures warm enough to generate meltwater to allow for their formation. The calcite crusts all have median ¹⁴C ages < 1589 cal yr bp, with most being < 800 cal yr bp. Over the Holocene period, the past 1500 years corresponds to a warm-wet climate interval (Fig. 6; Buizert et al. 2021). As such, even if the ice retreated from Aurkjosen Cirque and the lake drained by the Middle Holocene, climate conditions probably were not favourable for the development of calcite crusts until the late Holocene.

Figure 6. Comparison of radiocarbon age distributions obtained from evaporative calcite crusts in the Untersee Oasis, East Antarctica, with Holocene reconstructed temperatures and ice accumulation rates (Buizert et al. 2021).

Conclusions

Based on these results, the following conclusions can be drawn regarding the origin, age and palaeo-environmental significance of the calcite crusts in the Untersee Oasis:

• • The calcite crusts are located within the vicinity of residual snow patches.

• • The calcite crusts have some of the highest δ¹⁸O (up to +17.4‰ VSMOW) and δ¹³C (up to +14.6‰ VPDB) compositions of any carbonate deposits in terrestrial polar environments. Their δ¹⁸O and δ¹³C values are substantially enriched with respect to the δ¹⁸O (average of -33.5 ± 2.4‰) and δ¹³C_DIC of snow meltwater (inferred to be in the -2.8 to -1.4‰ range).

• • The formation of the calcite crusts is ascribed to the evaporation of residual snow meltwater and the low relative humidity and strong winds in the Untersee Oasis, which would favour a kinetic isotope effect during the precipitation of the evaporative calcite crusts.

• • The ¹⁴C age distribution of the calcite crusts (1589 cal yr bp to modern) provides a minimum age for ice retreat and drainage of the palaeo-lake in Aurkjosen Cirque. However, in this polar desert environment, the calcite crusts require suitable climate conditions for their formation (sufficient snow accumulation and temperatures warm enough to generate meltwater), and their age distribution corresponds to the late Holocene warm-wet climate period.