Water sources and composition of dissolved gases and bubbles in a saline high Arctic spring

Abstract

We investigate the water sources for a perennial spring, "Little Black Pond," located at Expedition Fiord, Axel Heiberg Island in the Canadian High Arctic based on dissolved gases. We measured the dissolved O2 in the likely sources Phantom Lake and Astro Lake and the composition of noble gases (3He/4He, 4He, Ne,36Ar, 40Ar, Kr, Xe), N2, O2, CO2, H2S, CH4, and tritium dissolved in the outflow water and bubbles emanating from the spring. The spring is associated with gypsum-anhydrite piercement structures and occurs in a region of thick, continuous permafrost (400-600 m). The water columns in Phantom and Astro lakes are uniform and saturated with O2. The high salinity of the water emanating from the spring, about twice sea water, affects the gas solubility. Oxygen in the water and bubbles is below the detection limit. The N2/Ar ratio in the bubbles and the salty water is 89.9 and 40, respectively, and the relative ratios of the noble gases, with the exception of Neon, are consistent with air dissolved in lake water mixed with air trapped in glacier bubbles as the source of the gases. The Ne/Ar ratio is ~62% of the air value. Our results indicate that about half (0.47±0.1) of the spring water derives from the lakes and the other half from subglacial melt. The tritium and helium results indicate that the groundwater residence time is over 70 years and could be thousands of years.

Fig 1. Context image showing the location of Little Black Pond (LBP) and Colour Peak Springs (CP), the glaciers and ice-dammed lakes: Astro Lake (AL) and Phantom Lake (PL).

Image from Copernicus Sentinel-2 L2A 2020-07-11 processed with EO Browser. Inserts showing the location of Axel Heiberg Island and the location of the site on the island.

Fig 2. Schematic diagram depicting the sources of gas and water showing the contribution of lakes and subglacial melt and the flow of subsurface water through the salt strata associated with the diapirs.

Modified from Andersen et al. [7].

Fig 3. Close-up of Little Black Pond in the Gypsum Hill springs with the Expedition River in the background.

The pond is about 1 m in diameter.

Fig 4. Aerial image of Phantom Lake looking toward the Thompson Glacier, May 2003.

The lake is 4 km long.

Fig 5. Temperature and dissolved oxygen in Phantom Lake in July of 2000 and 2002.

Fig 6. Profile of temperature and dissolved oxygen in Astro Lake on 7 July 2002.

Fig 7. The bubble gas pressure ratio is the measured partial pressure in the bubbles for each gas divided by the computed partial pressure in the bubble based on the measured salt concentration in the spring water.

This is plotted for each gas as a function of salinity. As the salinity increases, the solubility of gases decreases. The bubble gas pressure ratios are ~0.8 at the measured values for salinities near the level in the springs (72 g/kg), indicating that the gas in the bubbles is 20% lower than the equilibrium–the gases do not fully achieve equilibrium with the dissolved gases.

Fig 8. The calculated ratio of N₂ to Ar in the bubbles in the spring as a function of the fraction of the source water that comes from air-saturated fresh water lakes.

The other source is gas trapped in glacier ice carried into the springs. The dashed lines show the values of N₂/Ar based on measured gas values in the spring water and the bubbles (lower line). The upper line shows the ratio for and the measured gas values in the spring water and the assumption of equilibrium in the bubbles in the spring. The uncertainty in the N₂/Ar measurements is ±3%, comparable to the spacing between the two dashed lines (2%). The red line and the right-hand scale show the corresponding results for the ratio of Kr/Xe. The spacing between the red dashed lines is larger, ~ 6%. For both gas, ratios the fraction of liquid water that fits the data is 0.47±0.01, shown by the blue-shaded region.

Fig 9. Noble case ratios in the springs normalized to air-saturated values for the saline spring water.

The scale does not reach the level of He, which is hundreds of times higher in the spring bubbles than air levels due to He accumulation in the groundwater. Error bars are the standard deviation for four replicates.