The Spatiotemporal Variability of Snowpack and Snowmelt Water ¹⁸O and ²H Isotopes in a Subarctic Catchment

Abstract

This study provides a detailed characterization of spatiotemporal variations of stable water 18O and 2H isotopes in both snowpack and meltwater in a subarctic catchment. We performed extensive sampling and analysis of snowpack and meltwater isotopic compositions at 11 locations in 2019 and 2020 across three different landscape features: (a) forest hillslope, (b) mixed forest, and (c) open mires. The vertical isotope profiles in the snowpack's layered stratigraphy presented a consistent pattern in all locations before snowmelt, and isotope profiles homogenized during the peak melt period; represented by a 1–2‰ higher δ $\delta $18O value than prior to melting. Our data indicated that the liquid‐ice fractionation was the prime reason that caused the depletion of heavy isotopes in initial meltwater samples prior to the peak melt period. The liquid‐ice fractionation was influenced by snowmelt rate, with higher fractionation during slow melt. The kinetic liquid‐ice fractionation was evident only in close examination of meltwater lc‐excess values, not δ $\delta $18O values alone. Meltwater was isotopically heavier and more variable than the depth‐integrated snowpack; the weighted mean of meltwater isotope values was higher by 0.62–1.33‰ δ $\delta $18O than the weighted mean of snowpack isotope values in forest hillslope and mixed forest areas, and 1.51–6.37‰ δ $\delta $18O in open mires. Our results reveal close to 3.1‰ δ $\delta $18O disparity between the meltwater and depth‐integrated snowpack isotope values prior to the peak melt period, suggesting that proper characterization of meltwater δ $\delta $18O and δ $\delta $2H values is vital for tracer‐based ecohydrological studies and models.

Flushing or mixing? Stable water isotopes reveal differences in arctic forest and peatland soil water seasonality

“…Thus, isotope samples from snowmelt produce a better estimate of the snowmelt signal that enters the soil and presumably becomes available for root water uptake, compared to snowfall samples (Earman et al, 2006;Unnikrishna et al, 2002). Site-specific snowmelt isotope data were available for each sampling location due to the proximity of snow lysimeters (Noor et al, 2023), providing additional information on the spatial variability of snowmelt fluxes. The total amount of snowmelt was estimated as the snow water equivalent value (SWE) based on snowpack depth and snow density from snow surveys conducted biweekly during winter, and every 2 to 3 days during the snowmelt period (Figure S5).…”

“…The isotope values for snowmelt (δ w in Equation 3) in both seasons were calculated as the flow-weighted values from daily lysimeter samples. The flow-weighted snowmelt lysimeter data allowed us to take into account the typically observed gradual enrichment of heavy isotopes in the snowmelt water over the melt season (Noor et al, 2023, Ala-aho, Tetzlaff, McNamara, Laudon, Kormos, & Soulsby, 2017. Such data are not commonly available, but they enabled an representative estimate of the total snowmelt water isotope values entering the soil, eliminating the potential bias of preferentially sampling the early or late snowmelt with differing isotope values.…”

“…Although mid-winter melts are not common in Pallas, 2 out of 11 snowmelt lysimeters had received significant mid-winter snowmelt water (Noor et al, 2023), suggesting localized mid-winter snowmelt entering the soil. Continuous low rates of melting at snow-soil interface have been reported in many other places (Unnikrishna et al, 2002), while Sutanto et al (2014) described how two water compartments with different isotopic signals approach their weighted mean due to vapour exchange at equilibrium.…”

“…Snowmelt samples for isotopic analysis were collected daily throughout melting periods in 2019 and 2020 from 11 snow lysimeters along the snowline (Noor et al, 2023). Supplementary hydrometeric data comprise groundwater levels, temperatures and peatland water table measured at 15-min intervals with water level loggers (Solinst levelogger, Solinst Canada Ltd.) and daily snowmelt fluxes.…”

“…How to cite this article: Muhic, F., Ala-Aho, P., Noor, K., Welker, J. M., Klöve, B., & Marttila, H. (2023). Flushing or mixing?…”