Measurements of water isotope composition and depth variation reveal strong influence of hydroclimate on water balance and water levels of shallow lakes in the Peace-Athabasca Delta (northern Alberta, Canada)

Abstract

Shallow water bodies are abundant, productive ecosystems in northern high-latitude regions where they provide habitat for fauna and hold socio-economic and cultural importance for remote communities. In northeastern Alberta, the Peace-Athabasca Delta (PAD), located within Wood Buffalo National Park (WBNP), is a large (~6000 km2) inland freshwater-rich landscape with an abundance of shallow lakes that is listed as a Ramsar Wetland of International Importance and contributes to WBNP’s UNESCO World Heritage status. However, concerns have long persisted about the effects of primarily hydroelectric regulation of Peace River flow and climate change on widespread drawdown of lake water-levels in the delta. In response to these concerns, and pressures from local communities and an international agency, a federal Action Plan was developed with goals to improve aquatic ecosystem monitoring at the PAD. Despite considerable prior study, this requires better understanding of hydrological processes that influence lake water balances and water levels in the delta and advancement of methodological approaches. A 7-year research project was launched in 2015 to inform the design of a long-term aquatic ecosystem monitoring program at the PAD. Research presented in this thesis is an outcome of two key observations that were made during the last two years (2020-2021) of that project, which provided opportunity to explore the influence of unusually high relative humidity on isotope-based lake water-balance modelling and the influence of snowmelt runoff on lake depth when widespread ice-jam flooding occurred.

For two decades, the isotope composition of an index lake (informally named ‘PAD 18’) provided a constant and reliable estimate of δSSL (i.e., the water isotope composition of a terminal basin where evaporation = inflow), a key parameter used in an isotope framework for water-balance modelling of lakes at the PAD. In 2020-2021, a sudden and unprecedented shift occurred in the isotope composition of PAD 18, which required closer examination of the isotope framework to ensure accurate assessments of lake water balances. The shift in isotope composition of PAD 18 coincided with an increase in relative humidity by 4.6% from 66.2% (2015-2019) to 70.8% (2020-2021). An independent calculation of δSSL, using formulation by Gonfiantini (1986) and flux-weighted air temperature and relative humidity from 2020-2021, was determined to be equivalent to the observed isotope composition of PAD 18 during 2020-2021, confirming the shift in isotope composition reflected re-equilibration of the lake water with the atmosphere following the rise in relative humidity. The importance of accurate characterization of δSSL for isotope mass-balance modelling is demonstrated by estimating evaporation-to-inflow (E/I) ratios for 61 lakes across the PAD in July 2021 using the previous 2015-2019 isotope framework and the new 2020-2021 isotope framework. Results show differences between the two frameworks are minor when E/I ratios are low, but they become consequential when E/I ratios are high because of the non-linear effect of relative humidity on isotope framework parameters. Appropriate parametrization for isotope mass-balance modelling is critical for monitoring agencies utilizing isotope-based techniques at the PAD, and elsewhere. Thus, re-evaluation of isotope framework parameters may become increasingly important as climate warming fosters greater variability of meteorological conditions.

Snowmelt runoff is well-recognized for sustaining shallow waterbodies across semi-arid and sub-humid northern regions where evaporation rates are high. At the PAD, however, the influence of snowmelt runoff from within the delta remains largely unknown because the attention on drawdown of perched lakes has focused almost exclusively on alteration of the ice-jam flood regime. During spring of 2020, extensive ice-jam flooding occurred which raised water levels in flooded lakes across the PAD, but water levels rose also in many non-flooded lakes, providing an opportunity to explore the role of snowmelt runoff on lake levels. Using measurements of water depth and water isotope compositions, the contributions from snowmelt runoff and river floodwaters to lake-level rise were quantified at 52 lakes and spatial patterns were explored. Results reveal that the lake-level rise from snowmelt runoff at non-flooded lakes rivals the rise caused by floodwaters at flooded lakes. Geospatial analyses show lake-level rise by snowmelt was greatest in the northern Peace sector and the southwestern margin in the southern Athabasca sector where greater topographic relief and denser vegetation enhance snow accumulation via redistribution of snow by wind from adjacent flat, sparsely vegetated terrain. In contrast, the rise due to floodwater input was greatest in lower lying flood-prone central areas of the Athabasca sector. Exploration of meteorological records since 1963 reveal that similar contributions of snowmelt runoff to lake-level rise as occurred in 2020 were likely common between the early 1960s and late 1980s but only once since 1987 (2018), suggesting that reduced input from snowmelt runoff has contributed to drawdown of shallow perched lakes in recent decades.

Concern for drawdown of water levels in the PAD have prompted a need for enhanced monitoring of lake water balances and water levels, especially as mitigative actions are considered. The research presented in this thesis advances knowledge about the influence of shifts in relative humidity on isotope-based modelling of lake water balances and the role of snowmelt runoff on lake depth. The novel integration of water isotope composition and water depth measurements demonstrated here provides opportunity for assessing the success of mitigative actions such as strategic water releases from the W.A.C. Bennett Dam located upstream of the delta to enhance ice-jam flooding. While long-term aquatic ecosystem monitoring programs must be built on a foundation of knowledge generated through primary research, systematic and repeated measurements over sufficient scales of space and time may also trigger new research that leads to important new discoveries, as revealed in this thesis.