Watershed, climate, and stable isotope data (oxygen-18 and deuterium) for 50 boreal lakes in the oil sands region, northeastern Alberta, Canada, 2002–2017

Watershed data, climate and stable data collected over a 16-year period from a network of 50 lakes in northeastern Alberta, are provided to allow for broader incorporation into regional assessments of environmental impacts, particularly hydrologic and geochemical processes under changing climate and land use development. Oxygen-18 and deuterium analyses of water samples are provided from late summer surveys of 50 lakes with varying land cover and permafrost conditions. Six sub-groups of lakes are represented, including Stony Mountains, West Fort McMurray, Northeast Fort McMurray, Birch Mountains, Caribou Mountains and Shield. This dataset includes 1582 isotopic analyses made on 791 water samples and 3164 isotope mass balance model outputs, as well as 800 lake/watershed parameters, 5600 climate parameters, and 800 modelled values for isotopic composition of precipitation used in the computations. Model data are provided to facilitate evaluation of transferability of the model for other applications, and to permit more sophisticated spatial analysis and intercomparison with geochemical and biological datasets. Details and further discussion on the isotope mass balance approach are provided in “Regional trends in water balance and runoff to fifty boreal lakes: a 16-year isotope mass balance assessment including evaluation of hydrologic drivers” [1]. Overall, the data are expected to be useful, in comparison with local and regional datasets, for water resource management and planning, including design of monitoring networks and environmental impact assessments for oil sands projects.


Data
Lake, watershed, landcover, climate, stable isotope data (oxygen-18 and deuterium), water balance data, and Mann-Kendall statistics are provided from a program of hydrological and geochemical monitoring of 50 lakes in the oil sands region of northeastern Alberta over a 16-year period, during 2002e2017 ( Fig. 1, Tables 1e13, RAMPlakesWY.xlsx). Water sampling and analysis was supported by the University of Victoria, InnoTech Alberta, and Alberta Environment and Parks and its predecessors, and was designed to provide original data complimentary to geochemical characterization of lakewatershed systems for critical loads assessment.
Lake and watershed parameters for the monitored lakes (Table 1), including lake area, drainage basin area, watershed area, and lake elevation were acquired using 1:50:000 raster-format digital elevation model (DEM) data from Canada National Topographic Series (NTS) map sheets according to pre-established protocols [2]. Vector-format hydrometric data were obtained from the Canadian National Hydro Network data obtained from the GeoBase portal http://www.geobase.com [3]. Watershed delineation utilized the ArcGIS program applying the ArcHYDRO tools, aided by preprocessing to fill small DEM sinks. Each individual watershed was delineated upstream of a lake outlet determined by hydrographic and elevation datasets. Lake and watershed areas were calculated based on equal area projections, and lake volumes, maximum depths, and mean depths were estimated based on bathymetric surveys conducted by Alberta Environment and Parks and its predecessors, mainly prior to 2005. Drift thickness and distance to buried channels for each lake were calculated in ArcGIS based on geological and hydrostratigraphic data layers obtained from the Alberta Geological Survey web portal (https://ags.aer.ca/data-maps-models/digital-data). Detailed land cover classification mapping and assessment of permafrost conditions from 1:20,000 air photos was carried out prior to 2005 by R. Bloise, Southern Illinois University (pers. Comm.) based on the Alberta Wetland Classification methodology [4] which was then used to estimate areal extent of these terrain types in each watershed, as summarized previously [5].
Interpolated from the 32-km resolution North American Regional Reanalsis (NARR) monthly climatology [6] using the Grid Analysis Display System (GrADS) [7]. An evaporation flux-weighting protocol [8] was also applied to condition climate data to improve representativeness for assessment of isotope-based water balance, as used in numerous Canadian and international assessments [9e20].
From 2008 to 2017, a dual-inlet Thermofisher Scientific Isotope Ratio Mass Spectrometer, Delta V interfaced with a Gasbench peripheral (for oxygen-18) and H-Device peripheral (for deuterium) was used for isotopic analysis [21,22]. Comparable protocols were employed to measure isotopic content during 2002e2007 [5]. Results are reported in "d" notation in per mil (‰) relative to Vienna Standard Mean Ocean Water (V-SMOW), normalized on the SMOW/SLAP scale [23]. Analytical uncertainty, as estimated from standard deviation of repeats, is better than ±0.1‰ for d 18 O and ±1‰ for d 2 H. Raw isotopic data for lake water samples are provided in Tables 8 and 9 Mean values for each lake are summarized in Table 10 in comparison with interpolated estimates of isotope composition of precipitation and atmospheric moisture for each of the 50 sites. Monthly precipitation d 18 O estimates were obtained for each lake/ watershed location based on a protocol developed using empirically derived global relationships between latitude and elevation [24] fitted to regional precipitation data from the Canadian Network for Isotopes in Precipitation [25]. The d 2 H composition of monthly precipitation was calculated assuming that precipitation would follow the relationship defined by the Global Meteoric Water Line (GMWL; [26]).
Annual d 18 O and d 2 H in precipitation were then amount-weighted using monthly isotope data and NARR precipitation amounts. Isotope balance estimates of evaporation/inflow (Table 11) and water yield to lakes (Table 12) were based on a previously demonstrated model and protocols [12].

Lake
No.
Lake  Table 12 Site-specific annual water yield to lakes, RAMP sites, northeastern Alberta.
Lake No. Lake ID Wy(mm)   Table 14 Mann-Kendall trend results and number of years observations.

Water sampling and analysis
Acid-sensitive lakes were selected by the Regional Aquatics Monitoring Group from an initial regional survey of 449 lakes to be representative of lake and watershed characteristics and chemistry across six sub-regions within the study area [27]. Lakes were generally situated in remote locations accessible only by fixed-wing aircraft or helicopter. Water samples for analysis of the stable isotopes of water were collected for the purpose of establishing site-specific and year-specific water yield to lakes using an isotope balance method [12]. This, combined with concurrent geochemical sampling for base cations, was designed to enable estimation of critical loads of acidity to the lakes using a simple steadystate water chemistry model [28]. Critical loads of acidity is a measure of the buffering capacity of the lake-watershed system to potential acidifying emissions. In the case of the RAMP lakes network, assessment of potential impacts from emissions from oil sands operations on local watersheds and lakes was the primary objective of annual time-series monitoring. One complicating factor realized in the second decade of monitoring was the significant impact of permafrost thaw on runoff to many lakes in the Birch Mountains, Caribou Mountains and Northeast of Fort McMurray [see 1,4], which may significantly and differentially affect the long-term representativeness of the site-specific critical load of acidity calculations.
For deeper lakes (>3 m), lake water samples were collected near the centre of the major basin at a single deep-water site using weighted Tygon tubing and a one-way valve. This approach was used to provide vertically-integrated samples representative of the euphotic zone (defined as twice the Secchi disk depth). For shallow lakes (<3 m deep), composite samples were created from five to ten 1 L grab samples collected at 0.5 m depth along an upwind to downwind transect. Samples taken from a given lake were then combined to form a single composite sample. Euphotic zone samples from deep lakes, and composite samples from shallow lakes were then split according to requirements for specific analyses including an unfiltered, 30-mL sample in a high-density polyethylene bottle for stable isotope analysis, as well as various bottles for geochemical analyses. All bottles were subsequently refrigerated and returned to various labs for analysis (Colin Cooke, Alberta Environment and Parks, pers. Comm.)

Water balance data
Lake area, watershed area and NARR monthly climatology parameters (precipitation, temperature, relative humidity, evaporation and precipitation) were used in combination with isotopic data to estimate annual lake water balance by an established isotopic method [12]. Input to lakes was estimated based on amount-weighted isotopic composition of precipitation. Isotopic composition of atmospheric moisture was defined using the partial equilibrium approach [4], which involved fitting predicted oxygen-18 and deuterium enrichment to match the slope of the local evaporation line [1]; see also [29].