1. Introduction
Climate warming in regions characterized by nival-dominated hydrological regimes may exacerbate concerns about current and future water supplies. In fact, even slight changes in temperature and precipitation can affect freshwater availability through changes brought to hydrological cycle components and processes, and moreover in streamflow seasonality (Bonsal et al. 2019; Eum et al. 2017; Cohen et al. 2015; Koshida et al. 2015; Mortsch et al. 2015; Vincent et al. 2015; Barnett et al. 2005). Derksen et al. (2019) highlighted that the timing, duration, and intensity of melt periods, as well as the type of precipitation (rain versus snow), are influenced by the energy budget at the surface and resulting temperature. According to Bush and Lemmen (2019), warmer summers are also expected to increase evaporation of surface water and contribute to reduced summer water availability in the future despite more precipitation in some places.
The realities of climate change are requiring stakeholders to anticipate local changes for a variety of planning and adaptation needs, particularly in the water sector (Carlson et al. 2021). The Intergovernmental Panel on Climate Change Sixth Assessment Report (IPCC 2022) indicates an increase of 1.09°C in the global surface temperature during the period 2011–20 relative to the preindustrial period 1850–1900, with surface temperatures expected to keep rising. In Canada, where most watersheds are influenced by snow-dominated hydrologic patterns (Cohen et al. 2015), the warming rate is on average twice the magnitude of Earth’s climate warming signal (Bush et al. 2022; Jiang et al. 2017), and up to 3 times the global rate above 60°N (i.e., northern Canada), largely because of Arctic amplification (Braun et al. 2021; Stadnyk and Déry 2021). As an example, the global mean annual temperature has increased by 0.85°C over the period 1880–2012 (Bush et al. 2019), whereas between 1948 and 2016, an increase of 1.7° and 2.3°C was estimated for the whole of Canada and the northern regions of Canada, respectively (Zhang et al. 2019). Understanding the changes at the local scale are all that much more relevant.
Across western Canada, freshwater supply within their watersheds is dominated by glaciers and snowpack melting, since the eastern slopes of the Rocky Mountains form the headwaters of many rivers. This makes the region particularly susceptible to warmer temperatures, but also to intense water deficits and droughts as it lies under the rain-shadow side of the Rocky Mountains (Newton et al. 2021; Bonsal et al. 2020; Bonsal and Cuell 2017; Dibike et al. 2017; Jiang et al. 2017; Bonsal et al. 2013, 2011; Schindler and Donahue 2006; Barnett et al. 2005). For instance, Schindler and Donahue (2006) reported that summer flows in the Athabasca (1958–2003) and Peace (1915–2003) Rivers have declined by 20% and 42% at Fort McMurray and Peace River, respectively. They reported that reductions were seen in all the major rivers within the Canadian Prairies and were likely associated with the regulation of river flows due to dams, the withdrawals for irrigation, municipal, and industrial uses, and the temperature warming effects on evapotranspiration and winter snowpack.
Water availability is a crucial factor in sustaining natural ecosystems’ integrity and human well-being, and across western Canada, it is at the core of the region’s economic activities, including the oil sands industry. Canada possesses the third-largest oil reserves in the world (Sauchyn et al. 2015), 97% of which are in northern Alberta in the Athabasca, Peace River, and Cold Lake oil sands areas within the Athabasca River basin (ARB), the Peace River basin (PRB), and the Beaver River basin (BRB), respectively (Lunn 2013). The oil sands production requires large amounts of water and energy and leads to high greenhouse gas emissions (Wren et al. 2023). Water is generally withdrawn from groundwater or the nearest river, lake, or reservoir. The Athabasca River itself provides water for the Athabasca oil sands mining operation (Rood et al. 2015), and in 2010, the Alberta oil and gas industry required 74.5% of the total surface water licensed allocations in the ARB, representing 4.4% of the Athabasca River mean annual flow (Sauchyn et al. 2015). The neighboring PRB is home to important hydroelectricity infrastructure for BC Hydro (Prowse and Conly 2002). Other water demands within the region include Indigenous People’s needs (Westman and Joly 2019), the maintenance of aquatic ecosystems like the Peace–Athabasca Delta Ramsar Convention site (Dubé et al. 2022), and the Wood Buffalo National Park UNESCO world heritage site (Bonsal et al. 2020). Even though the northern river basins of Alberta have a relative abundance of water (Jordaan 2012), an ongoing warmer and drier climate across western Canada may affect water supplies and users because as suggested by Leong and Donner (2015), the future occurrences of low flows during summer in the ARB would be driven more by climate change than by oil sands water withdrawals. In the Alberta oil sands region, the Oil Sands Monitoring Program (OSMP 2019) funds research projects to assess environmental conditions, including detecting and quantifying changes in climate and water budget variables, thus supporting the regulation of the oil sands industry operations in an environmentally responsible way.
The ongoing monitoring of changes in the climate from global to local scale (Eum et al. 2023; IPCC 2022; Rajulapati et al. 2022; Bush and Lemmen 2019; Dibike et al. 2017; Barnett et al. 2005), has been useful to enhance our understanding of past, current, and climate future scenarios projected future hydroclimatic patterns, with the goal to provide decision-making information to support impacts management. Such research studies have been facilitated on the one hand, by the growing capacity to produce high-resolution gridded climate datasets, for instance from station-based and global climate model outputs (Eum et al. 2023; Braun et al. 2021; Eum and Gupta 2019). On the other hand, the development of standard and robust statistical-based indicators like the Expert Team on Climate Change Detection and Indices (ETCCDI) and its extended version the Expert Team on Sector-Specific Climate Indices (ET-SCI), enable to detect, monitor, and understand changes in mean, moderate, and extreme climatic conditions (Alexander and Herold 2016; Zhang et al. 2011; Alexander et al. 2006). These indicators, endorsed and recommended by the World Meteorological Organization (WMO), are derived from daily time series of total precipitation (PR), minimum temperature (TN), and maximum temperature (TX) climate variables, and allow large climate datasets to be synthetized into more digestible information (Eum et al. 2023; Rajulapati et al. 2022; Braun et al. 2021; Burhan et al. 2021; Sillmann et al. 2013a,b; Alexander et al. 2006) for decision-makers. For instance, based on patterns of indices such as the annual highest (and lowest) daily maximum (minimum) temperatures and the growing season length, Bush and Lemmen (2019) showed that the increasing (decreasing) of extreme hot (cold) events and the longer growing seasons are, among others, part of the climate change signals. The ET-SCI list of hydroclimate indices also include heatwave aspects indicators, which allow for characterizing summer climate state in an absolute sense (Nairn and Fawcett 2015; Perkins and Alexander 2013), and drought representative indices, the standardized precipitation index (SPI), and standardized precipitation evapotranspiration index (SPEI), which enable quantifying deficits and surpluses in regional water availability at multiple time scales (Bonsal et al. 2020; Bonsal and Cuell 2017; Dibike et al. 2017; Bonsal et al. 2013; Vicente-Serrano et al. 2010; McKee et al. 1993).
The scope of the present study is, first, to propose a set of indices from the ET-SCI list that inform on climate patterns and water availability. Second, we have designed a framework and produced a set of R programming language scripts that utilize reliable climate datasets to calculate the ET-SCI and use them to analyze and monitor changes in hydroclimatic conditions relative to an established baseline condition, and for any region of interest globally. Our approach is flexible enough to incorporate spatial and temporal analyses for various time periods as needed and to facilitate updating as the quality of available data evolves. Third, we used the Alberta oil sands region to implement our framework and contribute to the development of a high-level commentary on climate change and its potential implications for water supply in this region. Our goal with respect to this target region is to support sustainable adaptation to climate change by all water users.
2. Study framework
a. Study domain
The target region used in this study is the region formed by the Athabasca, Cold Lake, and Peace River oil sands deposits, located in northern Alberta, and encompassed by the bold black lines as shown in Fig. 1. This region is part of the boreal plains ecozone with low elevation profiles, south–north climate gradients, a hummocky landscape with numerous lakes and wetlands areas, extensive forests, sporadic discontinuous permafrost, and poor drainage (Newton et al. 2021). In terms of the Alberta natural regions (Eum et al. 2023), the study area is within the boreal forest natural region. A cool continental climate with mean annual temperature of approximately 0°C dominates the domain with low annual precipitation of ∼400 mm on average, of which about 50% falls in the June–August period (Nenzén et al. 2020).
Region encompassed by the Alberta oil sands areas in Canada.
Citation: Journal of Hydrometeorology 25, 1; 10.1175/JHM-D-23-0051.1
The Athabasca (273 808 km2) and Peace (335 331 km2) Rivers are the major systems spanning the study domain, with their headwaters originating respectively in southern Alberta and northern British Columbia, from the eastern slopes of the Rocky Mountains (Dubé et al. 2022). Both rivers are part of the Mackenzie River system, which flows north and covers a total drainage area of 1.7 million km2 within the pan-Arctic watershed domain (Stadnyk et al. 2021). The Athabasca (Peace) River basin has glaciers in its headwater regions with 44.5% (20%) of discontinuous permafrost, 17% (30.6%) of isolated permafrost, and 0.1% (0.1%) of glaciers (Bonsal et al. 2020).
b. Climate data
1) Global climate models
High-resolution gridded climate datasets produced from station-based observations and climate model outputs are widely employed to enable long-term impact studies that are consistent over time and across space. For our target study area, we used 12 climate simulations associated with seven global climate models (GCM) and four climate change scenarios. These models participate in the sixth phase of the Coupled Model Intercomparison Project (CMIP6), whose climate simulations support the IPCC Sixth Assessment Report (IPCC 2022). The details about the evaluated climate runs are presented in Table 1. The models were identified by the Resource Stewardship Division of Alberta’s Environment and Protected Areas (EPA), from a larger multimodel ensemble (see Table S-1 in Eum et al. 2023), for the province of Alberta’s impact studies. They implemented the integrated computational geometry algorithm developed by Farjad et al. (2019), and based on Cannon (2015), to enable an objective selection of a subset of climate simulations that fully capture relative changes in the ETCCDI’s 27 core indices for the province of Alberta. The datasets of interest associated with these simulations were downscaled and bias corrected by applying the MBCDS method (Eum et al. 2020), a multivariate bias correction (MBC) method that combines a quantile delta mapping with a distribution-free shuffle (DS) approach. The gridded hybrid historical dataset, produced at 10-km resolution by Eum and Gupta (2019), was employed as the climate observations at this end.
Details on the CMIP6 climate runs used in this study.
For each climate simulation, the historical period ends in 2014 since the future in the CMIP6 data starts in 2015 (Eyring et al. 2016). The future projections in the CMIP6 experiments are based on a new scenario modeling framework, which combine the well-known climate radiative forcing levels [i.e., representative concentration pathways (RCPs); van Vuuren et al. 2011] and the newly introduced socioeconomic development aspects [i.e., shared socioeconomic pathways (SSPs); Riahi et al. 2017], thus facilitating an integrated analysis of future climate impacts, vulnerabilities, adaptation, and mitigation. The SSPs provide five distinct pathways within which optimistic SSP1 and pessimistic SSP3 symmetric markers describe climate futures where challenges to adaptation and mitigation are both projected to be low (SSP1) and high (SSP3). Conversely, SSP4 and SSP5 denote two asymmetric markers where a duo of high challenges to mitigation and low challenges to adaptation is expected for SSP4, whereas the opposite is true for SSP5. The marker SSP2 represents intermediate challenges for both adaptation and mitigation efforts. For instance, in Table 1, the selected future scenario for the model GFDL-CM4 is SSP245, implying a combined SSP2 marker with RCP4.5 forcing level.
The 12 climate simulations assessed in this study include four future pathways where human development across the world could be more sustainable (SSP1), balanced (SSP2), regionally unequal (SSP3), or based on an intensive technology oriented-economy (SSP5), in the presence of low (RCP2.6), intermediate (RCP4.5), or high (RCP7.0, RCP8.5) emissions scenarios (Rajulapati et al. 2022; Riahi et al. 2017). Since most of the GCMs evaluated share few common future scenarios, each of the 12 climate runs is examined individually. The datasets for each climate simulation include a gridded daily time series of total precipitation (PR), maximum temperature (TX), and minimum temperature (TN) at 10-km resolution covering the period 1950–2100 and encompassing the study area.
2) Hydroclimate indicators
The ETCCDI and ET-SCI lists of indices enable us to represent hydroclimatic conditions for any region of interest globally (Alexander and Herold 2016; Zhang et al. 2011; Alexander et al. 2006). Monthly and annual time series can be derived for most of these indices, but others are annual only. For the Alberta oil sands region, we have selected 13 indicators (Table 2), and we examine their annual time series, which are derived from the identified climate simulations PR, TN, and TX. These indices contain information on hydroclimatic regional conditions, specifically evaluating patterns in extreme hot and cold events, multiday wet and dry cycles, and water availability within a season or a year (Rajulapati et al. 2022; Braun et al. 2021; Burhan et al. 2021; Perkins and Alexander 2013; Sillmann et al. 2013a,b; Vicente-Serrano et al. 2010).
Description of the hydroclimate indicators evaluated. Sensitive sectors are health (H), agriculture and food security (AFS), and water resources and hydrology (WRH).
Four temperature-related indices, derived from the climate variable TX or TN, are examined, and include the annual highest daily maximum temperatures (TXx) and the annual lowest daily minimum temperatures (TNn), which are both absolute indices (Alexander et al. 2006). The two other indices, namely, the warm spell duration index (WSDI) and the cold spell duration index (CSDI), depict respectively the annual count of days when at least six consecutive days of TX (and TN) are above (below) the 90th (10th) percentile of their corresponding calendar day TX (TN). For the percentile-based temperature indices, annual calendar daily values for each percentile are calculated over a selected reference period by applying a 5-day window centered on each calendar day (Sillmann et al. 2013a,b; Zhang et al. 2011, 2005), and used by default to generate the indices time series for the study period. Both the WSDI and CSDI are duration-based indicators and describe warm and cold spells that could occur at any time of the year (Nairn and Fawcett 2015; Perkins and Alexander 2013).
Among the four precipitation-based indices derived from the climate variable PR, there is the annual total precipitation on wet-days (PRCPTOT), which is not necessarily an extreme index but provides useful information to understand patterns in precipitation (Sillmann et al. 2013a,b). The remaining three indices which characterize extreme conditions of precipitation are the annual highest amounts of precipitation that fall in five consecutive days (RX5day), the fraction of very wet days contributing to PRCPTOT (R95pTOT), and the annual longest spell of consecutive dry days (CDD). The index RX5day falls in the absolute indicators category, whereas CDD and R95pTOT are duration- and threshold-based indices, respectively (Sillmann et al. 2013a,b; Alexander et al. 2006). Since R95pTOT is a percentile-based precipitation index, a single threshold is derived as the 95th percentile of the daily total precipitation whole sample over the defined baseline condition (Zhang et al. 2011).
The selected indices also include three aspects of heatwave events that describe their number, frequency, and severity. The first is the heatwave number (HWN), which shows the number of heatwave events occurring in the summer period, which extends from May to September for the Northern Hemisphere (Alexander and Herold 2016); the second index is the heatwave frequency (HWF), which describes the total days having contributed to heatwaves as identified by HWN each year; and last, the heatwave duration (HWD) depicting the duration of the annual longest heatwaves as identified by HWN. With respect to HWN, it is important to note that as much as this index characterizes extreme hot temperatures, its temporal evolution can remain constant or show a decreasing trend when a single event is spread out over a long period.
The assessment of the heatwave events requires a threshold of exceedance, and in this study, is the excess heat factor (EHF) whose calculation is based on the average daily temperature (Tavg) and the daily calendar 90th percentile of Tavg (Perkins and Alexander 2013). Both TX and TN are integrated in the EHF calculation process, making it more appealing from a purely climatological standpoint (Nairn and Fawcett 2015). A heatwave event is identified when the EHF threshold value is positive for three or more consecutive days. In the case of the heatwave aspects, the percentile-based threshold needed to calculate EHF is also estimated for each calendar day over a 30-yr baseline period but using a 15-day running window (Alexander and Herold 2016). This procedure helps to account for temporal variabilities and estimate each calendar day threshold from a large sample (Braun et al. 2021; Sillmann et al. 2013a,b; Zhang et al. 2005).
Finally, we considered the SPEI to assess drought conditions and regional water availability (Vicente-Serrano et al. 2010; Dibike et al. 2017). The SPEI is a water balance index determined by estimating the difference between monthly precipitation and potential evapotranspiration (PET), and then standardizing them at different time scales (Vicente-Serrano et al. 2010). The PET is estimated using the modified Hargreaves method (Droogers and Allen 2002; Hargreaves 1994), which requires data on latitude, and monthly TN, TX, and PR as input. The SPEI is an extension of the well-known SPI proposed by McKee et al. (1993) and is calculated monthly given a moving window of typically 1, 3, 6, 9, 12, 24, or 48. With respect to the purpose of the study, certain accumulation periods (i.e., time scales) are more relevant than others. We first focus on the 3-month SPEI by compiling the annual time series for the specific month of August, because these intensities reflect the regional water budget from June to August, and thus over the summer period. We hereafter refer to such an index as Aug_SPEI3. Then we retrieved the annual time series for the month of September from the 12-month SPEI to consider the regional water availability over the water-year period. This index is hereafter referred to as Sep_SPEI12 and includes the previous 11 months antecedent conditions (i.e., from October of the precedent year to September of the current evaluated year). The Aug_SPEI3 and Sep_SPEI12 aggregated indices represent a direct estimate of water availability during summer and over the water-year (Bonsal and Cuell 2017; Dibike et al. 2017; Bonsal et al. 2013). The SPEI or SPI with negative (positive) intensities I denote dry (wet) conditions, and based on the primary drought classification system described by McKee et al. (1993), from the normal condition (I = 0) going downward, the following moisture categories follow: mildly dry (−0.99 ≤ I < 0), moderately dry (−1.49 ≤ I ≤ −1.00), very dry (−1.99 ≤ I ≤ −1.50), and extremely dry (I ≤ −2.00). Conversely, similar categories apply to wet conditions.
c. Analysis methods
For a target region, the assessment of spatial patterns in the hydroclimatic conditions using indicators such as the ET-SCI require gridded indices calculated from gridded PR, TX, and TN datasets representative of the region. However, to evaluate the temporal patterns, the gridded indices can be first calculated then spatially averaged, or the input gridded climate datasets can be spatially averaged and used to calculate the indices regional time series. For the Alberta oil sands region, we examine only the temporal patterns in the annual time series of the selected indices. Because this region shows similar climatic, physiographic, and ecological characteristics (Eum et al. 2023; Newton et al. 2021), we have first applied a simple spatial average of the individual gridded PR, TN, and TX time series over the study area and for each of the 12 climate simulations evaluated, and subsequently calculated the indices. According to Avila et al. (2015), differences in gridscale annual values have less impact on long-term trends and interannual variability when regional averages are considered, regardless of the order of the spatial average.
The ClimPACT2 software, an open-source R package (https://github.com/ARCCSS-extremes/climpact) developed by the WMO’s ET-SCI to ensure a consistent calculation of the ET-SCI (Alexander and Herold 2016), was used to compute the annual time series of the selected indicators. The exception was the SPEI from which the original monthly time series data were further aggregated annually for the months of interest. The indices were all calculated for the study period 1950–2100, and for the indicators requiring percentile-based thresholds, the established baseline period 1985–2014 was considered.
The temporal assessment of patterns in the indices annual time series, with respect to each climate simulation, includes the generation of the time series plots from 1950 to 2100, a linear trend analysis over the historical (and future) period 1950–2014 (2015–2100), and the quantification of the projected future changes over the near future 2041–70 (F1) and the far future 2071–2100 (F2), relative to the baseline conditions of 1985–2014 (REF). The nonparametric Mann–Kendall test (Kendall 1975; Mann 1945) with the trend-free prewhitening procedure (Yue et al. 2002) was applied to improve the detection of significant trends when the presence of serially correlated time series was identified. The presence of a trend was evaluated at the 5% significance level and the slope of the trend over the analysis period was estimated using Sen’s method (Sen 1968). For the indices PRCPTOT and RX5day, the projected changes (i.e., deltas Δ) for a given year y, are relative values {i.e., Δy = [(F1y − REFy)/REFy] × 100}, while the changes for the remaining indices are expressed in absolute values (i.e., Δy = F2y − REFy). We used boxplots to represent the annual distributions of the future projected changes and referred to specific values derived from them to translate the average annual changes anticipated in the regional hydroclimatic annual conditions.
All dataset processing was done using R software (R version 4.3.0 as of August 2022), with packages like ncdf4 for multidimensional data manipulation, ggplot2 for creating graphs, reshape2 for converting the data frame to long format for multiple variables plotting, and sf for geospatial simple features.
3. Results
a. Historical and future trends
1) Temperature and heatwave indices
The time series of the annual coldest nights (TNn) and hottest days (TXx) presented in Fig. 2 indicate a consistent warming pattern in the extremes of minimum and maximum temperatures during the 1950–2100 period, except for the climate simulations associated with the ssp126 and ssp245 future scenarios, for which a lesser warming is unsurprisingly anticipated toward the end of the current century. Indeed, ssp126 depicts the optimistic shared socioeconomic pathway SSP1 combined with the low representative concentration pathway RCP2.6, and ssp245 represents the intermediate SSP2 combined with the medium emission RCP4.5 scenario. The long-term linear trend analysis presented in Table 3 reveals generally significant positive trends in both TXx and TNn, with a strong agreement among the models. Further, all the increasing (and decreasing) trends detected from the individual GCM-driven TNn and TXx indices during the historical period are anticipated to continue (become positive) into the future. However, compared to the historical trends, the slopes of the future trends, shown in curly braces, are expected to at least double for these indices with respect to the high emission scenarios in general (i.e., ssp370, ssp585). For instance, this is translated to a decadal increase of 0.5°C (and of 0.3°C) over the period 1950–2014 in the TNn (TXx) annual time series as driven by the BCC-CSM2-MR model, and the rate of change over the 2015–2100 period is projected to be a 1°C (0.6°C) increase every decade based on the BCC-CSM2-MR_ssp370 scenario. Despite small differences between the driven models, the trend slopes also highlight that the annual lowest daily minimum temperatures are projected to warm up faster than the annual highest daily maximum temperature.
Time series from 1950 to 2100 of the GCM-driven annual coldest nights (TNn) and hottest days (TXx) indices for the Alberta oil sands region. In the figure, each index encompasses six subplots that depict the patterns driven by the individual 12 climate simulations evaluated in this study. The future projections start in 2015 and end in 2100, and when the climate change scenarios have a common GCM, they share the same historical data that cover the 1950–2014 period. In each subplot, the lower (and higher) emission scenario is shown in green (orange) color.
Citation: Journal of Hydrometeorology 25, 1; 10.1175/JHM-D-23-0051.1
Linear trend slopes for the temperature and induced hazard indices, as driven by the individual climate simulations over the historical and future periods. Bold (regular) slope values indicate statistically significant (nonsignificant) trends, with respect to the 5% significance level.
The persistent warming pattern of the temperatures is also reflected in the annual total duration of the warm and cold spells indices WSDI and CSDI (Fig. 3). The individual models project in general an overall absence (and an increase in the duration) of the cold (warm) spell events into the future, with again lower intensity from models associated with the SSP126 and SSP245 scenarios. The historical trends for the CSDI show significant negative slope across all the GCM except for the BCC-CSM2-MR and GFDL-ESM4 models for which no trend was detected (Table 3). The duration of the cold spells has diminished for up to 3 days per decade for some models over the 1950–2014 period, leading to fewer or no cold spells during the studied future period, and thus the absence of a linear trend over the 2015–2100 period from all the climate projections. The WSDI depicts an opposite behavior moving from an absence of trend in its historical annual time series, to the presence of significant (except from the driven IPSL-CM6A-LR_ssp126 model) and positive trends from all the driven future scenarios. The highest increases projected in the duration of the warm spells are 10, 14, and 17 days per decade over the 2015–2100 period (Table 3), from the GFDL-ESM4_ssp585, IPSL-CM6A-LR_ssp370, and CNRM-CM6-1_ssp585 future scenarios, respectively.
As in Fig. 2, but for the warm and cold spells annual duration indices, WSDI and CSDI.
Citation: Journal of Hydrometeorology 25, 1; 10.1175/JHM-D-23-0051.1
The heatwave number (HWN), duration (HWD), and frequency (HWF) aspects are presented in Fig. 4 and show that all the climate projections reveal more heatwaves over the future period that will last longer and increase in frequency, compared to the historical period. The CNRM-CM6-1_ssp585 and IPSL-CM6A-LR_ssp370 scenarios, which project the highest increases in the WSDI, also suggest the most substantial increase in terms of the heatwave duration and frequency.
As in Fig. 2, but for the annual heatwaves number (HWN), duration (HWD), and frequency (HWF).
Citation: Journal of Hydrometeorology 25, 1; 10.1175/JHM-D-23-0051.1
Table 3 reveals, overall, an absence of trend in the HWN historical time series, and weak to no trend in the HWD ones, but significant positive trends over the future period for both indices regardless of the climate simulation. For instance, a trend slope of 6 days per decade is estimated over the 2015–2100 period for the duration of the annual longest heatwaves as driven by the CNRM-CM6-1_ssp585 and IPSL-CM6A-LR_ssp370 scenarios. With respect to the frequency of the heatwaves, persistent significant positive trends are observed during the entire study period from all the climate models except the EC-Earth3-Veg, GFDL-ESM4, and IPSL-CM6A-LR historical runs, for which no trend was detected. The trend slope in HWF is estimated to 17 days each decade over the 2015–2100 period, based on the CNRM-CM6-1_ssp585 and IPSL-CM6A-LR_ssp370 scenarios. Even though the annual number of individual heatwaves is projected not to increase much into the future compared to the historical period, more consecutive days showing positive EHF values are anticipated as shown by the HWF index, and consequently this will also affect the HWD index. In fact, individual heatwave events tend to spread out over a long period of time, which may limit their number but not their severity.
The annual time series of the seven temperature-related indices depict clear warming patterns for the Alberta oil sands region, with, on the one hand, extreme cold events diminishing and becoming less intense as shown by the future projections for the CSDI, and on the other hand, extreme heat becoming more intense and frequent as depicted by the frequency aspect of the heatwaves. Moreover, the climate models tend in general to agree on the trend direction of the temperature indices annual time series evaluated in this study.
2) Precipitation and water balance indices
The annual time series of the precipitation and water balance selected indices show no clear increasing or decreasing pattern during the study period as indicated in Figs. 5–7. The climate simulations overall depict comparable projections for the PRCPTOT and CDD indices, regardless of a low, medium, or high emission scenario (Fig. 5). Indeed, the trend analysis reveals nonsignificant increasing trends for PRCPTOT from all the driven-GCM during the historical period, and a mix of significant and nonsignificant positive trends over the future period, except for the BCC-CSM2-MR_ssp126 and BCC-CSM2-MR_ssp370 simulations, for which nonsignificant decreasing trends are found (Table 4). The significant positive trends in the PRCPTOT index are usually associated with the high emission scenarios, which suggest that as temperatures warm, more precipitation should be expected, and intuitively, more rainfall than snowfall based on the impact this has on the mountain-fed hydrologic systems. In terms of trends in the CDD annual time series, they are overall weak or absent during the entire study period.
Time series from 1950 to 2100 of the GCM-driven annual total precipitation on wet days (PRCPTOT) and longest spell of consecutive dry days (CDD) indices, for the Alberta oil sands region. In the figure, each index encompasses six subplots that depict the patterns driven by the individual 12 climate simulations evaluated in this study. The future projections start in 2015 and end in 2100, and when the climate change scenarios have a common GCM, they share the same historical data that cover the 1950–2014 period. In each subplot, the lower (and higher) emission scenario is shown in green (orange) color.
Citation: Journal of Hydrometeorology 25, 1; 10.1175/JHM-D-23-0051.1
As in Fig. 5, but for the contribution from very wet days (R95pTOT) and the maximum 5-day precipitation total (RX5day) indices.
Citation: Journal of Hydrometeorology 25, 1; 10.1175/JHM-D-23-0051.1
As in Fig. 5, but for the standardized precipitation evapotranspiration index at 3 and 12 months in August (Aug_SPEI3) and September (Sep_SPEI12), respectively.
Citation: Journal of Hydrometeorology 25, 1; 10.1175/JHM-D-23-0051.1
Similar small variations over time are observed with the indices R95pTOT and RX5day (Fig. 6). Table 4 also depicts in general a mix of positive and negative nonsignificant trends in these indices.
Overall, for the Alberta oil sands region, small changes have occurred during the 1950–2100 period in the GCM-driven annual amount of precipitation as depicted by the PRCPTOT, R95pTOT, and RX5day indices, but also in the annual longest multi-dry-days index CDD. Consequently, the Aug_SPEI3 and Sep_SPEI12 water budget indices reveal a decreasing pattern but with a weak signal regardless of the climate model (Fig. 7). The hydrologically drier conditions observed in these indices and corresponding to the SPEI values below 0 tend to be dominant into the future. Table 4 also shows for both indices more negative than positive trends, which are overall nonsignificant regarding the individual simulations. The negative (and positive) trend in the SPEI implies the predominance of negative (positive) values. However, only a few occurrences of extremely dry hydrologically conditions (SPEI ≤ −2) are recorded and mainly into the future in association with the high emission scenarios.
Even though the SPEI is not a streamflow index, it influences the water input (i.e., runoff) to the streamflow network, because increasing evapotranspiration, if not adequately balanced by sufficient precipitation, could reduce regional water availability and trigger drought episodes. However, we recall that trends in streamflow for mountain headwater basins like the Athabasca River will be dominated by headwater runoff generation processes (i.e., glacier melt), while regional changes in runoff (as indicated by SPEI) may not be as well reflected in the streamflow record. Nevertheless, with an ongoing warming climate in the Canadian cold regions, mountain-fed hydrologic systems are likely to be modified (e.g., transition from snowfall to rainfall), and indices like the SPEI can capture and reflect these changes, enhancing the understanding of the streamflow patterns.
b. Projected future changes
1) Temperature indices
Figure 8 presents the distributions of the projected annual changes over the near future 2041–70 and far future 2071–2100 for the temperature-related indices, with respect to the individual 12 climate simulations and the baseline 1985–2014. The medians show, on the one hand, values above zero for the annual coldest nights (TNn) and hottest days (TXx) indices, and the annual duration of the warm spells (WSDI). On the other hand, median values are equal to or below zero for the annual duration of the cold spells (CSDI), suggesting that temperatures will warm up during the future time periods evaluated. For the TNn and TXx indices, temperatures will be as expected, even warmer over the far-future period than the near-future period, with respect to the baseline conditions and from most of the driven climate scenarios. Indeed, for the TNn index, a smaller warming over the far future is projected only from the CNRM-CM6-1_ssp126 scenario, and for the TXx index, the exception is associated with the EC-Earth3-Veg_ssp126 and IPSL-CM6A-LR_ssp126 scenarios.
Distributions of the projected annual changes (Δ) in the temperature indices over the near future 2041–70 and the far future 2071–2100 periods, for the Alberta oil sands region. On the x axis, the numbering represents the climate simulations as follows: 1) BCC-CSM2-MR_ssp126; 2) BCC-CSM2-MR_ssp370; 3) CNRM-CM6-1_ssp126; 4) CNRM-CM6-1_ssp585; 5) EC-Earth3-Veg_ssp126; 6) EC-Earth3-Veg_ssp370; 7) GFDL-CM4_ssp245; 8) GFDL-ESM4_ssp585; 9) IPSL-CM6A-LR_ssp126; 10) IPSL-CM6A-LR_ssp370; 11) MRI-ESM2-0_ssp370; 12) MRI-ESM2-0_ssp585. For each climate change scenario, the near (and far) future is color-coded in blue (orange) and are side by side. For the boxplots, the box, when there is one, shows the data range between the 25th (Q25) and 75th (Q75) percentiles, also known as the interquartile range (IQR). The solid horizontal line within the box denotes the median, the lower whisker limit equals Q25 − 1.5 × IQR, and the upper whisker Q75 + 1.5 × IQR; the dots beyond the whiskers indicate the outliers.
Citation: Journal of Hydrometeorology 25, 1; 10.1175/JHM-D-23-0051.1
In terms of annual changes in the CSDI time series, all the scenarios agree in projecting similar median patterns over both future periods, either an absence of change or a diminution in the duration of these events, except for the IPSL-CM6A-LR_ssp126 (i.e., a diminution, and then, no change) and IPSL-CM6A-LR_ssp370 (i.e., no change, and then, a diminution) simulations, which highlight minor differences. The medians of the WSDI projected annual change distributions indicate that they are also in general anticipated to last longer into the far future compared to the near-future expected changes. Indeed, only the CNRM-CM6-1_ssp126 and EC-Earth3-Veg_ssp126 scenarios project similar or smaller duration of these events in the far future.
From the change distributions driven by the GCM with high emission scenarios (i.e., ssp370, ssp585) and shown in Fig. 8, an average increase of between 4° and 9°C (versus 9° and 17°C) is expected in the TNn annual time series during the 2041–70 (2071–2100) period, relative to these scenarios and the 1985–2014 baseline conditions. For the TXx index, the projected average increase is between 1.6° and 4°C for the near future, and between 2.9° and 7°C for the far future, resulting in extreme cold temperatures warming at least 2 times more than the extreme hot temperatures. The near-future projection for WSDI is expected to double in the far future, gaining an average of between 43 and 124 days compared to the baseline conditions and with respect to the high emission scenarios. As also shown in the trend analysis (Table 3), future changes in CSDI are projected to remain overall the same over both the near- and far-future periods, with an average diminution in baseline records of less than 10 days across climate models.
2) Precipitation indices
The projected future change distributions for the precipitation-related indices reveal medians slightly above zero from most of the climate simulations (Fig. 9), suggesting that there will be some increases, though at a small rate, in the annual total precipitation on wet days, the annual highest amount of precipitation that falls in five consecutive days, and the annual contribution to total precipitation from very wet days, but also in the annual longest dry spells.
As in Fig. 8, but for the precipitation indices.
Citation: Journal of Hydrometeorology 25, 1; 10.1175/JHM-D-23-0051.1
In Fig. 9, all the high emission scenarios except the IPSL-CM6A-LR_ssp370 suggest a continuous increase in PRCPTOT, with average relative changes ranging between 2% and 4% during the period 2041–70, and between 12% and 23% into the far future, relative to the baseline conditions. The R95pTOT index depicts the same patterns as PRCPTOT, and its projected future changes suggest that the average increase will not exceed 5% (and 10%) over the near (far) future period. The diminution and increase reflected in the CDD index are projected to not exceed on average 5 days during either of the 30-yr future time slices.
Across all the climate models, the average increase in the annual highest amount of precipitation falling in 5 consecutive days relative to the baseline is projected between 6% and 24% for the 2041–70 period, which is similar to the far future projections (i.e., 4% and 26%).
3) Heatwave and water balance indices
Figure 10 presents the projected future change distributions for the heatwaves and water balance indices. The GCM-driven median annual changes for HWN, HWF, and HWD are all above zero, indicating an increase in the extreme heat episodes within May to September.
As in Fig. 8, but for the heatwave and water balance indices.
Citation: Journal of Hydrometeorology 25, 1; 10.1175/JHM-D-23-0051.1
In the case of the water balance indices for the summer and water-year periods, the future scenarios generally project median annual changes below the normal moisture condition (i.e., zero), suggesting deficits in local water availability. A notable exception is with the EC-Earth3-Veg_ssp370 scenario, for which a surplus is projected into the far future period.
The annual changes driven by the climate simulations for the heatwave aspects suggest an average increase between three and five events for the HWN during the near future, versus four and seven into the far future relative to the baseline conditions (Fig. 10). Although these increases seem small, the projected patterns in the HWD index highlight that among the identified annual heatwaves, the longest will last even longer, increasing on average over the period 2071–2100, by at least (and at most) 13 (61) days above the baseline value with respect to the future projection. Consequently, this implies that more extreme heat days, as defined by the excess heat factor, will contribute to each identified annual heatwave, resulting in individual long heatwave events but not necessarily an increase in their annual number. The pattern of the HWF index illustrates this very well, with an average projected increase of between 58 and 113 days in the total number of days contributing to all heatwave events during the far future, relative to the climate model and the baseline conditions.
The annual projected changes in the water balance indices Aug_SPEI3 and Sep_SPEI12 (Fig. 10) suggest, on average, deficits in the regional water availability over the summer and water-year period, with the driest conditions being associated with the GFDL-ESM4_ssp585 and IPSL-CM6A-LR_ssp370 scenarios during the far-future period compared to the baseline. Only the EC-Earth3-Veg_ssp370 scenario projects surplus into the far future 2071–2100. With respect to the high emission scenario, the deficit is higher over the water-year than the summer period, except with the MRI-ESM2-0_ssp370 and MRI-ESM2-0_ssp585 simulations, for which the opposite is anticipated. Given that the warming of temperatures is persistent and patterns in precipitation do not show strong increases, using water budget indices like the SPEI appear to be well suited here to assess deficits or surpluses in water availability for the Alberta oil sands region.
4. Discussion
The hydroclimate indicators annual time series examined in this study reveal persistent warming of extreme temperatures and small changes in precipitation-related indices, throughout the 1950–2100 period and for the region encompassed by the Alberta oil sands areas. Such patterns raise concerns for freshwater availability, because on the one hand, as pointed out by Bonsal et al. (2013), across the Canadian Prairies, it is the combined effect of increasing temperature and slight changes in precipitation that will likely increase both the occurrence and severity of summer droughts over time. On the other hand, ongoing warmer temperatures are very likely and irreversibly increasing the changes to the cryosphere, including the reduction in the coming decades of glacier mass in western Canada and even their disappearance by the end of this century (Bush et al. 2022). Further, in regions where freshwater is primarily supplied from snowpacks and/or glacier melt, increasing temperatures cause earlier onset of melt-based runoff generation in spring, thus potentially affecting the availability of sufficient water supply over the mid- and late summer and autumn seasons (Bonsal et al. 2019; Barnett et al. 2005). For instance, Eum et al. (2017) reported an average projected decrease in summer flows by up to 21% for the Athabasca River basin in the 2080s relative to the 1990s. They suggested that these anticipated changes are largely induced by the combined effect of earlier snowmelt, increased evapotranspiration, and no significant increase in summer precipitation. Moreover, the climate warming in the mountain-fed hydrologic systems leads to the occurrence of more rainfall than snowfall (Jiang et al. 2017), changing precipitation regimes, challenging the recharge of freshwater sources, and impacting the river flows and lake levels (Carlson et al. 2021; Schindler and Donahue 2006).
The linear trend analysis and the projected changes assessment show that extreme cold temperatures and spells, depicted here by TNn and CSDI, are becoming less intense and diminishing. This is partly the consequence of the extreme hot temperatures becoming more intense and frequent, as reflected herein by the WSDI and heatwave aspects. Increasing (and decreasing) linear trends for the TXx, WSDI, and TNn indices (CSDI) were also detected in large Canadian cities during the 2020–2100 period by Rajulapati et al. (2022). The highest warming trends, ranging between 4° and 6°C, were also detected in winter and spring annual mean temperature over the 1948–2012 period across western Canada (Vincent et al. 2015). These unusually hot temperatures have implications on evapotranspiration, water supply shortages, and drought risks, among other resulting impacts (Bush et al. 2022; Burhan et al. 2021; Bonsal and Cuell 2017).
Small rates of increase are projected in precipitation amounts, and as would be expected, the strong climate warming signals detected through the evaluated temperature indices may not be offset, suggesting an increased risk of drier conditions in the Alberta oil sands region. Jiang et al. (2017) reported that over 75% of the positive trends detected in seasonal precipitation for the 1900–2011 period in Alberta were statistically nonsignificant. The deficits projected in the regional water availability by the Aug_SPEI3 and Sep_SPEI12 indices are depicting this threat of drying conditions. Indeed, these insufficient moisture conditions are likely caused by a deficit in precipitation and/or an increase in potential evapotranspiration driven by warmer temperatures (Vicente-Serrano et al. 2010; McKee et al. 1993). Eum et al. (2023) showed that under the global mean temperature changes scenarios from 1° to 5°C, drought conditions are expected to become more frequent and intense in Alberta as temperatures warm up and snow accumulation is reduced by more rain-phase precipitation and earlier snowmelt. Dibike et al. (2017) also reported a gradual increase in the magnitude and duration of water deficits over the water-year for the Columbia, Saskatchewan, Fraser, and Athabasca basins from 1950 to 2100. In their far future period (i.e., 2071–2100), a decrease in summer water availability for almost all river basins across western Canada is projected. Leong and Donner (2015) indicated that the occurrence of low summer flows in the Athabasca River basin is projected to increase by up to 85% by 2100, potentially declining water availability.
The climate simulations used to derive the hydroclimate indicators evaluated in this study show a general agreement with respect to the warming up of extreme temperatures and the minor change in precipitation amounts for the Alberta oil sands region. The warming of Canada’s climate is expected to keep occurring faster than the average global warming rate of change, largely based on the carbon dioxide emissions induced by human activities (Bush et al. 2022; IPCC 2022; Bush and Lemmen 2019). Since the oil sands operations require large amounts of water and energy but also lead to high greenhouse gases emissions (Wren et al. 2023), the findings of the present study show that temperatures are very likely warming up and will likely contribute to reduced water availability across the year as small changes are expected in annual total precipitation. The Alberta oil sands monitoring program may build upon the framework implemented for this study and use the set of hydroclimate indices proposed to develop, update, and improve adaptive strategies for effective climate actions that support a sustainable management of Alberta’s natural resources.
5. Conclusions
The assessment of climate risks for various sectors includes analyzing and monitoring observed and projected patterns in climate variables such as temperature and precipitation to detect and enhance our understanding of climate changes for global and regional scales. In this study, we have examined for the Alberta oil sands region the historical and future temporal patterns in hydroclimatic annual conditions using 13 hydroclimate indicators as simulated by 12 individual climate simulations.
Our study reveals that projected extreme temperatures are getting hotter into the future compared to the established baseline conditions, whereas little changes are anticipated for precipitation events. Since freshwater water in the Alberta oil sands region is primarily supplied from snowmelt and the melting of glaciers, as it is for the nival hydrological regimes, such findings suggest that water availability will likely continue to be threatened as temperature warming persists, impacting human activities and ecosystems integrity. In fact, warmer temperatures in cold regions that persist almost year-round can reduce winter snowfall, trigger earlier and faster freshet, lead to low summer flows, and impact the effective supplies of water to all users throughout the year. Moreover, small increases in precipitation are unlikely to offset the potential increased evapotranspiration under a strong warming climate. Water scarcity and droughts are already a concern in the Canadian Prairies, and with an expected warmer and drier climate, perspectives on expanding human activities should continually integrate the reality of climate change to accommodate all water users.
Acknowledgments.
We acknowledge the technical support from the Resource Stewardship Division of Alberta Environment and Protected Areas (AEPA), and the financial contributions of AEPA’s Oil Sands Monitoring Program as well as from Natural Sciences and Engineering Research Council of Canada.
Data availability statement.
The climate datasets analyzed in this study were developed by the Research Stewardship Division of Alberta’s Environment and Protected Areas for Alberta impact studies and can be shared upon request. The ClimPACT2 open-source software used to calculate the hydroclimate indicators is archived at https://github.com/ARCCSS-extremes/climpact. We compile all developed R scripts into a workflow and can make it available upon request.
REFERENCES
Alexander, L., and N. Herold, 2016: ClimPACT2: Indices and software. World Climate Programme Doc., 46 pp., https://epic.awi.de/id/eprint/49274/1/ClimPACTv2_manual.pdf.
Alexander, L. V., and Coauthors, 2006: Global observed changes in daily climate extremes of temperature and precipitation. J. Geophys. Res., 111, D05109, https://doi.org/10.1029/2005JD006290.
Avila, F. B., S. Dong, K. P. Menang, J. Rajczak, M. Renom, M. G. Donat, and L. V. Alexander, 2015: Systematic investigation of gridding-related scaling effects on annual statistics of daily temperature and precipitation maxima: A case study for South-East Australia. Wea. Climate Extremes, 9, 6–16, https://doi.org/10.1016/j.wace.2015.06.003.
Barnett, T. P., J. C. Adam, and D. P. Lettenmaier, 2005: Potential impacts of a warming climate on water availability in snow-dominated regions. Nature, 438, 303–309, https://doi.org/10.1038/nature04141.
Bonsal, B., R. R. Shrestha, Y. Dibike, D. L. Peters, C. Spence, L. Mudryk, and D. Yang, 2020: Western Canadian freshwater availability: Current and future vulnerabilities. Environ. Rev., 28, 528–545, https://doi.org/10.1139/er-2020-0040.
Bonsal, B. R., and C. Cuell, 2017: Hydro-climatic variability and extremes over the Athabasca River basin: Historical trends and projected future occurrence. Can. Water Resour. J., 42, 315–335, https://doi.org/10.1080/07011784.2017.1328288.
Bonsal, B. R., E. E. Wheaton, A. C. Chipanshi, C. Lin, D. J. Sauchyn, and L. Wen, 2011: Drought research in Canada: A review. Atmos.–Ocean, 49, 303–319, https://doi.org/10.1080/07055900.2011.555103.
Bonsal, B. R., R. Aider, P. Gachon, and S. Lapp, 2013: An assessment of Canadian prairie drought: Past, present, and future. Climate Dyn., 41, 501–516, https://doi.org/10.1007/s00382-012-1422-0.
Bonsal, B. R., D. L. Peters, F. Seglenieks, A. Rivera, and A. Berg, 2019: Changes in freshwater availability across Canada. Canada’s Changing Climate Report, E. Bush and D. S. Lemmen, Eds., Government of Canada, 261–342, https://changingclimate.ca/CCCR2019/chapter/6-0/.
Braun, M., A. N. Thiombiano, M. J. F. Vieira, and T. A. Stadnyk, 2021: Representing climate evolution in ensembles of GCM simulations for the Hudson Bay system. Elementa, 9, 00011, https://doi.org/10.1525/elementa.2021.00011.
Burhan, A. K., A. H. Khan, S. A. A. Bukhari, and K. Riaz, 2021: Univariate statistics of the RCPs forced ET–SCI based extreme climate indices over Pakistan. Eur. J. Sustainable Dev. Res., 5, em0166, https://doi.org/10.21601/ejosdr/11091.
Bush, E., and D. S. Lemmen, 2019: Canada’s Changing Climate Report. Government of Canada, 444 pp., https://changingclimate.ca/CCCR2019/chapter/6-0/.
Bush, E., N. Gillett, E. Watson, E. Fyfe, F. Vogel, and N. Swart, 2019: Understanding observed global climate change. Canada’s Changing Climate Report, E. Bush and D. S. Lemmen, Eds., Government of Canada, 25–72, https://changingclimate.ca/CCCR2019/chapter/2-0/.
Bush, E., and Coauthors, 2022: Canada’s changing climate report in light of the latest global science assessment. Government of Canada, 36 pp., https://changingclimate.ca/site/assets/uploads/sites/2/2022/03/CCCR-2022-Supplement-Final.pdf.
Cannon, A. J., 2015: Selecting GCM scenarios that span the range of changes in a multimodel ensemble: Application to CMIP5 climate extremes indices. J. Climate, 28, 1260–1267, https://doi.org/10.1175/JCLI-D-14-00636.1.
Carlson, H., A. Pietroniro, P. Gober, W. Leger, and S. Merrill, 2021: Water resources. Canada in a Changing Climate: National Issues Report, F. J. Warren and N. Lulham, Eds., Government of Canada, 186–263, https://changingclimate.ca/national-issues/chapter/4-0/.
Cohen, S., G. Koshida, and L. Mortsch, 2015: Climate and water availability indicators in Canada: Challenges and a way forward. Part III – Future scenarios. Can. Water Resour. J., 40, 160–172, https://doi.org/10.1080/07011784.2015.1006021.
Derksen, C., D. Burgess, C. Duguay, S. Howell, L. Mudryk, S. Smith, C. Thackeray, and M. Kirchmeier-Young, 2019: Changes in snow, ice, and permafrost across Canada. Canada’s Changing Climate Report, E. Bush and D. S. Lemmen, Ed., Government of Canada, 194–260, https://changingclimate.ca/CCCR2019/chapter/5-0/.
Dibike, Y., T. Prowse, B. Bonsal, and H. O’Neil, 2017: Implications of future climate on water availability in the western Canadian river basins. Int. J. Climatol., 37, 3247–3263, https://doi.org/10.1002/joc.4912.
Droogers, P., and R. G. Allen, 2002: Estimating reference evapotranspiration under inaccurate data conditions. Irrig. Drain. Syst., 16, 33–45, https://doi.org/10.1023/A:1015508322413.
Dubé, M. G., J. M. Dunlop, C. Davidson, D. L. Beausoleil, R. R. O. Hazewinkel, and F. Wyatt, 2022: History, overview, and governance of environmental monitoring in the oil sands region of Alberta, Canada. Integr. Environ. Assess. Manage., 18, 319–332, https://doi.org/10.1002/ieam.4490.
Eum, H.-I., and A. Gupta, 2019: Hybrid climate datasets from a climate data evaluation system and their impacts on hydrologic simulations for the Athabasca River basin in Canada. Hydrol. Earth Syst. Sci., 23, 5151–5173, https://doi.org/10.5194/hess-23-5151-2019.
Eum, H.-I., Y. Dibike, and T. Prowse, 2017: Climate-induced alteration of hydrologic indicators in the Athabasca River basin, Alberta, Canada. J. Hydrol., 544, 327–342, https://doi.org/10.1016/j.jhydrol.2016.11.034.
Eum, H.-I., A. Gupta, and Y. Dibike, 2020: Effects of univariate and multivariate statistical downscaling methods on climatic and hydrologic indicators for Alberta, Canada. J. Hydrol., 588, 125065, https://doi.org/10.1016/j.jhydrol.2020.125065.
Eum, H.-I., B. Fajard, T. Tang, and A. Gupta, 2023: Potential changes in climate indices in Alberta under projected global warming of 1.5°–5°C. J. Hydrol. Reg. Stud., 47, 101390, https://doi.org/10.1016/j.ejrh.2023.101390.
Eyring, V., S. Bony, G. A. Meehl, C. A. Senior, B. Stevens, R. J. Stouffer, and K. E. Taylor, 2016: Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev., 9, 1937–1958, https://doi.org/10.5194/gmd-9-1937-2016.
Farjad, B., A. Gupta, H. Sartipizadeh, and A. J. Cannon, 2019: A novel approach for selecting extreme climate change scenarios for climate change impact studies. Sci. Total Environ., 678, 476–485, https://doi.org/10.1016/j.scitotenv.2019.04.218.
Hargreaves, G. H., 1994: Defining and using reference evapotranspiration. J. Irrig. Drain. Eng., 120, 1132–1139, https://doi.org/10.1061/(ASCE)0733-9437(1994)120:6(1132).
IPCC, 2021: Annex II: Models. Climate Change 2021: The Physical Science Basis, V. Masson-Delmotte et al., Eds., Cambridge University Press, 2087–2138, https://doi.org/10.1017/9781009157896.018.
IPCC, 2022: Summary for policymakers. Climate Change 2022: Impacts, Adaptation and Vulnerability, H.-O. Pörtner et al., Eds., Cambridge University Press, 3–33, https://doi.org/10.1017/9781009325844.001.
Jiang, R., T. Y. Gan, J. Xie, N. Wang, and C.-C. Kuo, 2017: Historical and potential changes of precipitation and temperature of Alberta subjected to climate change impact: 1900–2100. Theor. Appl. Climatol., 127, 725–739, https://doi.org/10.1007/s00704-015-1664-y.
Jordaan, S. M., 2012: Land and water impacts of oil sands production in Alberta. Environ. Sci. Technol., 46, 3611–3617, https://doi.org/10.1021/es203682m.
Kendall, M. G., 1975: Rank Correlation Methods. 4th ed. Griffin, 202 pp.
Koshida, G., S. Cohen, and L. Mortsch, 2015: Climate and water availability indicators in Canada: Challenges and a way forward. Part I – Indicators. Can. Water Resour. J., 40, 133–145, https://doi.org/10.1080/07011784.2015.1006023.
Leong, D. N. S., and S. D. Donner, 2015: Climate change impacts on streamflow availability for the Athabasca oil sands. Climatic Change, 133, 651–663, https://doi.org/10.1007/s10584-015-1479-y.
Lunn, S., 2013: Water use in Canada’s oil-sands industry: The facts. SPE Econ. Manage., 5, 17–27, https://doi.org/10.2118/156676-PA.
Mann, H. B., 1945: Nonparametric tests against trend. Econometrica, 13, 245–259, https://doi.org/10.2307/1907187.
McKee, T. B., N. J. Doesken, and J. Kleist, 1993: The relationship of drought frequency and duration to time scales. Eighth Conf. on Applied Climatology, Anaheim, CA, Amer. Meteor. Soc., 179–184, https://climate.colostate.edu/pdfs/relationshipofdroughtfrequency.pdf.
Mortsch, L., S. Cohen, and G. Koshida, 2015: Climate and water availability indicators in Canada: Challenges and a way forward. Part II – Historic trends. Can. Water Resour. J., 40, 146–159, https://doi.org/10.1080/07011784.2015.1006024.
Nairn, J. R., and R. J. B. Fawcett, 2015: The excess heat factor: A metric for heatwave intensity and its use in classifying heatwave severity. Int. J. Environ. Res. Public Health, 12, 227–253, https://doi.org/10.3390/ijerph120100227.
Nenzén, H. K., D. T. Price, Y. Boulanger, A. R. Taylor, D. Cyr, and E. Campbell, 2020: Projected climate change effects on Alberta’s boreal forests imply future challenges for oil sands reclamation. Restor. Ecol., 28, 39–50, https://doi.org/10.1111/rec.13051.
Newton, B. W., B. Farjad, and J. F. Orwin, 2021: Spatial and temporal shifts in historic and future temperature and precipitation patterns related to snow accumulation and melt regimes in Alberta, Canada. Water, 13, 1013, https://doi.org/10.3390/w13081013.
OSMP, 2019: Oil sands monitoring program: Annual report for 2018–19, 123 pp., https://open.alberta.ca/publications/2562-9182.
Perkins, S. E., and L. V. Alexander, 2013: On the measurement of heat waves. J. Climate, 26, 4500–4517, https://doi.org/10.1175/JCLI-D-12-00383.1.
Prowse, T. D., and F. M. Conly, 2002: A review of hydroecological results of the Northern River Basins Study, Canada. Part 2. Peace-Athabasca delta. River Res. Appl., 18, 447–460, https://doi.org/10.1002/rra.682.
Rajulapati, C. R., R. K. Gaddam, S. D. Nerantzaki, S. M. Papalexiou, A. J. Cannon, and M. P. Clark, 2022: Exacerbated heat in large Canadian cities. Urban Climate, 42, 101097, https://doi.org/10.1016/j.uclim.2022.101097.
Riahi, K., and Coauthors, 2017: The shared socioeconomic pathways and their energy, land use, and greenhouse gas emissions implications: An overview. Global Environ. Change, 42, 153–168, https://doi.org/10.1016/j.gloenvcha.2016.05.009.
Rood, S. B., G. W. Stupple, and K. M. Gill, 2015: Century-long records reveal slight, ecoregion-localized changes in Athabasca River flows. Hydrol. Processes, 29, 805–816, https://doi.org/10.1002/hyp.10194.
Sauchyn, D. J., J.-M. St-Jacques, and B. H. Luckman, 2015: Long-term reliability of the Athabasca River (Alberta, Canada) as the water source for oil sands mining. Proc. Natl. Acad. Sci. USA, 112, 12 621–12 626, https://doi.org/10.1073/pnas.1509726112.
Schindler, D. W., and W. F. Donahue, 2006: An impending water crisis in Canada’s western prairie provinces. Proc. Natl. Acad. Sci. USA, 103, 7210–7216, https://doi.org/10.1073/pnas.0601568103.
Sen, P. K., 1968: Estimates of the regression coefficient based on Kendall’s Tau. J. Amer. Stat. Assoc., 63, 1379–1389, https://doi.org/10.1080/01621459.1968.10480934.
Sillmann, J., V. V. Kharin, X. Zhang, F. W. Zwiers, and D. Bronaugh, 2013a: Climate extremes indices in the CMIP5 multimodel ensemble: Part 1. Model evaluation in the present climate. J. Geophys. Res. Atmos., 118, 1716–1733, https://doi.org/10.1002/jgrd.50203.
Sillmann, J., V. V. Kharin, F. W. Zwiers, X. Zhang, and D. Bronaugh, 2013b: Climate extremes indices in the CMIP5 multimodel ensemble: Part 2. Future climate projections. J. Geophys. Res. Atmos., 118, 2473–2493, https://doi.org/10.1002/jgrd.50188.
Stadnyk, T. A., and S. J. Déry, 2021: Canadian continental-scale hydrology under a changing climate: A review. Water, 13, 906, https://doi.org/10.3390/w13070906.
Stadnyk, T. A., and Coauthors, 2021: Changing freshwater contributions to the Arctic: A 90-year trend analysis (1981–2070). Elementa, 9, 00098, https://doi.org/10.1525/elementa.2020.00098.
van Vuuren, D. P., and Coauthors, 2011: The representative concentration pathways: An overview. Climatic Change, 109, 5, https://doi.org/10.1007/s10584-011-0148-z.
Vicente-Serrano, S. M., S. Beguería, and J. I. López-Moreno, 2010: A multiscalar drought index sensitive to global warming: The standardized precipitation evapotranspiration index. J. Climate, 23, 1696–1718, https://doi.org/10.1175/2009JCLI2909.1.
Vincent, L. A., X. Zhang, R. D. Brown, Y. Feng, E. Mekis, E. J. Milewska, H. Wan, and X. L. Wang, 2015: Observed trends in Canada’s climate and influence of low-frequency variability modes. J. Climate, 28, 4545–4560, https://doi.org/10.1175/JCLI-D-14-00697.1.
Westman, C. N., and T. L. Joly, 2019: Oil sands extraction in Alberta, Canada: A review of impacts and processes concerning indigenous peoples. Hum. Ecol., 47, 233–243, https://doi.org/10.1007/s10745-019-0059-6.
Wren, S. N., and Coauthors, 2023: Aircraft and satellite observations reveal historical gap between top–down and bottom–up CO2 emissions from Canadian oil sands. Proc Natl. Acad. Sci. Nexus, 2, pgad140, https://doi.org/10.1093/pnasnexus/pgad140.
Yue, S., P. Pilon, B. Phinney, and G. Cavadias, 2002: The influence of autocorrelation on the ability to detect trend in hydrological series. Hydrol. Processes, 16, 1807–1829, https://doi.org/10.1002/hyp.1095.
Zhang, X., G. Hegerl, F. W. Zwiers, and J. Kenyon, 2005: Avoiding inhomogeneity in percentile-based indices of temperature extremes. J. Climate, 18, 1641–1651, https://doi.org/10.1175/JCLI3366.1.
Zhang, X., L. Alexander, G. C. Hegerl, P. Jones, A. K. Tank, T. C. Peterson, B. Trewin, and F. W. Zwiers, 2011: Indices for monitoring changes in extremes based on daily temperature and precipitation data. Wiley Interdiscip. Rev.: Climate Change, 2, 851–870, https://doi.org/10.1002/wcc.147.
Zhang, X., and Coauthors, 2019: Changes in temperature and precipitation across Canada. Canada’s Changing Climate Report, E. Bush and D. S. Lemmen, Eds., Government of Canada, 112–193, https://changingclimate.ca/CCCR2019/chapter/4-0/.
