ABSTRACT
This study aimed to investigate long-term (1969–2023) variability and trends in annual mean daily discharge (AMDD) and seasonal mean daily discharge (SMDD) in 10 rivers throughout Northern Finland in connection to climate teleconnections. The lowest AMDDs were mostly recorded during the first 12 years (1969–1980) of our study period, while the highest ones were during the last 12 years (2012–2023). Significant trends in AMDD were found only in three natural rivers of Tornionjoki, Simojoki, and Kuivajoki during 1969–2023. Such variations and trends in AMDD were most significantly associated with the Scandinavia (SCA) pattern, which is also an influential climate teleconnection for annual precipitation anomalies across Northern Finland. The highest (lowest) SMDDs were experienced in spring (winter). Only increasing trends in SMDD were statistically significant (p < 0.05). Such substantial increases in SMDDs were detected in winter, spring, and summer (autumn) in association with the Arctic Oscillation or AO and the North Atlantic Oscillation or NAO (East Atlantic/West Russia or EA/WR). Generally speaking, recent milder and wetter climatic conditions in association with strong positive AO and NAO (negative EA/WR) phases could increase SMDD for winter, spring, and summer (autumn) seasons. Hence, both AMDD and SMDD in rivers throughout Northern Finland were primarily influenced by precipitation.
HIGHLIGHTS
The Scandinavia (SCA) pattern affects annual mean daily river discharge in Northern Finland.
From winter to summer, daily river discharge in Northern Finland is associated with the Arctic Oscillation (OA).
The East Atlantic/West Russia (EA/WR) influences river discharge in Northern Finland during autumn.
Precipitation primarily controls both annual and seasonal mean daily river discharge in Northern Finland.
INTRODUCTION
Compared with 1850–1900, the Earth's mean global surface air temperature (SAT) was 1.1 °C warmer during 2011–2020 (IPCC 2021). Such a warming SAT was substantially higher over the lands (1.59 °C) than over the oceans (0.88 °C) (IPCC 2021). The primary cause of such increases in SAT is a rise in the atmospheric concentrations of greenhouse gas (GHG) emissions (IPCC 2021). This global warming has already influenced discharge regimes in rivers by substantially altering different hydrological cycle components, particularly precipitation, evapotranspiration, and runoff (Berghuijs et al. 2017; Konapala et al. 2020). Hence, evaluating historical river discharge time series has been one of the key study topics in international research communities focusing on climatic, hydrological, and environmental changes on local, regional, and global scales (Ashraf et al. 2018; Gudmundsson et al. 2021; Uvo et al. 2021).
At high latitudes, snowpack plays a key role in land surface hydrology by naturally storing water during winter and gradually releasing it into rivers in late spring and early summer as snowmelt (Jansson et al. 2003). Such meltwater from snowpack fundamentally controls the flow regime in northern rivers (Wilson et al. 2010). Hence, global warming and climate change extensively alter water resources in boreal environments by influencing snowmelt quantity and timing (Pohl et al. 2005). In general, warmer SAT reduces the number of cold days, the amount of snowfall, the accumulation of snowpack, and the snowmelt runoff, leading to significant changes in river discharge in cold climate regions, like Finland (Irannezhad et al. 2016, 2022b, 2024). Such impacts of climate warming on snow-dominated catchments, however, can be offset by more precipitation in winter, increasing snowfall, accumulating more snowpack, and delivering a higher rate of snowmelt runoff into northern rivers (Räisänen 2008). Hence, the river discharge regime at high latitudes principally depends upon changes in both regional SAT and precipitation patterns (Irannezhad et al. 2015b).
In general, variations in climatic conditions (particularly SAT and precipitation) across a region are influenced by large-scale atmospheric-oceanic circulation patterns, e.g. the North Atlantic Oscillation (NAO) (Glantz et al. 2009). The strength and natural influences of these patterns on regional climate variability are usually expressed by numeric teleconnection indices (hereafter climate teleconnections). Numerous previous studies have described the key features of such climate teleconnections (Glantz et al. 2009) and their effects on SAT, precipitation, and snowpack hydrological processes (SHPs) around the world (Bartolini et al. 2010; Hoy et al. 2013; Ghasemifar et al. 2022). In our previous studies, we mainly focused on the role of climate teleconnections on the historical variability and trends in precipitation (Irannezhad et al. 2014), SAT (Irannezhad et al. 2015a), and SHPs (Irannezhad et al. 2015b), but not much on river hydrology, in Finland. Although changes in the historical river discharge regime in Finland have already been evaluated (e.g., Korhonen & Kuusisto 2010; Gohari et al. 2022), our knowledge about the effects of climate teleconnections on such hydrological alterations is still limited to a few studies (e.g., Uvo et al. 2021; Irannezhad et al. 2022b).
The present study aimed to investigate the influence of climate teleconnections on historical changes in the river discharge regime in Northern Finland. The specific objectives were to: (1) analyze long-term (1969–2023) variability in annual (annual mean daily discharge (AMDD)) and seasonal mean daily discharge (SMDD) in 10 rivers (both natural and regulated) in Northern Finland; (2) determine statistically significant (p < 0.05) trends in such times series reflecting historical discharge regimes in northern Finnish rivers, and (3) measure the correlations of these times series with different large-scale climate teleconnections. The findings can lay a foundation for forecasting future daily river discharge in these rivers based on the coming positive and negative phases of influential climate teleconnections (e.g., Wanders & Wada 2015). Such prediction of daily river discharge can play an important role in the operational processes of regulated rivers and thereby in hydropower production (e.g., Engström & Uvo 2015) in Finland. Based on Irannezhad et al. (2022a), hence, such an observational-based study related to river discharge regimes at high latitudes can improve our knowledge about local/regional water and energy security and thereby act toward achieving the 2030 United Nations Agenda for Sustainable Development (UN 2015).
MATERIAL AND METHODS
Study area
The rivers in Finland are generally divided into three groups based on their flow regime (Korhonen & Kuusisto 2010). For the present study, accordingly, five natural (Torniojoki, Simojoki, Kuivajoki, Kiiminkijoki, and Temmesjoki) and five regulated (Kemijoki, Iijoki, Oulujoki, Siikajoki, and Pyhäjoki) rivers (Korhonen 2006) in Northern Finland were selected (Figure 1(b)). The River Oulujoki belongs to the first group (A), which encompasses the watershed of lakes in southern and central Finland. The Temmesjoki, Siikajoki, and Pykäjoki rivers are part of the second group (B), which comprises rivers of large sizes in Northern Finland. The third group (C), in contrast, includes medium- and small-sized river basins like Torniojoki, Simojoki, Kuivajoki, Kiiminkijoki, Kemijoki, and Iijoki rivers.
Data description
For all the natural and regulated rivers selected for this study, long-term (1968–2023) daily discharge time series, without any missing values, were collected from the Finnish Environment Institute (SYKE) (https://www.syke.fi/en-US/Services, accessed on 10 June 2024). Based on previous studies focusing on links between large-scale atmospheric-oceanic circulation patterns and climate variability across Finland (Irannezhad et al. 2014, 2015a), six climate teleconnections (Table 1) were selected. For these climate teleconnections, the monthly time series from January 1968 to December 2023 were obtained from the website of the Climate Prediction Center (CPC) at the National Oceanic and Atmospheric Administration (NOAA), USA, freely and openly available at https://www.cpc.ncep.noaa.gov/data/teledoc/telecontents.shtml (accessed on 10 June 2024). For this study, the calendar-based year (January–December) and climatological seasons (spring: March–May, summer: June–August, autumn: September–November, and winter: December to the following February) were considered as annual and seasonal scales, respectively. For example, the annual (winter season) time series of river discharge and climate teleconnections for 1969 referred to the period from Jan 1969 (Dec 1968) to Dec 1969 (Feb 1969).
No. . | ID . | Climate teleconnection . | Source . | Reference . |
---|---|---|---|---|
1 | AO | Arctic Oscillation | CPC | Thompson & Wallace (1998) |
2 | EA | East Atlantic | CPC | Barnston & Livezey (1987) |
3 | EA/WR | East Atlantic/West Russia | CPC | Barnston & Livezey (1987); Lim & Kim (2013) |
4 | NAO | North Atlantic Oscillation | CPC | Barnston & Livezey (1987) |
5 | POL | Polar/Eurasia pattern | CPC | Barnston & Livezey (1987) |
6 | SCA | Scandinavia pattern | CPC | Barnston & Livezey (1987); Bueh & Nakamura (2007) |
No. . | ID . | Climate teleconnection . | Source . | Reference . |
---|---|---|---|---|
1 | AO | Arctic Oscillation | CPC | Thompson & Wallace (1998) |
2 | EA | East Atlantic | CPC | Barnston & Livezey (1987) |
3 | EA/WR | East Atlantic/West Russia | CPC | Barnston & Livezey (1987); Lim & Kim (2013) |
4 | NAO | North Atlantic Oscillation | CPC | Barnston & Livezey (1987) |
5 | POL | Polar/Eurasia pattern | CPC | Barnston & Livezey (1987) |
6 | SCA | Scandinavia pattern | CPC | Barnston & Livezey (1987); Bueh & Nakamura (2007) |
Statistical analyses
The Mann–Kendall (MK) non-parametric test (Mann 1945) was used to detect statistically significant (p < 0.05, i.e. 5% significance level or 95% confidence interval) trends in AMDD and SMDD time series in selected natural and regulated rivers in Northern Finland. To quantify the magnitude of these significant trends, the Sen's Slope method (Sen 1968) was applied. This study employed the Spearman's rank correlation (Rho) to measure the connections of AMDD and SMDD time series with annual and seasonal climate teleconnections, respectively (Helsel & Hirsch 1992). The Rho, unlike the Pearson's correlation coefficient (r), allocates no specific distribution functions for AMDDF and SMDDF (Helsel & Hirsch 1992). If there is a positive autocorrelation in such time series, however, the trend-free pre-whitening (TFPW) method (Yue et al. 2002) was used to determine significant trends, while the residual bootstrap (RB) approach (Park & Lee 2001) with 5,000 independent replications was applied for examining the standard deviation of the Rho.
RESULTS
Annual mean daily discharge
Seasonal mean daily discharge
During 1969–2023, all statistically significant (p < 0.05) trends found in SMDD in both natural and regulated rivers in Northern Finland for spring, summer, autumn, and winter were positive or increasing (Figure 5). For spring, such increases were at the rates of 3.55, 1.62, and 0.30 (m3 s−1 year−1) for the Tornionjoki, Oulujoki, and Pyhäjoki rivers, respectively (Figure 5(a)). The first river (Tornionjoki) also experienced increasing trends in SMDD time series for autumn (2.56 m3 s−1 year−1) and winter (1.01 m3 s−1 year−1) (Figure 5(c) and 5(d)). For these two seasons (autumn and winter), substantial increases (ranging between 0.21 and 2.56 m3 s−1 year−1) were detected in SMDD in Simojoki, Kuivajoki, Kiiminkijoki, Kemijoki, and Iijoki (Figure 5(c) and 5(d)). Three of these rivers (Kemijoki, Kuivajoki, and Iijoki), moreover, showed statistically significant trends in SMDD time series during the summer season (Figure 5(b)). For both Temmesjoki and Siikajoki, however, no clear changes were found in SMDD during any of the four seasons over time (Figure 5).
DISCUSSION
River discharge regime in Northern Finland
In general, both climatic and human impacts play critical roles in significant changes in historical river discharge regimes around the world. Human activities such as land cover-land use changes, bog drainage, and regulations have already influenced discharge in some river basins (e.g., Irannezhad et al. 2018). Changes in climatic conditions have also altered different characteristics (in terms of magnitude and timing) of river discharge in different parts of our planet (e.g., Minaei & Irannezhad 2018). At high latitudes, accordingly, significant alterations in river discharge are one of the key hydrological responses to global warming, climate change, and human activities (Korhonen & Kuusisto 2010; Blöschl et al. 2019).
The present study found statistically significant (all increasing) trends in historical (1969–2023) AMDD in only three rivers of Tornionjoki, Simojoki, and Kuivajoki. All these natural rivers are located in Region C of the Finnish Rivers System, with abundant annual flow because there are not as many lakes as the water reservoirs (Korhonen & Kuusisto 2010). Such increases in AMDD in these rivers can be explained by more annual precipitation (Irannezhad et al. 2014), warmer annual mean SAT (Irannezhad et al. 2015a), and consequently a higher amount of rainfall across their basins. Besides such climatic factors, different human activities, like extensive drainage operations mainly in peatland forests, changed land cover in the basin of these rivers and thereby modified their AMDD over time (Meriö et al. 2019). Similar to our findings, on the other hand, previous studies also determined no significant trends in mean annual flow in regulated rivers in Finland during 1912–2004 (Korhonen & Kuusisto 2010).
For winter, this study found statistically significant increasing trends in SMDD in both natural and regulated rivers in Northern Finland during 1963–2010. Korhonen & Kuusisto (2010) also reported increases in wintertime river discharge in Northern Finland in recent decades. Similar to our findings, they concluded that the main cause of such increases, particularly for natural rivers, was more rainfall due to the milder and wetter climatic conditions during winter in recent decades. For the regulated rivers in Northern Finland, however, the other reason behind such SMDD increases in winter was the higher rates of water release from the hydropower dams during winter match the cold season energy needs and create enough storage capacity for snowmelt approaching the end of winter or the early spring (Korhonen & Kuusisto 2010).
For spring, SMDD increased in only one natural river (Tornionjoki) and two regulated rivers (Oulujoki and Pyhäjoki) across Northern Finland during 1969–2023. Korhonen & Kuusisto (2010) also found a significant increasing trend in historical springtime river discharge in both Tornionjoki and Oulujoki. This could be related to warmer springtime SAT resulting in a rapid snowpack ablation accelerated by more frequent rain on snow events (Uvo et al. 2021). Hence, toward the end of the spring season, snowmelt declined and rainfall runoff became the dominant contributor to river discharge in such snow-dominant catchments (Uvo et al. 2021). However, increases in springtime SMDD in regulated rivers can also be explained by releasing water for securing its supply and power generation at the end of spring and in the early summer, when droughts are common in Northern Finland due to less and early springtime snowmelt runoff (Okkonen & Kløve 2010).
Similar to Korhonen & Kuusisto (2010), the present study found considerable increases in summertime river discharge in two natural rivers (Simojoki and Kuivajoki) in Northern Finland. Such increases might be related to more summer precipitation across this part of the country (Irannezhad et al. 2014) and higher rates of drainage from both forest and peatland areas (Korhonen & Kuusisto 2010). For autumn, the present study found only significant increases in SMDD in both natural (Tornionjoki, Simojoki, Kuivajoki, and Kiiminkijoki) and regulated rivers (Kemijoki and Iijoki) during 1969–2023. In the era of global warming, higher SAT during autumn can result in slower and later soil frost processes before winter in Northern Finland, thereby accelerating infiltration, enhancing deeper flow paths, mobilizing subsurface water, and consequently increasing river discharge (Evans et al. 2020). More rainfall on snow events during the late autumn might also increase river discharge in northern Finnish rivers (Rawlins et al. 2009; Gohari et al. 2022).
The role of climate teleconnections
In general, the SCA was the most significant climate teleconnection controlling variability and trends in AMDD in Northern Finland's rivers during 1969–2023. Similarly, Uvo et al. (2021) concluded that discharge regimes in river basins at high latitudes in Northern Europe were negatively associated with variations in the SCA during 1960–2010. According to previous studies, the SCA was also one of the climate teleconnections negatively influencing precipitation (mainly rainfall) (Irannezhad et al. 2017) variability across Northern Finland (Irannezhad et al. 2014). Similar to Korhonen (2006), this confirmed that AMDD in rivers across Northern Finland is mainly influenced by precipitation variability rather than changes in snowmelt under global warming. The primary circulation center of this climate teleconnection (SCA) is located across the Scandinavian Peninsula and the Siberian segment of the Arctic Ocean (Barnston & Livezey 1987). The other two weaker centers of opposite anomalies sign are located across Western Europe and eastern Russia/western Mongolia (Bueh & Nakamura 2007). The SCA negative (positive) phase expresses low (high) pressure anomalies across the Norwegian Seam Greenland, and Scandinavian region, bringing wetter (drier) airflow toward Finland (Bueh & Nakamura 2007). However, the positive phase of SCA may sometimes describe the main blocking anticyclones across Scandinavia and Western Russia. Such a natural sign of SCA was clearly reflected by the significant negative relationships found by this study between AMDD in Northern Finland's rivers and the SCA. In fact, the positive (negative) values of SCA were associated with lower (high) AMDD than normal in Northern Finland's rivers, particularly during the first (last) 12 years of our study period from 1969 to 1980 (2012–2023).
For winter, spring, and summer, the AO and NAO were generally the most significant teleconnection positively influencing SMDD in rivers across Northern Finland during 1969–2023. Similarly, Uvo et al. (2021) concluded that streamflows in snow-dominated catchments at high latitudes in northern Europe were positively associated with the AO during 1960–2010. This climate teleconnection also influenced peak spring flood discharge in rivers across Northern Finland during 1967–2011 (Irannezhad et al. 2022b). The AO indicates the power of the circumpolar vortex, while the NAO describes the intensity of west circulation from the North Atlantic to the Atlantic European section (Thompson & Wallace 1998). In general, hence, the NAO is considered a main component of the AO (Serreze et al. 2000). Their positive and negative phases naturally correspond to strong and weak westerly circulations, transporting mild maritime and cold airflows over Northern Europe, respectively, during the winter season (Gormsen et al. 2005; Jaagus 2006). In recent decades, the AO (NAO) power increased by 0.26 (0.20) decade−1 (Wang et al. 2005), resulting in milder and wetter winters throughout Northern Finland (Irannezhad et al. 2014, 2015a). This could significantly warm SAT, thereby increasing rainfall (Irannezhad et al. 2016), and consequently raising the rate of discharge in rivers across Northern Finland during the winter, spring, and summer seasons in 1969–2023.
In Northern Finnish rivers, SMDD time series for autumn were negatively associated with the EA/WR anomalies. Previous studies also reported statistically significant negative correlations between the EA/WR and SMDD in rivers at high latitudes in Northern Europe during the spring, summer, and autumn seasons (Uvo et al. 2021; Irannezhad et al. 2022b). The EA/WR describes the meridional circulation for Finland that naturally weakens with the strengthening of the westerly airflows. It is a zonally oriented climate teleconnection consisting of (i) two anomaly centers located across the Caspian Sea and Western Europe in winter, and (ii) three anomaly centers across the coast of Portugal, north-west Europe, and west–north–west Russia during spring–autumn. In general, the negative (positive) EA/WR phase is associated with the anomalous southerly and southeasterly (northerly and northwesterly) airflows, bringing wetter (drier) and milder (colder) climatic conditions across Northern Europe, including Finland (Lim & Kim 2013; Irannezhad et al. 2014, 2015a). In rivers across Northern Finland, hence, significant increases in SMDD time series for autumn can be explained by their negative relationships with the EA/WR that similarly generate precipitation runoff rates in Northern Europe.
Interestingly, we did not notice major differences in the effects of climate teleconnections on annual and seasonal mean discharge between regulated and unregulated river systems. The regulated river systems selected by this study are runoff-river types (Ashraf et al. 2016) with limited storage capacity in the mainstream, while the main storages in the upper catchments are natural or artificial (only for the Kemijoki River) lakes that play an important role in the regulation processes (Ashraf et al. 2016). Hence, the main storage capacity in these regulated river systems is lower than in many other hydropower rivers worldwide (Marttila et al. 2024). Such regulations influence river discharge primarily within day and week scales (Ashraf et al. 2018; Virk et al. 2024), shift peak spring flood discharge timing (Irannezhad et al. 2022b), and increase wintertime base flow due to higher energy need during this season (Ashraf et al. 2018; Virk et al. 2024). However, these effects were not seen that strongly in our analyses focusing on long-term variability and trends in annual and seasonal mean daily discharge time series and their connections to teleconnections.
CONCLUSIONS
The present study analyzed long-term (1969–2023) historical variations and trends in river discharge throughout Northern Finland and their relationships to large-scale climate teleconnections. The major findings were as follows:
In both natural and regulated rivers in Northern Finland, the lowest and highest AMDDF were generally recorded in the first (1969–1980) and last (2012–2023) 12 years of our study period, respectively. Statistically significant trends in AMDD were found only for three natural rivers of Tornionjoki, Simojoki, and Kuivajoki during 1969–2023. Such historical variations and changes in AMDD time series were most significantly controlled by the SCA, which is also one of the influential climate teleconnections for interannual precipitation variability across Northern Finland. Although snowpack hydrological processes have already been affected by global warming, our findings might prove that precipitation variability more than snowmelt controls AMDD in rivers throughout Northern Finland under climate change.
The highest and lowest SMDD in rivers across Northern Finland were generally experienced in spring and winter, respectively. For these rivers, only increasing trends in SMDD were statistically significant (p < 0.05). These substantial increases during winter, spring, and summer were generally associated with the stronger AO and NAO positive phases in recent decades, bringing more precipitation, milder SAT, and consequently more rainfall across Northern Finland. For autumn, SMDD increased in those rivers with statistically significant correlations with the EA/WR, the climate teleconnection negatively controlling both precipitation and SAT over Finland. Generally speaking, the recent stronger negative phases of this climate teleconnection (EA/WR) modified different hydrological processes (mainly infiltration) in Northern Finland by not only increasing precipitation in autumn but also postponing the winter season. In conclusion, hence, SMDD in rivers in Northern Finland are generally controlled by the climate teleconnections primarily affecting seasonal precipitation across this country: the AO and NAO for winter, spring, and summer, while the EA/WR for autumn.
ACKNOWLEDGEMENTS
The authors are grateful to the Finnish Environment Institute (SYKE) for monitoring and recording historical daily river discharge in Finland. They would also acknowledge the Climate Prediction Center (CPC) at the National Oceanic and Atmospheric Administration (NOAA) of the United States for making available online the standardized monthly values of climate teleconnections used in this study.
FUNDING
This study was funded by the Sakari Alhopuro Foundation (Grants 20220247 and 20230218), and Maa- ja vesitekniikan tuky r.y. (Grants 44008 and 45599).
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.