Hydrological drought assessment of the Sava River basin in South-Eastern Europe

Carbon emissions from production and consumption systems are the reasons for the extreme weather events such as drought and intensive rainfall (Abbas et al. 2021a, 2021b, 2022; Ehsan et al. 2024; Jiang et al. 2024). Hydrological drought is characterized by an insufficient presence of water in the hydrological system, leading to unusually low streamflow in rivers and diminished levels in lakes, reservoirs, and groundwater (Van Loon 2015). This precipitation deficit may accumulate swiftly or take months before manifesting as lowered lake levels, lowered river discharges, or deeper levels of groundwater. The far-reaching consequences of drought extend beyond its immediate hydrological impact, affecting human activities, lives, and Earth's ecosystems (Sardou & Bahremand 2014). This phenomenon has engaged interdisciplinary attention from scientists in geography, ecology, hydrology, meteorology, and agriculture, each contributing unique perspectives (Tigkas et al. 2015). Of several forms of droughts, the hydrological component is particularly critical due to its profound implications for hydropower production, water supply in urban areas, and industrial activity which significantly depend on surface water supplies (Vasiliades et al. 2011). Initiated by rainfall deficits, hydrological droughts are commonly associated with reservoirs/lake water levels within a basin, contributing to widespread impacts such as water supply reduction, deteriorating water quality, irrigation restrictions, agricultural failure, diminished energy production, disturbance of riparian habitats, limited outdoor activities, and disruptions to diverse social and economic endeavours (Mishra & Singh 2010). Despite European countries being generally considered to possess sufficient water resources, the increasing occurrence and spread of water scarcity and droughts have become evident. The period from 1980 to 2022 witnessed total economic losses due to weather- and climate-related events in 32 countries of the European Economic Area estimated at 650 billion euros, of which 52.3 billion euros in 2022 and 59.4 billion euros in 2021 (European Environment Agency 2023). The accumulating impacts of climate change, global warming, and human activities pose unprecedented challenges to exacerbate the drought problem (Asadieh & Krakauer 2015; IPCC 2021, 2021).

The driving force behind droughts lies in the decrease in precipitation, resulting in diminished storage volumes and fluxes within the hydrological cycle (Mosley 2015). Essential for comprehending and monitoring drought events, drought indicators rely heavily on physical datasets encompassing river discharge, groundwater, precipitation, reservoir storage, and soil moisture. These indicators are categorized as hydrological, meteorological, and agricultural (Wable et al. 2019).

Streamflow is one of several hydrological variables used for drought characterization that holds paramount significance in determining water quantity. Hydrological droughts, identified by stream flow deficits relative to normal conditions, are characterized by severity, onset time, length, areal frequency, and areal coverage (Sardou & Bahremand 2014). The streamflow drought index (SDI) employs cumulative streamflow volumes to predict the onset, duration, and severity of droughts (Nalbantis & Tsakiris 2009). Recognized for fulfilling essential criteria for drought indices, the SDI has been globally applied, from Greece (Tigkas et al. 2012) and the Yangtze River in China (Li et al. 2012), Iran (Tabari et al. 2013) to the shared Diyala River Basin in Iraq and Iran (Al-Faraj et al. 2014). SDI has been instrumental in various studies across Europe, encompassing the Neman and the Vistula Basins in Eastern Europe (Rimkus et al. 2013; Kubiak-Wójcicka & Bąk 2018), in Croatia, this index was applied on the Cetina River basin (Ljubenkov & Kalin 2016), in UK it was applied on the 121 watersheds (Barker et al. 2016), and in Central Europe, on the Tisza (Leščešen et al. 2020), and in Lithuania it was applied on 15 catchments (Nazarenko et al. 2023).

The main goal of this research was to investigate the occurrence, frequency, and magnitude of hydrological droughts in the Sava River basin (SRB), the biggest basin in South-Eastern Europe, by employing the SDI. Discharge values from 12 monitoring stations along the Sava River and its affluents, located in Slovenia, Croatia, Bosnia & Herzegovina, and Serbia (Table 1), form the basis of our analysis.

SRB is under the influence of different river regimes, upriver parts, from the Litija station to the Četeš station, Alpine nival-pluvial regime is represented (Frantar et al. 2008). Downriver from the Čatež station, the regime shifts to an Alpine pluvial-nival regime (Ulaga et al. 2008). The Peripannonian pluvial-nival type is represented in the middle part of the SRB, from Zagreb to Jasenovac stations, and it includes Novi Grad, Prijedor, Delibašino selo, Doboj, Bajina bašta, and Slovac stations. Finally, Županja, and Sremska Mitrovica stations are under the Pannonian pluvial-nival type influence (Čanjevac & Orešić 2018; Leščešen et al. 2022a).

SEE, including the SRB, is a climate change hotspot characterized by above-average warming and highly vulnerable populations (Leščešen et al. 2023). Previous research that most commonly used the hydrological and meteorological data from the second half of the 20th century and first decades of the 21st century indicate that SEE has already experienced a significant increase in summer temperatures (Trbić et al. 2017), a decrease in summer precipitation (Milošević et al. 2021), and an increase in consecutive dry days (Popov et al. 2019), all of which contribute to increased overall dryness. The most important climate simulations predict a temperature increase of 3.5°C for the Western Balkans with moderate greenhouse gas emissions up to 8.8°C in the high emissions scenario (RCP 8.5) by the end of the century (IPCC 2021). There is very high confidence – indicating robust empirical evidence and high agreement between the reviewed studies – that summer temperatures in this region will increase more than the global terrestrial average, leading to more intense evapotranspiration that will further contribute to intensification of hydrological droughts during the warm period of the year (IPCC 2021). All of these results indicate that in the future a lack of water can be expected in the region, especially during the warm period of the year, when water is needed the most for agriculture, energy production, industries, and maintaining ecological balance.

In this paper, a 60-year (1961–2020) database of mean monthly streamflows for 12 selected stations that are located within the SRB (Figure 1) was utilized. This database can be considered satisfactorily long as usually a database of at least 50 years is necessary to differentiate variability from trends (Kundzewicz & Robson 2000). Streamflow data were obtained from four different national agencies (Slovenian Environment Agency; Meteorological and Hydrological Service, Croatia; Republic Hydrometeorological Service of Serbia; and the Republic Hydrometeorological Service – Republic of Srpska) (Table 1). The dataset was separated into hydrological summer (April–October) and hydrological winter (November–March).

Data quality control measures were rigorously implemented by each national organization responsible for the collection and maintenance of hydrological data. All organizations adhered to strict standards to ensure the accuracy and consistency of the data series used in this study. In doing so, they followed their internal instructions, but also the recommendations of the World Meteorological Organization (WMO). To ensure the homogeneity of the data, the Mann–Kendall (MK) and Pettitt tests, a non-parametric method often used to detect change points in time series data (Ferraz et al. 2022), was applied. The Pettitt test is particularly useful for detecting shifts in median values and thus assessing the consistency of hydrological data over time (Kocsis et al. 2020). The results of the MK and Pettitt tests can be found in Table 2.

Table 2 summarizes the results of the MK and Pettitt tests carried out with the data from the selected stations. It is noteworthy that all locations except Novi grad, Prijedor, and Slovac show a significant decreasing trend in streamflow, as evidenced by low p-values in the MK test. The results of the Pettitt test indicate that there are no significant changes at the 0.05 significance level at all sites, as the p-values are above 0.05, which in combination with the MK test results indicates homogeneity of the data for all sites. The data presented in Table 2 show a consistent downward trend in flow for most locations, with the Pettitt test revealing no significant changes. This consistency, combined with the classification of the data as homogeneous, indicates that the dataset is reliable for further analysis.

The natural logarithms of cumulative streamflow are calculated using the mean and standard deviation s_y, with these statistical parameters estimated over an extensive period. The classification of drought states mirrors those used in meteorological drought indices, such as the standardized precipitation index. This classification encompasses five states, as outlined in Table 3 (Tigkas et al. 2015).

For SDI trend analysis the MK test, a non-parametric method relying on ranks, is applied. This test is commonly used for detecting trends in climatological and hydrological time series (Bezak et al. 2016; Leščešen 2022b). In the MK test, the null hypothesis (H₀) posits the absence of any trend in the series, while the alternative hypothesis (H_A) asserts the presence of a trend, which could be either positive or negative. In this study, a significance level of 5% is applied, indicating that statistical significance is acknowledged if the calculated p-value is less than or equal to 0.05. This approach helps to determine whether the observed trends in SDI data are statistically meaningful or simply a result of random variability.

The results presented in Table 3 provide valuable insights into the hydrological dynamics of each station. Several stations along the Sava River, including Litija and Čatež, displayed significant negative trends in both annual (p = 0.035; p = 0.018, respectively) and summer streamflow (p = 0.001 at both stations), indicating a decline in water availability during the summer period. This consistent pattern was further emphasized by Jasenovac, corroborating the existence of a noteworthy negative trend in the SRB (annual p = 0.023; summer p = 0.001). Interestingly, Zagreb did not exhibit a significant annual trend, while a considerable negative trend in summer (p = 0.007) suggests potential seasonal variations impacting streamflow. Conversely, stations such as Županja and Sremska Mitrovica also did not show significant annual trends, yet both displayed a negative trend during summer (p = 0,013; p = 0.010, respectively). These findings further underscore the importance of considering seasonality in hydrological assessments. The most extreme drought on the annual level was recorded during 2011–2012 with an average SDI value of −2.48. Overall, from eight drought events with SDI values lower than −1, five were recorded after the year 2000.

As results in Table 2 indicate that only the summer season has shown statistically significant changes in SDI over the whole SRB, plots representing these changes were created and presented in Figures 3 and 4.

The SDI values fluctuate during the summer season over the observed period, reflecting the hydrological variability within the SRB. Notably, the findings reveal discernible patterns of hydrological drought intensity along the entire stretch of the river over this period, spanning from comparatively humid conditions at the outset to drier conditions in recent decades.

The hydrological trends observed in the right tributaries of the Sava River, including the Una, Sana, Vrbas, Bosna, Drina, and Kolubara Rivers, provide valuable insights into the changing hydrological conditions in this region.

The hydrological conditions across various tributaries of the Sava River, as exemplified by the Una River–Novi Grad station, the Sana River–Prijedor station, the Vrbas River–Delibašino Selo station, the Bosna River–Doboj station, the Drina River–Bajina Bašta station, and the Kolubara River–Slovac station, reveal intricate patterns of drought occurrences from 1961 to 2020. Examining the Una River, the data reflect a mix of drought states over the years. Notable periods of extreme drought, observed in 2002–2003 (SDI = −2.55) and 2010–2011 (SDI = −2.39), underscore the vulnerability of this station to severe hydrological stress. The Sana River exhibits a similar trend, with peaks of extreme drought aligning with regional drought events in 1999–2000 and 2002–2003. In contrast, the Vrbas River–Delibašino Selo station experiences more variability, with intermittent extreme drought events, such as those in 2002–2003 (SDI = −2.32) and 2019–2020 (SDI = −1.69). The Bosna River displays periods of extreme drought, particularly in 2002–2003 (SDI = −2.52) and 1999–2000 (SDI = −2.19), highlighting the shared vulnerabilities of these interconnected river systems. The 2003 drought is considered as the most widespread hydrological drought in Europe (Hanel et al. 2018). Analysing the Drina River, a pronounced drought period was observed from 2006 to 2012 with the lowest SDI value in 2007–2008 at −1.71. Finally, the Kolubara River station showcases a distinctive pattern, with a prominent spike in extreme drought during 1999–2000 (SDI = −1.86), highlighting the localized nature of hydrological dynamics. The rise in the number of consecutive dry days in recent decades, with anticipated further increases in the future (Popov et al. 2018), highlights the potential for extended periods of meteorological droughts to transform into impactful hydrological droughts, affecting water supply significantly (Van Lanen et al. 2016). Furthermore, the amount of summer precipitation over the region has shown a statistically significant decrease in the recent period (Gajić-Čapka et al. 2015; Leščešen et al. 2022b). Precipitation analysis over the Sava River watershed in B&H revealed mainly insignificant trends in both signs, whereas a prominent decrease (especially in the north of B&H) was found in the summer season (up to 14 mm/decade). On the other hand, air temperature displayed prominent upward trends across the entire study area. The most pronounced warming (0.4–0.6 °C/decade) was present during the summer which further supports hydrological drought intensification in this part of the SRB (Gnjato et al. 2021). In the same study, authors emphasize a significant negative correlation between air temperature and river discharges and the strongest positive correlation between precipitation and river discharges, underpinning the worsening hydroclimatic conditions during the warmest part of the year.

The study has several potential limitations that should be considered when interpreting the results. One key limitation is the representativeness of the monitoring stations used, which may not adequately capture the spatial variability of hydrological conditions across the entire study area. This could lead to biases in the data and, consequently, the findings. Additionally, uncertainties in the data collection and processing methods, such as measurement errors, inconsistencies in data recording, and gaps in the time series, could further affect the reliability of the results. Another significant limitation is the methods’ inability to account for intricate processes such as groundwater–surface water interactions, land use changes, and feedback mechanisms within the hydrological cycle.

Given these limitations, future research should focus on investigating the spatial analysis of droughts within SRB as well as the impact of anticipated climate variability on future drought scenarios and evaluating the effectiveness of various drought mitigation strategies to enhance our preparedness for and management of hydrological extremes. Furthermore, it is recommended to use hydrometeorological drought indices across various regions in the world to create more comprehensive systems for drought monitoring and prediction (Abbas et al. 2021a, 2021b), correlating hydrological with meteorological drought indexes (Sarwar et al. 2022). Additionally, incorporating anthropogenic information, such as changes in crop varieties, irrigation practices, and other adaptive measures, is essential for conceptualizing the full scope of drought impacts and designing effective mitigation plans (Waseem et al. 2022; Abbas et al. 2023). By addressing these research areas, we can develop more robust strategies to mitigate the effects of drought and improve water resource management in the face of climate change.

In conclusion, this study delves into the intricate dynamics of hydrological droughts in the expansive SRB, revealing significant trends and patterns that underscore the region's vulnerability to changing hydrological conditions. The observed decrease in streamflow, particularly during the critical summer season, has been a prevalent trend across numerous monitoring stations, indicating a heightened risk of hydrological droughts in the basin. The MK test results highlight that a substantial majority of stations, 83.3% during the hydrological summer season, recorded statistically significant decreases in streamflow, emphasizing the severity and widespread nature of this issue.

The findings elucidate the interconnectedness of hydrological systems within the SRB, emphasizing the cumulative downstream intensification of hydrological drought severity. The analysed tributaries, including the Una, Sana, Vrbas, Bosna, Drina, and Kolubara Rivers, exhibit diverse responses to changing hydrological conditions, reflecting the complex and localized nature of drought occurrences. Notably, the increased frequency and severity of hydrological droughts in recent decades, particularly evident in the last two, emphasize the escalating risk and sustained impact of prolonged drought events in the SRB.

Projected climate change, coupled with the compounding effects of global warming and human activities, poses unprecedented challenges that exacerbate the hydrological drought problem. The anticipated shifts in precipitation patterns and temperature, as indicated by climate change projections, are likely to further intensify the occurrence and severity of hydrological droughts in the SRB. This necessitates urgent attention to adaptive water resource management strategies to mitigate the potential socioeconomic and environmental consequences.

The implications of hydrological droughts extend far beyond reduced water availability, affecting agriculture, industries, public water supply, ecosystems, energy production, food security, and overall social well-being. As water scarcity becomes more prevalent, the livelihoods of the people living within the SRB are at stake. Agriculture, a cornerstone of the regional economy, faces challenges such as reduced crop yields and livestock productivity. Industries dependent on water resources are at risk, and public water supply may be compromised. Ecosystems, energy production, and food security are also threatened, highlighting the need for comprehensive and sustainable water management policies.

Given these findings, it is crucial for the inhabitants of the SRB to adopt proactive measures such as developing robust water management strategies, sustainable agricultural practices, and climate-resilient infrastructure. The implementation of advanced real-time monitoring systems and the enforcement of water use restrictions during critical drought periods can greatly improve water conservation efforts. Furthermore, encouraging the use of water-saving technologies and upgrading existing infrastructure will help to mitigate the effects of increasing hydrological droughts. Additionally, fostering international collaboration among the countries sharing the basin is crucial to addressing the transboundary nature of water resources. This study serves as a clarion call for timely and concerted efforts to enhance the region's resilience to the escalating challenges of hydrological droughts in the face of a changing climate.

Future studies in the SRB should delve into climate change impacts, integrated water resource management, and the ecological and socioeconomic repercussions of hydrological droughts. Investigating land use changes, water demand dynamics, and applying advanced monitoring techniques will be essential for informing adaptive measures and sustaining the region's water resources.

This research was supported by ClearClimate project which has received funding from the European Union's Horizon Europe research and innovation programme under grant agreement No 101131220.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.