Most European countries are considered to have sufficient water resources; however, water scarcity and droughts are increasing and spreading. From 1980–2020, the total economic loss from weather- and climate-related events was EUR 450–520 billion (in 32 countries of the European Economic Area [1
]). Climate change, global warming and human activity may unprecedentedly exacerbate the problem of drought [2
Even among drought experts, there is no single definition of drought that everyone would agree on [5
]. Water Directors within the CIS (a Common Strategy for the implementation of the Water Framework Directive) process have decided on the following definition of drought: it is a temporary, negative and severe deviation along a significant period and over a large region from average precipitation values (a rainfall deficit), which may lead to meteorological, agricultural, hydrological and socio-economic drought, depending on its severity and duration [8
]. Water deficit typically propagates through the hydrological cycle, impacting different ecosystems and human activities accordingly [9
Accomplished studies challenge the view that hydrological drought can only be described as a lack of precipitation and show many gaps and uncertainties in our knowledge of this extreme event. A number of interrelated phenomena may cause hydrological drought [10
]; many efforts are being made to study the various aspects of droughts and aim to provide early warning and information to decision-makers, policy-makers, water managers, water users and the general public about droughts. To prevent or at least mitigate the effects of a drought, it is necessary to understand this phenomenon, identify its signs as quickly as possible and prepare for a drought’s impact [12
Scientists have developed numerous methods to identify hydrological drought. Criteria for identifying an impending hydrological drought and its beginning or end can include the simplest indicators (e.g., river or groundwater level, flow rate) or complex drought indices (e.g., aggregate dryness index, palmer hydrological drought severity index, surface water supply index) that require several or more indicators. When managing drought, it is convenient to use indices to reduce the complex problem to one single number. However, water managers should be cautious in choosing indices [6
]. It would be beneficial to develop a composite drought index that integrates all relevant data and drought descriptions, considering the predominant types of droughts in time and space and climate change scenarios [16
]. However, a recent report published by the Intergovernmental Panel on Climate Change [4
] warns that droughts are a complex and difficult-to-predict natural phenomenon, and that differences between drought types are not unambiguous and cannot be described by a single universal definition or directly measured by a single variable. The National Meteorological and Hydrological Services around the world are encouraged to use the standardized precipitation index (SPI) to characterize meteorological drought; however, a comprehensive indicator to describe agricultural and hydrological droughts still needs to be proposed [17
Although Lithuania belongs to a humid continental climate, the drought phenomenon is quite well known. The recent dry and warm summers are causing major changes in river runoff. For three consecutive summers of 2018–2020, a hydrological drought for the entire country was declared. Although scientific studies based on available observational data do not reveal clear trends in rising dryness and extreme droughts [18
], end-of-century climate change may enhance the likelihood of more intense and frequent meteorological droughts, which may increase the threat of hydrological droughts [20
]. Rising warm-season temperatures and consequent increasing evaporation are likely to have a particular impact on runoff during the warm season, which is a critical time for water users and aquatic ecosystems, even under normal climatic conditions. With the growing evidence of climate change, Lithuanian scientists have been paying closer attention to droughts in recent years. The meteorological effective drought index (EDI) proposed by Byun and Wilhite [22
] was used to identify hydrological drought during the warm period [23
]. A study of drought dynamics during the warm period of the year using the meteorological standardized precipitation index (SPI) proposed by McKee et al. [24
] and the hydrological standardized water level index offered by Nalbantis and Tsakiris [25
] was carried out [19
]. The suitability of the hydrological standardized runoff index (SRI) proposed by Shukla and Wood [26
] to determine hydrological drought was also investigated [27
]. The ability of the analyzed indices to identify hydrological droughts in Lithuanian rivers mainly depended on the nature of river feeding, e.g., some indices performed better on groundwater-fed rivers and others on snowmelt-fed rivers. To our knowledge, thus far, only one scientific study has been devoted to assessing hydrological drought in Lithuania using river water levels [28
]. This study aimed to identify the warm-period hydrological drought cases in Lithuania using the streamflow drought index (SDI; calculated based on discharge data) and standardized water level index (SWLI; calculated based on water level data) to compare and evaluate the possibilities of their practical application. The findings based on data from seven rivers (eight water gauging stations) revealed that a modified SDI methodology based on water level data (i.e., SWLI) could become a good alternative for detecting hydrological droughts in Lithuania.
As the operational information of the hydrology network in Lithuania consists of (hourly) water level data, it should be used to characterize the hydrological drought and declare the state of severe hydrological drought. Such an assessment has a significant advantage. Water levels can be easily measured directly, while discharge is estimated indirectly from the water level using a water level-discharge ratio. This study aimed to analyze the past conditions of hydrological drought and project future drought scenarios for the entire territory of Lithuania based on its major sub-basins using the improved methodology for calculating the standardized water level index (SWLI).
4. Discussion and Conclusions
The present study was designed to determine the suitability of the standardized water level index (SWLI) to monitor, follow and forecast hydrological drought conditions in Lithuania.
The hydrological drought index SDI, in turn, was developed [25
] based on the concept of the widely known and recognized standardized precipitation index (SPI) [17
]. A number of other hydrological indices are calculated similarly to SPI (e.g., standardized reservoir supply index (SRSI), standardized streamflow index (SSFI), standardized water-level index (SWI)) [15
The developed methodology was adapted to analyze hydrological droughts in Lithuanian river catchments over the past three decades. The hydrological identification and quantification of droughts using the modified SWLI have led to the discovery of past and future trends.
The results indicated the most severe droughts over the 30 years in 1992, 2002, 2006, 2019 and 2020. Droughts in Lithuania in 1992, 2002 and 2006 were identified using other methodologies [19
]. The vegetation seasons of 1992 and 2002 were also described as extremely dry in the eastern Baltic Sea region [38
]. The highest rates of flow intermittence in 1992, 2002, 2006 and 2018 were established by Šarauskienė et al. [30
]. According to agricultural drought criteria, from 1992 to 2006, as well as in 2018 and 2019, large areas of Lithuania suffered from extreme dryness [39
]. According to [40
], over the studied period of 1950–2012, the longest and most severe and widespread drought event in the Baltic States was recorded during 2005–2009; in Eastern Europe, the most prolonged was in 1992–1995 and the most severe in 1989–1991. The year 2019 was the warmest year on record in Poland [41
]. Blauhut et al. [42
] listed the Lithuanian neighbors—Belarus, Latvia and Poland—as particularly affected by the multi-year drought of 2018–2019. Furthermore, the 2018–2020 drought event of extraordinary intensity covered a significant part of Europe [43
]. Moreover, it was followed by a drought in 2022 that was considered the worst in at least 500 years [44
In general, using the developed methodology, a positive trend in the number of severely dry days was detected over the last three decades. A similar pattern of results was obtained in the neighboring northern part of Poland: based on different indices, river flow decrease was identified for the period of 1981–2016 [45
]. These basic findings are consistent with research [46
] showing the ongoing negative water balance of the Greater Poland region in the years following 1988. In Latvia, since the early 1990s, remarkably drier conditions have been observed more often as well [47
]. Our findings are consistent with what was found in the study [48
], which analyzed long-term changes in drought indices of central and eastern European countries during 1949–2018. These authors estimated drying trends in the north, the Baltic countries and northern Belarus.
In individual rivers, the maximum duration of severe droughts lasted from 3 to 161 days. According to SWLI, in 1992, hydrological drought covered eight sub-basins (out of 15) and had the maximum duration from 2 to 134 days in 2006–2012 and 5–112 days in 2019–2020—14 rivers in each year—with a maximum duration of 131 days in 2019 (Nevėžis River) and 161 days in 2020 (Merkys River). According to SWLI, in 1992, hydrological drought covered 8 sub-basins (out of 15) and had the maximum duration from 2 to 134 days; 2006–2012, with 5–112 days; and in 2019–2020 period—14 rivers in each year, with maximum duration 131 days in 2019 (Nevėžis River) and 161 days in 2020 (Merkys River). At the beginning of the study period, hydrological drought events were identified in the southeastern catchments, while, in the first decade of this century, they were indicated in the rivers of the central part of Lithuania. However, more recently (2016–2020), drought events were detected in each analyzed river catchment. The most prone to the hydrological drought was the Nevėžis river, where the percentage of severe droughts in the warm period was 7.81% (when, on average in Lithuania, it is 4.08%). These findings agree with a previous study [31
], which, based on three different drought indices, revealed different patterns of drought in the hydrological regions of Lithuania.
As was already mentioned, the lowest amount of drought events were detected in the Svyla river. Since it is considered intermittent [30
], we expected that this small river would distinguish itself by the most prolonged drought. A possible explanation for this case might be that our study applied the drought index based on river water levels. We suppose that during the period of low flow (which almost coincides with the warm period), the vegetation of the channel might have changed the hydrodynamics, i.e., the river stopped flowing. However, some water (the level of which can be measured) was still available in the river channel (the complex influence of aquatic macrophytes in regulating flow rates and water levels is discussed by [49
]). Therefore, the case of this intermittent river shows some limitations of the SWLI methodology.
The developed methodology was applied to forecasting hydrological drought. In general, the obtained results demonstrated that the selected river catchments would likely suffer from more extreme hydrological droughts, especially under RCP8.5 at the end of the century. At the same time, it is evident that, in climate change conditions, the behavior of river catchments with different physical-geographical features is complex and challenging to predict. These findings support the arguments that the results of drought projections highly depend on the regions and drought indices considered [50
It should be emphasized that the results obtained using the widely recognized stream drought index (SDI) developed by Nalbantis and Tsakiris [25
] with the standardized water level index (SWLI) proposed by Kugytė and Valiuškevičius [28
] are rather similar. SWLI can, therefore, be used as an operational index for hydrological drought monitoring and severe drought detection. It covers the essential criteria of a (hydrological) drought index [15
] as it is simple (can be understood by non-experts), easily calculated, based on available real-time data, has a physical meaning, is sensitive to various drought conditions and can be used for forecasting.