Evaluating changes in water levels during periods of normal flow and drought with a specific emphasis on water withdrawal

Freshwater scarcity is increasing rapidly due to global population growth (Veettil & Mishra 2016; Veettil et al. 2022). In addition, drought, which is the most expensive natural disaster and causes freshwater scarcity, affecting many water-using sectors, including domestic, agricultural, industrial, and public, further danger the security of freshwater resources that are already under pressure (Naumann et al. 2015; Veijalainen et al. 2019). In recent years, the adverse socio-economic effects of drought have escalated. The growing frequency of droughts underscores the urgency for proactive and preventive measures, moving beyond mere crisis management in disaster response. Effective strategies should incorporate innovative, well-coordinated responses tailored to address specific impacts and vulnerable populations (Wilhite & Pulwarty 2017).

70% of the water abstracted globally is used for crop irrigation to support food supply worldwide (Thenkabail et al. 2011; Koech & Langat 2018). Moreover, in developing countries where economies heavily rely on agricultural production, agricultural water usage can account for up to 90% of total water abstraction (FAO 2009). Consequently, agriculture stands as the predominant sector contributing to freshwater scarcity (Lamastra et al. 2014; Pellegrini et al. 2016; Ingrao et al. 2023). Sustainable Development Goal 6 (SDG 6) targets sustainable water management under the concept of water and sanitation across all sectors particularly in agriculture (Lee et al. 2016; Biancalani & Marinelli 2021). The primary objective of this goal is to enhance water security and ensure the sustainable protection of freshwater resources by promoting efficient water resource utilization.

Precipitation variability increasing with climate change and global warming leads to an increase in the frequency and intensity of drought events worldwide (Dai 2013; Mosley 2015). Water storage in lakes is vulnerable to climate change which responds rapidly to relevant impacts (Woolway et al. 2020). Shallow lakes are particularly vulnerable to extreme drought conditions because even slight changes in water level can affect a larger proportion of their surface area and volume compared to deeper lakes (Jones et al. 2013). During drought periods, freshwater lakes, which serve as an important water supply with their available storage, experience significant reductions in water levels (Edalat & Stephen 2019). Consequently, water supply security for reservoirs and other water resources becomes increasingly vulnerable under the combined pressures of climate change and other human-induced impacts (Zhang et al. 2023). Freshwater lakes play a critical role in mitigating the adverse impacts of drought on human activities and can serve as indicators of vulnerability to these impacts (Edalat & Stephen 2019).

During drought periods water deficiency in climate extends through the hydrologic cycle further decreasing groundwater levels, stream flows, and lake water levels (Mosley 2015). Drought is categorized into four types: meteorological, hydrological, agricultural, and socio-economic drought (Rajsekhar et al. 2015). Hydrological drought specifically refers to insufficient water for potential uses in managing a particular water resource (Zhong et al. 2020). Therefore, hydrologic drought conditions can be considered as directly representing threats to water security. Hydrologic drought is a product of meteorological drought but unlike meteorological drought, it can prolong for much longer durations (Huang et al. 2017). Under hydrologic drought, the impacts of anthropogenic activities such as uncontrolled irrigation, land use changes, and increased agricultural practices are often more significant than the impacts of climate change alone (Wanders & Wada 2015; Vicente-Serrano et al. 2022; Raposo et al. 2023).

Hydrological models, which simulate rainfall–runoff processes, are essential tools for quantifying potential water deficiencies during drought conditions and identifying effective measures to address them. Hydrological models simulate hydrologic response to climate conditions in basin, regional, or global scales (Kizza et al. 2013; Li et al. 2015). Thus, potential hydrologic impacts of climate change under future projections can be simulated by using hydrological models (Li et al. 2015).

Various types of models have been developed to simulate rainfall–runoff processes, each with different complexities such as physically based, conceptual, and artificial intelligence-based black box models (Gichamo et al. 2024). Each type of model offers advantages in simulating different aspects of real-world rainfall–runoff processes (Li et al. 2015). However, they also come with their own disadvantages, such as being non-user-friendly, requiring large amounts of data, or lacking clarity regarding their limitations (Devia et al. 2015). An ideal model generates results that reflect the real conditions as much as possible, with low data requirements and complexity (Suliman et al. 2015; Bihon et al. 2024). Hence, a model should not be more complex than necessary and should be fit for purpose (Beven & Young 2013; Horton et al. 2022).

In the calibration of hydrological models, one of the most critical challenges is the availability of continuous and long-term flow data (Patil & Stieglitz 2014; Folton 2024). Globally, resolution of the monitoring networks in most of the water basins is low resulting in significant data gaps (Samaniego et al. 2010; Medina & Muñoz 2020). The quality of available monitoring data also greatly influences the calibration process of hydrological models (Huang & Bardossy 2020). Typically, the more reliable and comprehensive the input data are, the less effort is necessary for the calibration (Jin et al. 2015). However, in regions where data deficiencies are prevalent, the calibration of hydrological models becomes much more complex and demanding (Jin & Jin 2020).

In developing countries, hydrometeorological data often suffer from insufficiency and poor quality due to inadequate monitoring networks (Boongaling et al. 2018; Hurtado-Pidal et al. 2020). This creates an urgent need to assess the usability of water resources for agriculture and other human activities under conditions such as drought, flood risks, hydropower potential, and the impacts of upstream land use changes, especially in rural basins where historical data are lacking (Melo et al. 2022). Conceptual models such as GR2M, GR4J, AWBM, HYMOD, MIKE NAM, HBV, TÉMEZ, etc. are less data-driven tools for such studies (Bihon et al. 2024). These models are based on mass balance principles and incorporate simplified energy balances, and therefore, resemble physically based models (Hrachowitz & Clark 2017; Ahmed et al. 2020). In hydrology, the use of conceptual rainfall–runoff models is particularly crucial for studying the impacts of climate change, especially when data are scarce (Patil & Stieglitz 2014; Yang et al. 2019; Folton 2024). Therefore, in regions with data deficiencies, conceptual models provide a versatile and practical means to conduct basic and rapid simulations of basin behavior (Ahmed et al. 2020).

The study area, Egirdir Lake, situated in the rural part of Türkiye – a developing country – is of significant importance as the second-largest potable water resource in the country. The lake plays a crucial role in supporting local ecosystems and sustaining the livelihoods of nearby residents. However, in recent years, the impacts of climate change, particularly drought, have led to a persistent decrease in the water level of Egirdir Lake. These climate-related impacts, compounded by agricultural water extraction from the lake, have exacerbated the decline in water levels to critical lows. This situation poses serious threats to both water and food security in the region.

Due to the decreasing water level in Egirdir Lake, changes in lake circulation patterns, increased water temperatures, and other related impacts, the lake has become more susceptible to eutrophication. This vulnerability is exacerbated by ongoing anthropogenic activities such as intensive agriculture and the discharge of untreated wastewater, which contribute significant nutrient loads. The modeling studies conducted as part of the ‘Revision of Egirdir Lake Special Provisions’ and related findings published in this study provided evidence of the potential decline in Egirdir Lake's water level <913 meters above sea level (mASL) in the near future. If this occurs, Egirdir Lake is anticipated to be physically separated into two sections in the ‘Kemer Bogazi’ area. Within that context, it is necessary to identify and propose urgent, effective measures to ensure the water security of Egirdir Lake, the second-largest freshwater resource in Türkiye. To the best of our knowledge, in the study area, there have been no previous studies to assess and test potential protection measure alternatives to satisfy sustainable use conditions under climate change impacts.

This study aims to address the critical research question: How can an optimized mitigation strategy be developed to safeguard Egirdir Lake's water security under the pressures of human activities and climate change? A key component of this research is the active involvement of stakeholders in developing mitigation alternatives. The study systematically evaluates various mitigation strategies and identifies an optimized cumulative measure that balances ecosystem protection with the sustainable use of the lake's resources. Additionally, this study examines policies designed to ensure water security, particularly for communities that depend on Egirdir Lake. It emphasizes the necessity of well-coordinated and timely interventions to mitigate risks and preserve the lake as a crucial resource. The findings provide guidance on developing effective water management strategies for lake systems with scarce monitoring data under anthropogenic and climatic pressures.

AQUATOOL + a Decision Support System (DSS) software developed by Valencia Technical University (Solera et al. 2015), is utilized for hydrological and water management modeling of the Egirdir Lake drainage basin. The EVALHID module of AQUATOOL + is employed for rainfall–runoff simulation of drainage basins, allowing for hydrological modeling of the Egirdir Lake basin. Within the scope of the study, (i) Témez Hydrological Model (Témez 1977), (ii) Hydrologiska Byråns Vattenbalansavdelning (HBV) (Bergström 1976) and (iii) GR2M Hydrological Model (Mouelhi et al. 2006) hydrological models are tested. The simulation performances of these models are compared and the model that provides the highest performance at each calibration point is selected.

The Témez Hydrological Model (Témez 1977), commonly used in Spain and other countries for the management of water resources (Jódar et al. 2017; Zambrano Mera et al. 2018; Rivadeneira Vera et al. 2020; Muñoz-Mas et al. 2024), is a four-parameter-based deterministic water balance model (Estrela Monreal et al. 1999; Marcos-Garcia et al. 2017). The HBV model, an eight-parameter-based hydrological model, is successfully used in more than 80 countries (Li et al. 2015) in various basins with various physical and climatological characteristics (Rientjes et al. 2013). Over the years HBV is developed through various studies to become a versatile multi-purpose model for the simulation of flood projections (Kobold & Brilly 2006), water resources management (Medina & Muñoz 2020) and climate change impact assessment (Huang & Bardossy 2020; Abdulahi et al. 2022). The GR2M model, a two-parameter-based model, is widely utilized worldwide (Okkan & Fistikoglu 2014; Lyon et al. 2017; Fathi et al. 2019; Llauca et al. 2021; Ditthakit et al. 2023; Mahdaoui et al. 2024). There are various versions of this model that have proven effective (Kabouya 1990; Kabouya & Michel 1991; Makhlouf 1994; Makhlouf & Michel 1994; Mouelhi 2003; Mouelhi et al. 2006). This study employs the latest GR2M version developed in 2006 (Mouelhi et al. 2006). Témez, HBV, and GR2M models are semi-distributed models rooted in mass balance principles. These models with low data and computing source requirements are practically used for hydrological simulations for areas that suffer from data deficiency (Oñate-Valdivieso et al. 2016; Bui et al. 2020; Ditthakit et al. 2021).

Meteorological time-series data used in the model is obtained from grid-based precipitation, temperature, and evapotranspiration data generated within the scope of the project titled ‘Impact of Climate Change on Water Resources in Türkiye’ conducted by the Turkish Directorate General of Water Management (DGWM) (DGWM 2018). Grid-based meteorological data with a 10 × 10 km resolution is generated using the Inverse Distance Weighting (IDW) approach, incorporating data from the 12 closest ground-based monitoring stations with the shortest 25-year observational record for each grid.

Snowmelt data, including snow depth and snow water equivalent, are unavailable for the study area. Consequently, the hydrological modeling is based solely on the available hydrometeorological data to minimize uncertainties in the model.

Most of the available studies on climate change impacts focused on hydrological and ecological responses of streams typically use naturalized flow projections (Poff et al. 1997; Terrier et al. 2021) due to the complexity of simultaneously assessing climate change and changes in water use patterns (Terrier et al. 2021). In this study, for the calibration of stream flows in the Egirdir Lake Basin naturalized flows in Antalya Basin are utilized. Recently, certain studies have suggested that the traditional approach of hydrological model calibration–validation processes may be ineffective and potentially misleading. It has been proposed that instead of separating the dataset into calibration and validation sets, the entire dataset should be used for calibration to identify hydrological parameters accurately (Arsenault et al. 2018). This approach is particularly relevant in situations where data are limited. In this study, due to data deficiencies in the Egirdir Lake area, the entire data set is utilized for calibrating the hydrological model. Validation of the model is conducted using observed water levels as benchmarks compared to modeled lake water levels during the following water management modeling stage. Importantly, this stage in which validation is conducted does not involve parameter optimization after water abstractions are input into the model.

The calibration period spans from 1990 to 2014, encompassing 25 water years, aligning with the conclusions of the flow naturalization study. The naturalized flows are derived using a water balance approach that considers upstream–downstream relationships under conditions free from anthropogenic interference (SHW 2018).

Once the hydrological modeling stage is completed, the SIMGES module is utilized to determine hydrologically altered flows. At this second stage, the model outputs are validated with observational data. SIMGES module of AQUATOOL + is used to simulate stream flows in the basin and water levels in Egirdir Lake under various water abstractions and effluent discharges. The main water abstraction in the basin is for agricultural irrigation, with mean monthly volumes obtained from the Hydrological Master Plan of Antalya Catchment (SHW 2018). Egirdir Lake also serves as the potable water supply for several settlements and facilities in the region, including the drinking water supply of Isparta Province in the Antalya Catchment. Historical monthly records of direct water abstractions for all sectors (agriculture, drinking water, and industry) from Egirdir Lake, as well as lake water levels, are obtained from the SHW for the period between 1980 and 2021. The simulated water level in Egirdir Lake is validated for the period between 2016 and 2021 with observed data.

Lake inflow data from the EVALHID model and observed water levels in Egirdir Lake are analyzed for the determination of the reference drought period in the study. For this purpose, annual flow anomalies and the variation in the lake water level are comparatively evaluated for the period between 1990 and 2021 water years.

Thirdly, for the assessment of the drought severity, the Water Depletion Index (WDI) is calculated based on monthly water availability and consumption for the selection of the Reference Drought Period. WDI, suggested by Brauman et al. (2016) is based on the ratio of the consumptive water use to the total inflows into the system.

Water level projections for Egirdir Lake are generated under two scenarios: a normal scenario (Scenario 1) and a drought scenario (Scenario 2). The simulation aims to test whether the water level drops below a critical threshold for the lake in either scenario.

The duration of the reference drought period is determined based on the findings reported in the Antalya Basin Drought Management Plan by DGWM (2018). This study analyzed historical data from 1965 to 2016 using various 12-month hydrological drought indices, including the Palmer Drought Severity Index, Standardized Precipitation Index, Standardized Precipitation–Evapotranspiration Index, and Palmer Hydrological Drought Index, to identify dry periods and their durations in the basin. The results revealed 26 historical hydrological drought periods, ranging from 2 to 19 months in duration. Therefore, to ensure a conservative approach, the dry period in the model is set to 3 years.

The effect of site-specific mitigation measures on the water level change is also tested through simulations under both scenarios. The mitigation measures focus on actions to improve water use efficiency, such as using drip irrigation instead of flood irrigation, changing irrigation patterns for less water consumption, recycling treated wastewater effluent as irrigation water, and using water derivation from other drainage basins. These site-specific measures include:

• Application of 50 and 30% restricted/deficit irrigation alternatives;

• Utilizing treated effluent from the Isparta Province Wastewater Treatment Plant (WWTP) for irrigation, replacing water abstracted from Egirdir Lake;

• Increase of average total annual derivation to Egirdir Lake to 83.97 Mm³ through 42.10 Mm³ additional derivation from Bagilli Regulator in Antalya Catchment; and

• Rehabilitation and conversion of the irrigation systems in the Egirdir basin from flood irrigation to drip irrigation type.

As seen in Table 1, the HBV model exhibits the best modeling performance for NF 1, NF 2, NF 3, and NF 5. The NSE and NNSE values obtained by the HBV model for the calibration period are 0.48, and 0.66 in NF 1, 0.44, and 0.64 in NF 2, 0.47 and 0.65 in NF 3, and 0.38 and 0.62 in NF 5, respectively. For these calibration points PBIAS values in NF 1, NF 2, NF 3, and NF 5 are 3.36, 2.16, 1.13, and 1.60%, respectively. For NF 4, located in the western part of the study area, GR2M demonstrates the best modeling performance. NSE and NNSE values in NF 4 obtained by the GR2M model are 0.63 and 0.73, respectively. The PBIAS value for the calibration point is 1.75%.

The thresholds for the performance indicators are dynamic values obtained from the compilation of the published studies. Moriasi et al. (2015), in their study, emphasized that the calculated thresholds are obtained from the compilation of only the study outcomes in which good performances could be achieved. Moreover, in their study, it is recommended that thresholds are proposed through the processing of a more extensive dataset including published and unpublished studies (Moriasi et al. 2015). Hence there are no definitive threshold values for the evaluation of the statistical performance indicators. However, in this study, like other similar studies, the commonly applied threshold values in the literature are taken into consideration. Accordingly, for flow rate simulations NSE > 0.75 value indicates a good modeling performance, and NSE > 0.36 indicates a satisfactory modeling performance (Motovilov et al. 1999; Muñoz et al. 2011; Luo et al. 2012; Hu et al. 2024). Regarding the PBIAS value PBIAS < ±5% indicates very good modeling performance. For satisfactory modeling outcomes PBIAS < ±15% is expected (Moriasi et al. 2015). In the study area, calculated NSE values range from 0.38 to 0.63 and PBIAS values range between 1.13 and 3.36%. Hence, NSE and PBIAS values calculated based on the model outcomes verify satisfactory modeling performance.

In the second stage of the modeling study, the SIMGES module is used in integration with the EVALHID module, utilizing outputs from the hydrological model. SIMGES module simulates the water balances for hydrologically altered flow by including extractions for water demands, evaporation, and water transfers (Pedro-Monzonís et al. 2016). Providing amenable water quality Egirdir Lake water is utilized both for drinking and agricultural irrigation purposes. The lake also receives water transferred from the Koprucay subbasin through the Yılanlı Derivation Line. SIMGES module enables input of these water abstractions and transfers in the basin water balance model.

Water stress in Egirdir Lake is primarily caused by two factors: a decrease in surface runoff within the drainage basin (due to hydrologic drought) and an increase in water consumption from the lake and its drainage basin. To assess the relationship between water scarcity and drought periods, change in runoff, and water level in time is examined. Additionally, the impact of consumption and hydrologic drought pressures during water scarcity periods, as identified by the WDI, is evaluated.

The analysis reveals that annual water storage first declined below the mean annual storage volume of 3,428.69 Mm³ after 1990, a period characterized by several successive years of negative runoff anomalies. Although successive positive runoff anomalies in the following years led to an increase in storage volume above the mean annual storage, the overall trend shows an apparent decline in annual storage throughout the study period. The highest negative anomaly rates, indicating the lowest flows, are observed in the water years 2001 (−44%) and 2021 (−50%).

The comparison of calculated WDI, consumptive water use, and annual runoff confirms that the primary cause of the increased water scarcity in the 2001 water year, which is also reflected in the decreasing water level in the lake, is hydrological drought rather than the consumptive water use that remains below the mean consumption during this period. On the other hand, after 2007 the consumptive water use is seen to increase tremendously creating severe water scarcity despite higher annual runoff into the lake. Consequently, Egirdir Lake experiences high water stress and is characterized as depleted in 2007, 2008, 2010, 2011, 2012, 2016, and 2021 water years. The main cause of the high water scarcity after 2007 is the substantial increase in consumptive water use from the lake and its drainage area.

To simulate the impact of future hydrological drought on Egirdir Lake, the 2001 water year is selected as the reference drought period for water level projections and testing of mitigation measures.

Mitigation measures are developed in line with the 2023–2033 Action Plan and Strategy Document for Water Efficiency within the Framework of Changing Climate (Ministry of Agriculture & Forestry 2023) and Methodologic Guideline on Water Efficiency in Agriculture Sector (DGWM 2021). Alternatives 1, 2, 3 include combinations of measures including deficit irrigation, rehabilitation of the irrigation systems and reclamation of treated wastewater effluent as well as freshwater derivation into the Egirdir Lake.

Deficit irrigation or conventional deficit irrigation is a water conservation method applied in water scarce regions or during water scarce periods and aims to restrict the irrigation season to the periods during which crops are more vulnerable to the drought stress without significant loss in productivity (DGWM 2021). Deficit irrigation has been verified to provide water use efficiency and water savings up to 53% compared to the full irrigation regime (Saitta et al. 2021). The use of deficit irrigation regime is classified as a measure that fosters the ‘absorptive capacity’ of agricultural production against the impacts of drought by Lankford et al. (2023). Alternative 1 and Alternative 2 propose irrigation reductions up to 30 and 50%, respectively, by 2050 to maximize water savings from Egirdir Lake's irrigation withdrawals. In contrast, Alternative 3 applies the deficit irrigation approach only during the 2025–2026 water years.

All alternatives incorporate the rehabilitation of irrigation systems by transitioning to a closed pipe water distribution network and installing drip irrigation systems. These technologies have been shown to enhance crop yields and achieve over 85% water savings in agricultural production (Bhalage et al. 2015; Muralikrishnan et al. 2021), thereby reducing the water footprint of crop cultivation (Nouri et al. 2019). The modernization of irrigation infrastructure is considered an adaptive measure to mitigate drought impacts (Lankford et al. 2023). In fact, rehabilitation of irrigation systems to use more efficient irrigation and conveyance systems is estimated to provide 35% water savings in Mediterranean region and even higher savings in Turkey and is designated as a strategy to compensate impacts of climate change and population growth regarding the water and food security (Fader et al. 2016). Beyond water conservation, drip irrigation offers multiple benefits, including higher productivity and crop yields, reduced weed growth and soil erosion, lower salinity issues, labor and energy savings, and improved fertilizer use efficiency (Bhamoriya & Mathew 2014). However, in developing countries, integrating advanced irrigation technologies presents challenges such as high initial investment costs, unreliable electricity supply, limited technical expertise, and lower public acceptance of new systems. Therefore, implementing these adaptive measures requires additional investments, incentives, and capacity-building efforts, including training programs to enhance adoption (Dawit et al. 2020; Oiganji et al. 2025).

The third measure included in all three alternatives is the reclamation of treated wastewater in part of the irrigation areas in the region. As water scarcity intensifies, using treated wastewater instead of freshwater for irrigation emerges as a sustainable strategy for preserving water resources. This approach is particularly beneficial in semi-arid regions, as treated effluents provide consistent water availability throughout the year, unlike freshwater sources, which fluctuate based on climatic conditions (Ungureanu et al. 2018). Additionally, wastewater reclamation offers a cost-effective alternative to desalination, which is commonly used in semi-arid and arid regions (Khan et al. 2022). However, if not properly treated, reclaimed water may pose risks such as heavy metal contamination, public health concerns, and salinity issues (Ungureanu et al. 2018; Khan et al. 2022; Kama et al. 2023). Therefore, to ensure its safe use, wastewater reclamation must be accompanied by adequate treatment processes and supported by capacity-building initiatives, including the development of wastewater reuse guidelines, public education, and awareness programs (Khan et al. 2022; Kama et al. 2023).

The final measure in the cumulative strategy is the freshwater derivation to Egirdir Lake. This approach utilizes the existing Yılanlı Derivation Line to transfer additional freshwater into the lake, aiming to enhance its sustainability and ecological integrity. Given that Egirdir Lake is a crucial local water source, this measure supports its long-term viability and helps mitigate the impacts of water scarcity.

From a socio-economic perspective, Egirdir Lake is of critical importance to the local community. In fact, sustainable water resource management that ensures the balance between conservation and utilization is essential, particularly to safeguard vulnerable groups whose sole livelihood depends on the lake. To achieve this, public participation meetings that bring decision-makers and users together were organized to ensure active participation of all stakeholders, including the local population, in the decision-making process. It is concluded that Alternative 3 is the most suitable option for sustainable water resource management. Based on these findings the relevant measures are brought into force by the decision of the Ministry of Agriculture and Forestry as of 15 June 2024.

In this study, proactive and preventive strategies are created using the hydrological modeling EVALHID module and the SIMGES water management modeling module from the AQUATOOL + DSS to ensure the water security of Egirdir Lake, Türkiye's second-largest freshwater resource. Located in a rural area of this developing country, Egirdir Lake faces challenges due to limited data and its multiple uses as a water resource.

An analysis of the lake's water balance reveals that evaporation losses are compensated through natural processes. However, water withdrawals for irrigation disrupt this balance, leading to a decline in the lake's water level. The simulation of Egirdir Lake's water level suggests that it is critical to apply mitigation measures, and the lake could fall below the critical drying point, potentially leading to the water body dividing into two separate sections. In Scenario 1 (normal scenario) without mitigation measures, Egirdir Lake is expected to physically split into two parts after 2038. In Scenario 2 (drought scenario) without mitigation measures, this division is projected to occur earlier, after 2028. However, the modeling study suggests that implementing mitigation measures can prevent or reduce the impact of the lake splitting into two parts. The study assesses the effectiveness of three cumulative measures (Alternatives 1, 2, and 3) under two climate scenarios (Scenario 1 and Scenario 2).

The cumulative mitigation measure, Alternative 3, which includes applying a 50% irrigation limitation, rehabilitating irrigation systems to convert to drip irrigation, recycling effluent from the Isparta WWTP for irrigation, and increasing water transfer to the lake, is identified as the most effective solution. The optimum Alternative of cumulative measures identified in this study was evaluated by the decision-makers and has been deemed sufficient and implemented as the revised Egirdir Lake Special Provisions.

Egirdir Lake, the study area, serves as a vital water source for meeting the demands of various sectors. However, due to scarce monitoring data and the undeniable impacts of climate change, ensuring water security requires the development of urgent action plans. The methodology and findings of this study provide guidance for such efforts, enhancing the study's originality and significance.

The study's findings offer important insights for regional water security and highlight the need for more comprehensive data collection and modeling in future research to address data limitations. The meteorological data in the area are scarce, and there is no available information on snowmelt, such as snow depth or snow water equivalent. To reduce uncertainties in the model caused by these data gaps, the analysis relies on the available hydrological data. Future research should focus on incorporating snowmelt data and evaluating its impact on lake discharges.

The authors extend thanks to the ongoing Project titled ‘Technical Assistance on Preparation of River Basin Management Plans for Six Basins (EuropeAid/140294/IH/SER/TR)’. The authors thank the Directorate General for Water Management, the State Hydraulic Works, and the General Directorate of Meteorology of Türkiye for providing data.

This research received no external funding.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.