ORIGINAL RESEARCH article

Front. Water, 16 October 2026

Sec. Water Resource Management

Volume 8 - 2026 | https://doi.org/10.3389/frwa.2026.1932313

A multi-scale framework for identifying and validating managed aquifer recharge sites in semi-arid Kazakhstan

  • 1. Akhmedsafin Institute of Hydrogeology and Environmental Geoscience, Satbayev University, Almaty, Kazakhstan

  • 2. School of Information Technology and Engineering, Kazakh-British Technical University, Almaty, Kazakhstan

Abstract

Managed Aquifer Recharge (MAR) has emerged as an effective approach for enhancing groundwater resources in water-scarce regions under increasing climate variability and growing water demand. However, most GIS-based MAR studies are limited to regional suitability mapping, whereas the hydrogeological feasibility of identified sites is rarely verified under field conditions. This study presents a multi-scale framework integrating regional GIS-based screening with detailed site-scale verification to improve the reliability of MAR suitability assessment in southeastern Kazakhstan. This research was conducted within the Zhetysu Alatau–Balkhash First-Order Hydrogeological Basin to develop and evaluate a multi-scale framework for identifying suitable MAR sites. Regional suitability assessment combined Geographic Information Systems (GIS), the Analytic Hierarchy Process (AHP), Monte Carlo simulation, Fuzzy Gamma analysis, remote sensing, and regional hydrochemical data. Based on the regional assessment, the Byzhy River Basin was selected as a representative site for detailed hydrological, hydrogeological, hydrochemical, LiDAR-based, and field investigations to verify regional suitability under actual hydrogeological conditions. The regional suitability assessment identified 21.82% (58.41 km2) of the representative study area as highly suitable for MAR, while Monte Carlo simulation and Fuzzy Gamma analysis confirmed the robustness of the suitability model. Hydrological assessment indicated a positive long-term mean potential water balance of 27.69 mm·year−1, suggesting sufficient water resources for MAR. Three double-ring infiltration tests yielded a mean field-saturated hydraulic conductivity of 1.92 m·day−1 (range: 1.86–1.96 m·day−1), indicating high infiltration capacity of the Quaternary alluvial deposits. Groundwater mineralization remained below 1 g·L−1, and hydrogeological investigations confirmed a 50 m thick Quaternary alluvial aquifer suitable for MAR. Long-term daily river discharge records indicated a potential annual recharge volume of 17.44 million m3, representing only 2.33% of the effective aquifer storage capacity (746.8 million m3), demonstrating substantial storage potential. The close agreement between the regional suitability assessment and field investigations demonstrates that the multi-scale framework substantially reduces uncertainty in GIS-based MAR site identification. By integrating regional suitability mapping with hydrological, hydrogeological, hydrochemical, LiDAR, and field investigations, the framework provides a transferable methodology for identifying and verifying MAR sites in arid and semi-arid regions, supporting groundwater management and long-term water resource planning.

1 Introduction

Groundwater is one of the main sources of freshwater and plays an important role in ensuring water security, particularly under conditions of climate change and increasing water scarcity (Taylor et al., 2013; Famiglietti, 2014) in Kazakhstan, where a large part of the territory is located within arid and semi-arid climatic zones, the sustainable use and conservation of groundwater resources are of strategic importance. Under these conditions, Managed Aquifer Recharge (MAR), which is based on the controlled infiltration of surface water into aquifers, is considered one of the most promising approaches for restoring groundwater reserves and supporting their sustainable use (Gleeson et al., 2012; Wada et al., 2010; Scanlon et al., 2002; Green et al., 2011).

In this context, Managed Aquifer Recharge (MAR) is considered a promising approach for the restoration and sustainable management of groundwater resources through the infiltration of surface water, floodwater, and treated water into aquifer systems. Managed Aquifer Recharge (MAR) has been widely adopted in Europe and North America as an effective approach for sustainable water resource management (Bouwer, 2002; Dillon et al., 2019). While this method has demonstrated positive outcomes in several countries, its implementation remains very limited in Central Asian states. Therefore, the adoption of groundwater management practices such as MAR represents a critical need for the region.

Recent MAR studies have advanced from conventional GIS-based suitability mapping toward broader feasibility assessments incorporating water availability, water demand, stakeholder participation, sensitivity analysis, and hydrogeological criteria (Martins et al., 2024; Panagiotou et al., 2024; Ebrahim et al., 2024; Seif et al., 2024; Chekirbane et al., 2025; Bilal et al., 2025). Numerical groundwater modeling has also been used to assess site-specific MAR performance and the redistribution of excess surface water (Ferencz et al., 2024). However, a gap remains in linking regional-scale MAR suitability and uncertainty assessment with field-based verification and quantitative evaluation of recharge-water availability at a representative site, particularly in semi-arid Central Asian environments. The present study addresses this gap through a multi-scale regional-to-site framework combining spatial screening, uncertainty analysis, hydrochemical assessment, high-resolution terrain characterization, field infiltration testing, and quantitative water-availability assessment.

For developing countries, this approach offers significant benefits, including ecosystem preservation, support for livelihoods, promotion of economic development, and mitigation of drought impacts (Tuinhof and Heederik, 2003; Dillon et al., 2009). Moreover, the relative reliability of groundwater compared to surface water resources makes it a key component of water security in arid regions (Taylor et al., 2013).

However, due to the complex interactions of geological, geomorphological, hydrological, climatic, and anthropogenic factors, groundwater availability and quality are spatially heterogeneous, a challenge that is particularly evident in Central Asia. Semi-arid basins are characterized by highly variable infiltration conditions, complex lithological structures, and spatial heterogeneity of groundwater recharge processes, which complicate MAR suitability assessment. The identification of suitable areas for MAR implementation is aimed at integrating hydrogeological data, remote sensing materials, and Geographic Information Systems (GIS) (Kazakis et al., 2015).

In European countries, MAR studies have primarily focused on sustainable water resource management and the protection of groundwater quality. In Mediterranean countries such as Spain, Italy, and Greece, GIS, high-resolution Digital Elevation Models (DEM), LiDAR data, the Analytic Hierarchy Process (AHP), and GIS-based Multi-Criteria Decision Analysis (MCDA) methods have been applied to identify areas suitable for infiltration (Sallwey et al., 2019; Kazakis et al., 2018). A distinctive feature of European research is its emphasis on environmental sustainability and the long-term efficiency of MAR systems (Stefan and Ansems, 2018).

In the Americas, particularly in the United States, MAR technologies have been investigated mainly from engineering and hydrogeological perspectives. Studies conducted in California integrate GIS, hydrogeological modeling, Monte Carlo simulation, and field hydraulic testing (Russo et al., 2014; Scanlon et al., 2016). The primary objective of these studies is to capture floodwaters and ensure the long-term preservation of groundwater resources. Unlike conventional fixed-weight approaches, the integration of Monte Carlo simulation enables uncertainty-aware evaluation of MAR suitability and improves the robustness of spatial decision-making (Neshat et al., 2014).

In Asia, the Middle East, and Africa, MAR research is largely driven by the need to address water scarcity and drought conditions. In countries such as India, China, Saudi Arabia, and several African states, Landsat and Sentinel satellite imagery, DEM data, and GIS–AHP approaches have been employed to delineate areas suitable for infiltration (Khan et al., 2023; Machiwal et al., 2011). The main characteristic of these studies is the application of MAR systems for agricultural purposes, water security enhancement, and the establishment of strategic water reserves during severe drought periods.

In Central Asian countries, studies on the application of MAR using GIS, DEM, and remote sensing techniques have been conducted; however, compared to Europe and the United States, this field remains relatively underdeveloped. In Kazakhstan, Uzbekistan, Kyrgyzstan, and Tajikistan, MAR-related research is associated with issues of water scarcity, irrigation, and seasonal runoff management (Micklin, 2007; Karthe et al., 2015). GIS-based hydrological and hydrogeological investigations are currently being carried out within the Syr Darya and Amu Darya river basins.

However, most existing MAR suitability studies in Central Asia are primarily based on conventional GIS–AHP approaches and often lack uncertainty assessment, field hydraulic validation, and high-resolution terrain analysis. As a result, the reliability and spatial accuracy of MAR suitability mapping in heterogeneous arid and semi-arid environments remain insufficiently investigated. Furthermore, the limited integration of hydrogeological field data with GIS-based spatial modeling restricts the practical implementation of MAR technologies in Kazakhstan.

In Kazakhstan, research on MAR technologies has only recently begun to develop. Current studies employ GIS, remote sensing, DEM, and AHP methods to identify areas suitable for groundwater recharge. Studies by (Onglassynov et al., 2025; Sotnikov et al., 2025; Sallwey et al., 2024; Amanzholova et al., 2024; Jabassov et al., 2026) were among the first GIS-based studies on MAR in Kazakhstan to apply the analytical hierarchy process (AHP) to identify zones with high groundwater recharge potential. These studies incorporated factors such as slope, topography, lithology, groundwater depth, normalized difference vegetation index (NDVI), drainage density, and climatic conditions. However, Monte Carlo simulations and integrated hydraulic testing approaches remain insufficiently applied in MAR assessments in Kazakhstan. Therefore, current research is focused on developing integrated MAR models that combine GIS, remote sensing, and field-based hydrogeological investigations.

The Zhetysu Alatau–Balkhash First-Order Hydrogeological Basin located within the Almaty and Zhetysu regions, is delineated in the Hydrogeological Atlas of Kazakhstan prepared (U.M. Akhmedsafin Institute of Hydrogeology and Environmental Geosciences, 2022), is one of the key hydrogeological regions of southeastern Kazakhstan. The area is characterized by semi-arid natural and climatic conditions, agricultural development, and increasing pressure on surface water and groundwater resources, which highlights the need to introduce modern approaches for their sustainable management. Research on MAR in Kazakhstan is still at an early stage; therefore, this study provides a comprehensive assessment of the MAR potential of the study area.

Our study is aimed at assessing the potential of Managed Aquifer Recharge (MAR) using an integrated regional-to-site-scale framework. This objective was achieved through a two-stage approach. In the first stage, a regional assessment of the suitability of the Zhetysu Alatau–Balkhash First-Order Hydrogeological Basin was conducted using eight spatial criteria, GIS-based AHP, Monte Carlo uncertainty analysis, and regional hydrochemical data. Based on the regional assessment, the Byzhy River Basin was selected as a representative site for detailed investigations. In the second stage, LiDAR-based terrain analysis, hydrogeological investigations, water balance assessment, infiltration and filtration tests, and an assessment of the volume of river runoff that can be used for artificial groundwater recharge were carried out. This two-stage approach makes it possible not only to identify the most promising areas for MAR implementation at the regional scale, but also to quantitatively assess their practical potential and the efficiency of using river runoff under real hydrogeological conditions.

2 Materials and methods

2.1 Study area

The study area is located in southeastern part of Kazakhstan, encompassing the Almaty and Zhetysu regions, and is situated within the Zhetysu Alatau–Balkhash First-Order Hydrogeological basin (Figure 1). The hydrogeological basin is characterized by complex hydrogeological conditions and significant groundwater resources (U.M. Akhmedsafin Institute of Hydrogeology and Environmental Geosciences, 2022; Smolyar and Mustafayev, 2007).

Figure 1

The study area is characterized by extensive Quaternary alluvial, alluvial–proluvial, and lacustrine–alluvial aquifers and sedimentary deposits distributed across arid and semi-arid environments, which are considered favorable for MAR implementation (Smolyar and Mustafayev, 2007). It is bounded by Lake Balkhash to the north and northwest, the high mountain ranges of Kungey Alatau and Ile Alatau to the south, and Zhetysu Alatau to the east. The territory runs from latitude 43°01′–46°25′N and 75°10′–82°15′E longitude, covering a total area of 267,688 km2 (Figure 1), is characterized by relatively gentle topography sloping toward Lake Balkhash, with mountainous terrain dominating the southeastern part and elevations ranging from 269 to 4,925 m above lake level. Several perennial rivers, including the Ili River, Koksu River, Karatal River, Aksu River, and Lepsy River, flow through the study area toward the north, contributing significantly to regional groundwater recharge processes and surface water dynamics (Environment and Natural Resources Kazakhstan Institute of Geography (KIG), 2022).

According to the hydrogeological conditions, both unconfined and confined groundwater occur within the study area. Unconfined groundwater is distributed within modern alluvial and Middle Quaternary lacustrine–alluvial deposits. Confined groundwater occurs within Lower Quaternary, Neogene, and Paleogene formations (Veselov, 2002).

The natural conditions of the region are represented by several climatic zones, ranging from glacial environments in the south to desert landscapes in the north. The climate is sharply continental due to the considerable distance from oceans and seas. It is characterized by predominantly hot and dry weather conditions, a high number of cloudless days, and pronounced seasonal and diurnal air temperature fluctuations. Summers are generally hot, whereas winters are moderately cold and relatively mild. Humid years are frequently followed by arid periods accompanied by droughts and dry winds. Intensive wind activity is often associated with snowstorms and dust storms, particularly across the plains. The average summer temperature across the plains ranges from 20 to 26 °C and decreases with elevation by approximately 0.4–0.8 °C per 100 m increase in altitude. During summer, air temperatures may reach up to 45 °C, whereas in winter they can decrease to −40 °C (Kazhydromet, 2025).

Atmospheric precipitation is unevenly distributed: the lowest amounts are observed near Lake Balkhash, while the highest values occur in the high-mountain regions of the Ile and Zhetysu Alatau. On the sandy plains of the Southern Balkhash region, annual precipitation ranges from 200 to 300 mm, whereas on the northern slope of the Ile Alatau at an elevation of 3,500 m it reaches 1,300 mm (Kazhydromet, 2025). However, precipitation occurring during the summer period, combined with high air temperatures, reduces the moisture availability of the territory. Throughout the year, 40–50% of precipitation falls during the vegetation period; however, a significant part is lost through evaporation and surface runoff, which reduces the role of atmospheric precipitation in groundwater recharge.

2.1.1 Representative study sub-area: Byzhy river basin

Within the regional study area, the Byzhy River Basin in Figure 1 was selected as a representative sub-area due to its arid natural-climatic conditions and hydrogeological features favorable for assessing the MAR suitability assessment. The basin was selected as a representative site because it reflects the hydrogeological and climatic conditions commonly observed across the broader study area.

The Byzhy River originates in the Zhetysu Alatau and is mainly fed by snowmelt and glacier meltwater, resulting in pronounced seasonal runoff variability and strong hydraulic interaction between surface water and groundwater. The river is approximately 177 km long, and the catchment area covers about 5,500 km2 (Kaldzbazev et al., 2016; Issaldayeva et al., 2023).

The basin is characterized by distinct altitudinal zonation, including mountainous areas, foothill plains, river valleys, and extensive Quaternary alluvial and alluvial–proluvial deposits. These deposits provide favorable hydrogeological conditions for groundwater accumulation and infiltration processes, making the basin suitable for detailed hydrogeological investigation (Smolyar and Mustafayev, 2007).

The representative study area was selected to verify the regional MAR suitability assessment through detailed field investigations, including LiDAR-based terrain analysis, hydrogeological characterization, water balance assessment, infiltration and hydraulic testing, and river discharge evaluation. These investigations provide the basis for assessing the practical potential of Managed Aquifer Recharge under representative hydrogeological conditions.

2.2 Methodology of research

2.2.1 Input data

The study identified potential groundwater zones within the Zhetysu Alatau–Balkhash First-Order Hydrogeological basin, covering the territories of the Almaty and Zhetysu regions, based on various input datasets. The semi-arid Byzhy river basin was selected as a representative study sub-area within the investigated region. The datasets used in the study included remote sensing (RS) products and data obtained from monitoring stations. All collected datasets were integrated using Geographic Information Systems (GIS) and the Analytic Hierarchy Process (AHP) and Monte Carlo simulation to assess MAR suitability and groundwater recharge potential (Yazdani Mina et al., 2026; Kazem et al., 2026). To further evaluate groundwater recharge potential and the suitability of the study area for MAR implementation, water balance analysis was performed, while the Monte Carlo method was used to assess the stability of weighting coefficients and the reliability of the model results. Table 1 gives characteristics and sources of the datasets used for generating the thematic layers.

Table 1

S. noThematic layerSourceDateResolutionPurpose
1Geological/LithologicalGLiM dataset (Esri living atlas)2013250 m (resampled to 30 m)Geological map
2SoilFAO/UNESCO digital soil map of the world (DSMW).200330 mSoil map
3Land use/land coverESA WorldCover (Esri living atlas)202110 mLULC map
4Precipitation (regional study area)WorldClim v2.11970–200030 arc-sec (~1 km), resampled to 30 mPrecipitation map
5ASTER DEMNASA SRTM DEM200030 mSlope map
6ASTER DEMNASA SRTM DEM200030 mDrainage density map
7Groundwater depth dataThe Atlas of Hydrogeological Maps of Kazakhstan202230 mGroundwater level map
8Groundwater mineralizationThe Atlas of Hydrogeological Maps of Kazakhstan202230 mGroundwater Mineralization Map
9Water balanceWater balance assessment (watershed- Byzhy)
9.1PrecipitationTerraClimate2000–20254 kmPrecipitation data
9.2EvapotranspirationERA5-Land2000–20254 kmEvapotranspiration data
9.3Surface runoff/river dischargeTerraClimate2000–20254 kmRiver discharge data
9.4Soil MoistureTerraClimate2000–20254 kmSoil Moisture data
10Infiltration capacity and field-saturated hydraulic conductivity (K_fs)Field tests2025Assessment of Infiltration Capacity (watershed Byzhy)
11Hydrochemical data (Piper–Durov Diagrams)Field water sampling and laboratory analysis2025Identification of hydrochemical facies
12UAV and LiDAR survey dataField survey20255 cmDetailed topographic mapping and validation of MAR suitability at the selected Byzy site
13Hydrogeological conditionsHydrogeological map of Kazakhstan (1:200,000 scale)2024Hydrogeological conditions and cross-sections for MAR

Characteristics of the input datasets and thematic layers used in the study.

During the database development process, eight thematic layers were generated based on remote sensing products. In particular, the drainage density and slope layers were prepared using the ASTER DEM digital elevation model with a spatial resolution of 30 m, while soil data were obtained from the FAO digital soil map (FAO/UNESCO Digital Soil Map of the World (DSMW), an international soil dataset published by FAO, Rome, Italy). The Geological and Land Use/Land Cover layers was identified using Esri living Atlas (Esri, United States) and satellite imagery data, while precipitation data were obtained from the WorldClim database (University of California, Davis, United States).

The groundwater level and mineralization layers were developed using Hydrogeological Atlas of Kazakhstan data prepared by the Akhmedsafin Institute of Hydrogeology and Environmental Geosciences.

For the representative study area, the Byzhy River Basin, precipitation, surface runoff, and soil moisture data were obtained from the TerraClimate dataset (University of Idaho, 2026), whereas actual evapotranspiration data were derived from the ERA5-Land dataset (European Union—Copernicus Climate Change Service (C3S), implemented by ECMWF) (ECMWF, 2019). The datasets covered the period from 2000 to 2025 and were aggregated to calculate long-term mean annual precipitation, actual evapotranspiration, surface runoff, and soil moisture conditions within the basin.

Field investigations were carried out in 2025 within the representative Byzhy watershed. UAV-based aerial surveys and LiDAR scanning were conducted at selected MAR test sites to obtain detailed topographic information and support site characterization. In addition, infiltration tests were performed to determine infiltration capacity and field-saturated hydraulic conductivity (Kfs). Groundwater and surface water samples were collected for hydrochemical analysis. The chemical composition of the samples was evaluated using Piper and Durov diagrams to identify hydrochemical facies and assess groundwater recharge conditions. The field data were used to support and verify the GIS-based MAR suitability assessment.

The importance level of each factor was evaluated using the AHP method, and the final MAR suitability map was generated through weighted overlay analysis in the GIS environment using ArcGIS Pro 3.6.2 (Esri, Redlands, CA, United States). Furthermore, the Monte Carlo method was applied to assess the stability of the thematic weighting coefficients and the reliability of the results obtained. To further evaluate the uncertainty associated with the suitability assessment, fuzzy overlay approaches, including Fuzzy Gamma and Fuzzy Minimum operators, were also applied and compared with the AHP-based results.

2.2.2 Data processing

Currently, a universal and generally accepted set of criteria for identifying suitable areas for MAR implementation in semi-arid regions has not been fully developed. Nevertheless, morphometrical, hydrological, hydrogeological, geological, and environmental factors are commonly considered during MAR suitability assessment. In particular, comprehensive evaluation of hydrogeological and climatic indicators is important for effective MAR planning in water-limited semi-arid basins. The number of studies integrating such factors within GIS-based multi-criteria MAR suitability analysis remains relatively limited. Therefore, in this study, several thematic layers were jointly used to provide a more accurate assessment of the factors influencing MAR processes.

The integrated thematic layers were generated through preliminary processing within a unified geospatial environment. Spatial modeling, raster processing, and suitability assessment.

To ensure the accuracy of spatial analysis and the consistency of spatial datasets, all spatial data were projected into the WGS 1984 UTM Zone 43 N coordinate system (WGS-1984). All raster layers derived from ASTER DEM and remote sensing data were standardized to a common spatial resolution of 30 m to ensure consistency in spatial analysis.

The Geological factors: geological database used in this study was obtained from the Global Lithological Map (GLiM) dataset available through the Esri Living Atlas data portal. The dataset was processed within the GIS environment, clipped to the study area boundary, and converted into thematic lithological layers for integration into the AHP-based suitability analysis.

The Environmental factors: layers included soil and land use/land cover (LULC) datasets. Soil data obtained from the FAO were clipped to the study area boundary, converted into raster format, and reclassified according to original soil classes were preserved for AHP integration. The ESA WorldCover LULC dataset, originally provided at a spatial resolution of 10 m, was resampled to 30 m using the nearest-neighbor method to ensure consistency with the spatial resolution of the other thematic layers. The nearest-neighbor approach was selected because LULC represents categorical data and therefore requires preservation of the original class values.”

The hydrological-morphometric factors: the slope layer was derived from the ASTER DEM using the slope tool. Flow direction and flow accumulation analyses were performed after sink filling, and stream networks were extracted to generate the drainage density layer. The drainage network was then extracted using a flow accumulation threshold of 1,500 cells, corresponding to approximately 1.35 km2 at a spatial resolution of 30 m.

The Hydrological Factors: layer included precipitation, drainage density, and distance-to-river datasets derived from the WorldClim v2.1 database, while drainage density and distance-to-river layers were derived from DEM-based hydrological analysis. The datasets were processed and integrated into the hydrological assessment for the overall study area.

For the representative Byzhy River Basin, watershed delineation was performed using the Flow Direction and Flow Accumulation tools within the Spatial Analyst and Hydrology toolsets. Precipitation, surface runoff, and soil moisture data were obtained from the TerraClimate dataset, whereas actual evapotranspiration data were derived from the ERA5-Land dataset for the period 2000–2025. These datasets were used to characterize the basin’s hydrological conditions and to calculate the long-term mean potential water balance.

The hydrogeological factors: layers included groundwater depth and groundwater mineralization data obtained from the Atlas of Hydrogeological of Kazakhstan prepared by the Akhmedsafin Institute. The original datasets were processed within the GIS environment and transformed into raster thematic layers for subsequent spatial and overlay analysis.

Additional field datasets were collected in 2025 within representative Byzhy watershed to support MAR site assessment. These included UAV imagery, LiDAR data, infiltration tests for determining field-saturated hydraulic conductivity (Kfs), and hydrochemical analyses of groundwater and surface water samples. The collected information was used to characterize local recharge conditions and support the interpretation of GIS-based suitability assessment results.

Continuous 8 raster variables, (slope, soil, drainage density, precipitation, groundwater depth, groundwater mineralization, geology and LULC), were classified into five classes using the Natural Breaks (Jenks) classification method prior to integration into the AHP-based multicriteria analysis. The classification procedure was applied to minimize intra-class variance and maximize inter-class differentiation among thematic factors. The standardized thematic layers were subsequently prepared within the GIS environment to represent the spatial variability of factors influencing groundwater recharge and MAR suitability within the study area. The thematic layers were further integrated using an AHP-based multicriteria analysis to identify suitable zones for MAR implementation. To evaluate the uncertainty and robustness of the AHP-derived weighting scheme, Monte Carlo simulation was additionally applied through iterative variation of thematic factor weights within the multicriteria analysis framework. Figure 2 schematically illustrates the methodological framework used in this research.

Figure 2

2.2.3 Analysis of thematic layers

Slope: Slope is one of the important topographic factors influencing infiltration processes and groundwater recharge. The gradient slope directly affects surface runoff and water infiltration conditions within the watershed. Areas with steep slopes are generally characterized by higher runoff and lower infiltration conditions, whereas gentle slopes promote water accumulation and increase groundwater recharge. In this study, the slope map was generated from the ASTER DEM with a spatial resolution of 30 m using the Slope tool in the ArcGIS Pro 3.6.2 environment. The obtained slope values ranged from 0° to 75° (Figure 3a), reflecting the topographic variability of the study area and its influence on MAR suitability conditions.

Figure 3

Geology: Geologically (Figure 3b), the study area is composed of unconsolidated sediments, carbonate sedimentary rocks, siliciclastic sedimentary formations, volcanic rocks, and plutonic complexes distributed across both mountainous and plain terrains. Unconsolidated Quaternary alluvial and alluvial–proluvial deposits are widely developed within the plains, river valleys, and intermountain depressions, where they form favorable hydrogeological conditions for groundwater accumulation and infiltration processes. Carbonate and siliciclastic sedimentary rocks are mainly associated with foothill and basin structures, whereas volcanic and plutonic formations are predominantly concentrated within the mountainous regions of the Ile and Zhetysu Alatau (U.M. Akhmedsafin Institute of Hydrogeology and Environmental Geosciences, 2022). The geological heterogeneity of the region significantly influences groundwater distribution, recharge conditions, permeability characteristics, and the potential suitability of the area for Managed Aquifer Recharge (MAR) implementation.

Soil: Soil characteristics (Figure 3c) play an important role in infiltration processes, groundwater recharge, and MAR suitability assessment. In this study, the soil types were derived from the FAO/UNESCO Digital Soil Map of the World (DSMW) (FAO/UNESCO, 2003) dataset with a spatial resolution of 30 m. Five major soil types were identified within the study area: sandy loam, loam, clay loam, sandy clay, and silty clay. The spatial distribution of these soil groups reflects variations in infiltration capacity and water retention conditions across the area (Carsel and Parrish, 1988). In general, sandy loam and loam soils are considered more favorable for infiltration, whereas clay-rich soils may reduce groundwater recharge potential due to lower permeability. Therefore, the soil layer was incorporated as one of the key thematic factors in the GIS-based MAR suitability analysis. Soil-moisture conditions were additionally evaluated using the TerraClimate “soil” variable, expressed in millimeters (mm) of water stored in the total soil column at the end of each month. Monthly values for 2000–2025 were averaged to represent long-term mean soil-moisture conditions, with hir values indicating greater water storage within the soil profile.

Land use land cover: Land use and land cover (LULC) (Figure 3d) play an important role in controlling infiltration processes, surface runoff, groundwater recharge, and soil moisture distribution. Different land-cover types influence hydrological processes in different ways. Vegetated areas such as grasslands, croplands, and tree-covered regions generally enhance infiltration and reduce surface runoff, whereas built-up areas with impervious surfaces limit infiltration and increase runoff generation. Wetlands contribute to temporary water storage and support local groundwater recharge processes. In this study, the LULC map was derived from the ESA WorldCover dataset (Esri Living Atlas, 2021; Turner et al., 1995) with a spatial resolution of 10 m. Five major land-cover categories were identified within the study area, including grassland, cropland, tree cover, wetlands, and built-up areas. Cropland and grassland represent the dominant land-cover types across most parts of the study area, while tree cover and wetlands are mainly distributed in mountainous and lowland areas, respectively. Built-up areas occupy relatively limited zones and are primarily concentrated near settlements and infrastructure corridors. The spatial distribution of LULC classes significantly affects infiltration capacity, runoff formation, and potential groundwater recharge conditions across the study area.

Drainage density: Drainage density (DD) represents the total length of stream channels per unit basin area and is considered an important hydrological factor influencing surface runoff, infiltration processes, and groundwater recharge potential. The drainage system reflects areas where fluvial processes dominate over hill-slope processes, contributing to the formation and maintenance of channel networks. High drainage density regions are generally associated with low infiltration capacity and rapid runoff concentration, whereas low drainage density areas indicate relatively favorable conditions for groundwater recharge. In this study, the drainage density map was derived from the ASTER DEM dataset obtained from the NASA SRTM DEM database (2024) (NASA, United States) with a spatial resolution of 30 m. The drainage network was extracted in the ArcGIS Pro 3.6.2. environment using hydrological analysis tools based on flow direction and flow accumulation processes. Within the study area, drainage density values range from 0 to 0.78 km/km2 (Figure 3e). Spatial analysis indicates that relatively higher drainage density values are mainly distributed in mountainous and dissected terrain areas, while lower values dominate relatively flat regions. The spatial distribution of drainage density significantly affects runoff concentration, infiltration conditions, and potential groundwater recharge processes across the study area.

Precipitation: Precipitation is considered one of the main climatic factors influencing infiltration processes, surface runoff formation, soil moisture conditions, and groundwater recharge potential within the study area. The spatial distribution, intensity, and amount of precipitation directly affect hydrological response and recharge conditions, particularly in mountainous regions where increased precipitation contributes to higher infiltration and potential groundwater replenishment. In this study, long-term mean precipitation data from WorldClim v2.1 for the 1970–2000 climatological period were used as a consistent baseline for the AHP-based MAR suitability assessment. The original 30 arc-second (~1 km) dataset was resampled to 30 m for integration with the other thematic layers. Recent hydroclimatic variability was assessed separately in the water-balance analysis using data for 2000–2025. The resulting precipitation map demonstrates considerable spatial variability across the study area, with annual precipitation values ranging from approximately 136 to 770 mm/year (Figure 3f). Relatively higher precipitation values are mainly distributed in mountainous and elevated terrain areas, whereas lower precipitation conditions dominate flat and lowland regions. The spatial variability of precipitation significantly influences infiltration capacity, runoff generation, and potential groundwater recharge processes within the study area.

Groundwater depth to aquifer: Groundwater depth is an important hydrogeological factor influencing the suitability of aquifers for MAR implementation. In this study, the groundwater-depth layer was derived from the Atlas of Hydrogeological Maps of Kazakhstan. The source atlas does not provide a specific temporal attribution for this layer (e.g., seasonal conditions, long-term mean values, or measurements from a specific date or period). Therefore, the groundwater-depth layer was interpreted as a regional cartographic representation of groundwater-depth conditions, without assigning a specific temporal reference beyond the information provided in the source. The spatial distribution of groundwater depth was classified according to the depth intervals available in the source hydrogeological atlas: <5 m, 5–10 m, 10–15 m, 15–30 m, and >50 m. These classes reproduce the categories available in the original cartographic source. The resulting map indicates that shallower groundwater conditions are mainly represented in the northern and western parts of the study area, whereas greater groundwater depths predominantly occur in the southern part (Figure 3g).

Groundwater mineralization: Groundwater mineralization (Figure 3h) is an important hydrogeological factor influencing groundwater quality and the suitability of aquifers for MAR implementation. In this study, the groundwater mineralization layer was developed using data from the Atlas of Hydrogeological Maps of Kazakhstan. The spatial distribution of mineralization values was classified into five categories, including <1 g/L, 1–3 g/L, 3–5 g/L, >5 g/L, and >10 g/L. The obtained results showed that most of the study area is characterized by groundwater mineralization values below 1 g/L, whereas zones with higher mineralization are mainly distributed in the northern and northeastern parts of the basin. Lower groundwater mineralization was considered more favorable for MAR from a water-quality and geochemical compatibility perspective, as it may reduce the risk of salinity-related deterioration of groundwater quality during recharge. Therefore, groundwater mineralization was incorporated as one of the thematic factors in the GIS-based MAR suitability analysis.

2.2.4 Analytical hierarchy process (AHP)

The Analytical Hierarchy Process (AHP) was applied to identify suitable areas for Managed Aquifer Recharge (MAR) within the study area using a GIS-based multicriteria decision analysis (Nogueira, 2017).

The normalized thematic layers were integrated using the Analytic Hierarchy Process (AHP) to evaluate the suitability of potential Managed Aquifer Recharge (MAR) sites. The relative importance of each criterion was determined through pairwise comparison using Saaty’s nine-point scale (Saaty, 1980). Based on the comparison matrix, normalized weights were calculated for all thematic criteria, and the consistency of expert judgments was verified prior to the suitability analysis.

Following the weighting procedure, each thematic layer was reclassified into five suitability classes representing different levels of favorability for MAR implementation. Higher scores were assigned to conditions promoting groundwater recharge, whereas lower scores represented unfavorable hydrogeological settings. The standardized raster layers were subsequently combined using the Weighted Overlay approach to produce the final AHP-based MAR suitability map.

Eight thematic criteria were incorporated into the multi-criteria evaluation. Groundwater depth received the highest influence with a normalized weight of 0.33, followed by groundwater mineralization (0.23), slope (0.16), precipitation (0.11), soil (0.07), drainage density (0.05), geology (0.03), and land use/land cover (0.02) (Table 2) The assigned weights reflect the relative contribution of each factor to groundwater recharge potential under the hydrogeological conditions of the Almaty and Zhetysu regions.

Table 2

CriterionDTAMINSLOPEPRECSOILDDGEOLULCWeight
Depth to aquifer (DTA)122357990.33
Mineralization (MIN)1/212235890.23
Slope (SLOPE)1/21/21223580.16
Precipitation (PREC)1/31/21/2122460.11
Soil (SOIL)1/51/31/21/211240.07
Drainage density (DD)1/71/51/31/211230.05
Geology (GEO)1/91/81/51/41/21/2120.03
LULC1/91/91/81/61/41/31/210.02
Total weight1.00

Pairwise comparison matrix and criteria weights used in the AHP-based MAR suitability assessment.

Groundwater depth was included as an AHP criterion because it represents an important hydrogeological boundary condition for assessing the suitability of an area for MAR implementation. It was not used as an independent validation variable because the groundwater-depth dataset constituted one of the regional hydrogeological inputs used to construct the suitability model and therefore was not independent of the model itself. Independent validation requires observations that are not used in model construction, such as field-based infiltration measurements, groundwater-level responses to recharge, or observations from existing MAR systems.

The weighted overlay analysis generated a continuous suitability index, which was subsequently classified into five categories ranging from unsuitable to very suitable. The resulting AHP suitability map served as the baseline model for subsequent Monte Carlo simulation and fuzzy logic analyses.

The Analytical Hierarchy Process (AHP) was applied to determine the relative importance of the eight thematic criteria used for Managed Aquifer Recharge (MAR) suitability assessment (Nogueira, 2017). The criteria included groundwater depth, groundwater mineralization, slope, precipitation, soil, drainage density, geology, and land use/land cover (LULC).

The relative importance of the criteria was formalized using a pairwise comparison matrix based on Saaty’s nine-point scale. In the pairwise comparison matrix, a value of 1 indicates equal importance between two criteria, whereas values of 3, 5, 7, and 9 indicate moderate, strong, very strong, and extreme importance, respectively. Reciprocal values were assigned for inverse comparisons.

The pairwise comparison matrix can be expressed as Equation 1:

where represents the relative importance of criterion compared with criterion. The reciprocal relationship between criteria is expressed as Equation 2:

and all diagonal elements are equal to 1, as shown in Equation 3:

The normalized priority vector was calculated from the principal eigenvector of the pairwise comparison matrix using Equation 4:

where A is the pairwise comparison matrix, is the vector of criterion weights, and is the maximum eigenvalue of the matrix.

The calculated priority weights were 0.3244 for groundwater depth, 0.2285 for groundwater mineralization, 0.1586 for slope, 0.1124 for precipitation, 0.0674 for soil, 0.0555 for drainage density, 0.0320 for geology, and 0.0212 for LULC. For implementation in the GIS-based weighted-overlay model, these values were rounded to 0.33, 0.23, 0.16, 0.11, 0.07, 0.05, 0.03, and 0.02, respectively, while maintaining a total weight of 1.00.

The consistency of the pairwise comparison matrix was evaluated using the Consistency Index (), calculated using Equation 5:

where is the Consistency Index, is the maximum eigenvalue of the pairwise comparison matrix, and n is the number of criteria included in the analysis (n = 8 in this study).

The Consistency Ratio (CR) was subsequently calculated using Equation 6:

where CR is the Consistency Ratio and RI is the Random Index, which represents the average consistency index of randomly generated reciprocal matrices of the same order. For a matrix with eight criteria,  = 1.41.

For the developed pairwise comparison matrix, the maximum eigenvalue was  = 8.1373. The resulting and CR values were 0.0196 and 0.0139, respectively. Since the CR value was below the commonly accepted threshold of 0.10, the pairwise comparison matrix was considered to have an acceptable level of consistency.

Following the consistency assessment, the standardized thematic layers were integrated using a weighted linear combination to calculate the final MAR suitability index using Equation 7:

where SI is the MAR suitability index, is the normalized weight of criterion is the standardized suitability score of criterion and is the total number of criteria, as shown in Equation 8.

where DTA is groundwater depth to aquifer, MIN is groundwater mineralization, SLOPE is terrain slope, PREC is precipitation, SOIL is soil suitability, DD is drainage density, GEO is geology, and LULC is land use/land cover.

The resulting suitability index was subsequently classified into five classes ranging from unsuitable to very suitable. The AHP-based suitability map was used as the baseline model for the subsequent Monte Carlo simulation and Fuzzy Gamma analysis.

2.2.5 Monte Carlo simulation and fuzzy logic analysis

To ensure consistency between the ArcGIS- and Google Earth Engine (GEE)-based analyses, the same eight thematic layers (groundwater depth, groundwater mineralization, slope, precipitation, soil, drainage density, geology, and land use/land cover) were used in both environments. All thematic rasters retained the same five-class suitability classification (1–5), spatial extent, raster alignment, coordinate reference system, and spatial resolution of 30 m. The baseline AHP model implemented in GEE used the same criterion weights as the ArcGIS weighted-overlay analysis: 0.33 for groundwater depth, 0.23 for groundwater mineralization, 0.16 for slope, 0.11 for precipitation, 0.07 for soil, 0.05 for drainage density, 0.03 for geology, and 0.02 for land use/land cover. Thus, the baseline AHP model implemented in GEE was spatially and parametrically consistent with the ArcGIS-based AHP model.

To evaluate the robustness and uncertainty of the regional MAR suitability assessment, Monte Carlo simulation was applied following the generation of the baseline AHP suitability model (Kazem et al., 2026). The simulation was designed to assess the sensitivity of the suitability model to uncertainties associated with thematic criterion weights and raster values.

The Monte Carlo framework was implemented using 200 simulation iterations. During each iteration, the AHP-derived criterion weights were randomly perturbed within ±10% of their original values, while raster values were simultaneously varied within ±3% to account for possible classification and measurement uncertainties. A new MAR suitability map was generated for each simulation using the weighted linear combination approach.

The ensemble of simulation outputs was subsequently used to calculate the mean suitability index MC_mean –0.64, standard deviation −0.07, coefficient of variation MC_cv –0.11, and the probability of suitability exceeding the threshold value of 0.70 (MC_prob ≥ 0.70). These statistical indicators were used to evaluate the spatial variability, uncertainty, and stability of the regional MAR suitability model.

In addition to the AHP-based weighted overlay, Fuzzy Gamma analysis was performed as an independent nonlinear spatial integration approach. The fuzzy procedure applied in this study represents a Fuzzy Gamma overlay rather than Fuzzy AHP, because fuzzy logic was not used to derive the AHP criterion weights. Instead, the standardized thematic layers were integrated using the Fuzzy Gamma operator to evaluate the robustness of the spatial suitability patterns. The same thematic input layers, spatial extent, raster alignment, coordinate reference system, and 30 m spatial resolution were retained for the Fuzzy Gamma analysis; however, the AHP-derived criterion weights were not applied because the thematic layers were integrated through their fuzzy membership values using the Fuzzy Gamma operator.

Prior to the fuzzy overlay, the five ordinal suitability classes of each thematic criterion were transformed into fuzzy membership values on a common scale ranging from 0.10 to 1.00. Suitability classes 1, 2, 3, 4, and 5 were assigned membership values of 0.10, 0.325, 0.55, 0.775, and 1.00, respectively, with increasing membership representing increasingly favorable conditions for MAR. The transformation preserved the ordinal ranking and equal spacing among the five suitability classes. A non-zero minimum membership value (0.10) was retained because the Fuzzy Product operator is multiplicative; assigning a membership value of zero to the lowest suitability class would force the product, and consequently the Fuzzy Gamma result, to zero wherever any single criterion belonged to that class.

The Fuzzy Product operator was calculated as using Equation 9:

where is the resulting Fuzzy Product membership value, is the fuzzy membership value of criterion i, and n is the total number of criteria.

The Fuzzy Sum operator was calculated as using Equation 10:

where is the resulting Fuzzy Sum membership value and is the fuzzy membership value of criterion i.

The Fuzzy Product and Fuzzy Sum were subsequently combined using the Fuzzy Gamma operator, as shown in Equation 11:

where is the resulting Fuzzy Gamma membership value and γ is the gamma parameter controlling the balance between the Fuzzy Sum and Fuzzy Product operators. The gamma parameter ranges from 0 to 1; values approaching 1 increase the influence of the Fuzzy Sum, whereas values approaching 0 increase the influence of the Fuzzy Product. In this study, γ = 0.88 was applied.

Finally, a difference map between the AHP-based and Fuzzy Gamma suitability models was generated to evaluate spatial differences in suitability estimates obtained using the two integration approaches. The outputs of the Monte Carlo simulation and Fuzzy Gamma analysis were subsequently used to assess the uncertainty and robustness of the regional MAR suitability assessment.

It should be emphasized that the Monte Carlo simulation and the comparison between the AHP and Fuzzy Gamma models were not treated as independent validation procedures. Monte Carlo simulation was used to quantify parameter sensitivity and spatial uncertainty, whereas comparison with the Fuzzy Gamma model was used to evaluate methodological sensitivity to the choice of multicriteria aggregation approach. Independent field verification of the regional suitability assessment was conducted separately at the representative Byzhy River Basin using site-scale observations and measurements not used to construct the regional AHP suitability model, particularly field infiltration tests and field-saturated hydraulic conductivity measurements, together with detailed hydrogeological and hydrochemical investigations.

3 Results

3.1 AHP-based MAR suitability

The AHP-based MAR suitability map generated through the weighted overlay analysis is presented in Figure 4. The final suitability map was produced by integrating eight thematic criteria, including groundwater depth, groundwater mineralization, slope, precipitation, soil, drainage density, geology, and land use/land cover, using the normalized weights derived from the Analytic Hierarchy Process.

Figure 4

Based on the resulting suitability index, the study area was classified into five categories: unsuitable, low suitable, moderately suitable, suitable, and very suitable. The spatial distribution of these classes reflects the heterogeneity of hydrogeological, geological, and topographic conditions across the Almaty and Zhetysu regions.

The very suitable class is mainly associated with areas characterized by favorable groundwater depth, low groundwater mineralization, gentle terrain, permeable geological formations, and suitable drainage conditions, indicating high potential for managed aquifer recharge. The suitable class occupies areas where most hydrogeological conditions remain favorable, although one or more controlling factors become moderately limiting. Areas classified as moderately suitable represent transitional zones with intermediate recharge potential and may require additional site-specific investigations before MAR implementation.

In contrast, the low suitable and unsuitable classes are primarily related to areas with unfavorable hydrogeological conditions, including relatively deep groundwater levels, elevated groundwater mineralization, steep slopes, less permeable geological formations, or unfavorable land-cover characteristics, all of which reduce the potential for groundwater recharge.

The proportional distribution of the five suitability classes was estimated from the classified AHP map. Approximately 6.8% of the study area was classified as unsuitable, 3.24% as low suitable, 34.88% as moderately suitable, 33.26% as suitable, and 21.82% as very suitable for MAR implementation. These values will be replaced with the final area statistics calculated from the classified raster.

Overall, the AHP-based suitability map provides the primary spatial framework for identifying potential MAR sites within the study area and served as the baseline dataset for the subsequent Monte Carlo simulation and fuzzy logic analyses.

3.2 Uncertainty and fuzzy logic analysis

To evaluate the robustness and reliability of the MAR suitability assessment, an uncertainty analysis based on Monte Carlo simulation and fuzzy logic integration was additionally performed. The procedure was implemented in the Google Earth Engine (GEE) environment using eight normalized thematic layers, including depth to aquifer, mineralization, slope, precipitation, soil characteristics, drainage density, geology, and land use/land cover.

Initially, all thematic layers were normalized to a common suitability scale ranging from 0 to 1, where higher values represented more favorable conditions for managed aquifer recharge. The baseline AHP suitability model was calculated using weighted linear combination according to the final AHP-derived weights. The assigned weights were 0.33 for depth to aquifer, 0.23 for mineralization, 0.16 for slope, 0.11 for precipitation, 0.07 for soil, 0.05 for drainage density, 0.03 for geology, and 0.02 for land use/land cover.

To quantify model sensitivity and spatial uncertainty, a Monte Carlo framework (Figure 5) was applied. Random perturbations were introduced both to the AHP weights and to the raster layers themselves. The weights were varied within ±10% of their original values, while raster values were additionally perturbed by ±3% to account for possible classification and measurement uncertainties. A total of 200 simulation iterations were generated for the full analysis.

Figure 5

For each iteration, a new suitability map was calculated, producing an ensemble of realizations from which several statistical indicators were derived, including the mean suitability index (MC_mean = 0.64), standard deviation (MC_std = 0.07), coefficient of variation (MC_cv = 0.11), and probability of suitability exceeding a threshold value of 0.70 (MC_prob ≥ 0.7). The coefficient of variation was used as an indicator of model stability, where low values indicated spatially robust suitability estimates and high values reflected increased uncertainty.

In parallel, fuzzy logic analysis was performed using the Fuzzy Gamma operator as an independent non-linear multicriteria integration approach for comparison with the conventional AHP-based weighted overlay model (Supplementary Figure S1). For the fuzzy analysis, the five ordinal suitability classes were transformed into fuzzy membership values of 0.10, 0.325, 0.55, 0.775, and 1.00 for classes 1–5, respectively. The transformation preserved the ordinal ranking and equal spacing between suitability classes while maintaining a non-zero minimum membership value for the multiplicative Fuzzy Product operator. The Fuzzy Product and Fuzzy Sum were calculated independently from the transformed thematic layers and subsequently combined using the Fuzzy Gamma operator with γ = 0.88 (μγ = μsum0.88 × μproduct0.12).

Comparison of the AHP-based and Fuzzy Gamma suitability models showed a moderate positive spatial correlation (Pearson’s r = 0.598; R2 = 0.358). The mean suitability index was 0.640 for the AHP model and 0.600 for the Fuzzy Gamma model. The mean difference between the Fuzzy Gamma and AHP estimates was −0.041, indicating that the fuzzy approach generally produced slightly lower suitability estimates. The mean absolute error (MAE) between the two model outputs was 0.079, while the root mean square error (RMSE) was 0.101.

The comparison indicates an overall spatial correspondence between the two multicriteria integration approaches, while the observed local differences demonstrate the sensitivity of suitability estimates to the aggregation method. In combination with the Monte Carlo analysis, these results provide complementary information on both parameter uncertainty and methodological uncertainty in the regional MAR suitability assessment.

3.3 Site-scale field verification in the Byzhy river basin

Based on the regional AHP suitability assessment, the Byzhy River Basin was selected as a representative high-suitability sub-area for independent site-scale field verification. Importantly, the field infiltration measurements used at this stage were not incorporated into the construction or weighting of the regional AHP suitability model. Therefore, they provide an independent hydrogeological test of whether areas identified as favorable by the regional GIS-based screening exhibit suitable infiltration characteristics under field conditions. The field verification was complemented by detailed hydrogeological characterization, hydrochemical observations, UAV-LiDAR terrain analysis, water-balance assessment, and evaluation of available river discharge and aquifer storage capacity.

3.3.1 Hydrological assessment of potential water balance for groundwater recharge

The availability of sufficient water resources is one of the fundamental prerequisites for the implementation of Managed Aquifer Recharge (MAR). Following the regional MAR suitability assessment, a hydrological evaluation of the representative Byzhy River Basin was conducted to assess the potential water balance and determine whether the basin possesses sufficient water resources to support MAR implementation (see Figure 6). The potential water balance was estimated to quantify the amount of water potentially available for groundwater recharge.

Figure 6

Hydrological conditions were evaluated using long-term mean precipitation, actual evapotranspiration, surface runoff, and soil moisture data for the period 2000–2025. Precipitation, surface runoff, and soil moisture data were obtained from the TerraClimate dataset (Liu et al., 2020), whereas actual evapotranspiration (ET) data were derived from the ERA5-Land dataset. Since ERA5-Land provides evapotranspiration values in meters, all ET values were converted to millimeters to ensure consistency among the hydrological variables. Long-term mean values were calculated by aggregating annual raster datasets using the Aggregate Multidimensional Raster tool with the Mean statistical function in ArcGIS Pro 3.6.2. To estimate the amount of water potentially available for groundwater recharge, the potential water balance (PWB) was calculated using the following Equation 12:

where PWB is the net potential water balance (mm), P is precipitation (mm), ET is actual evapotranspiration (mm), and R is surface runoff (mm). Soil moisture was evaluated separately as an additional hydrological indicator reflecting the capacity of the soil profile to retain water and support infiltration processes, but it was not included directly in the potential water balance calculation.

The long-term mean hydrological variables were 34.84 mm for precipitation, 1.36 mm for actual evapotranspiration, 5.79 mm for surface runoff, and 16.93 mm for soil moisture. Based on the calculated long-term mean values, the potential water balance was estimated at 27.69 mm/year.

The spatial distribution of the potential water balance is presented in Figure 7. Higher values indicate areas with greater potential water availability for groundwater recharge, whereas lower values indicate relatively limited potential water availability.

Figure 7

The spatial distribution of the potential water balance is presented in Figure 8. Higher values indicate areas where a greater proportion of precipitation remains after accounting for losses due to actual evapotranspiration and surface runoff, reflecting more favorable hydrological conditions for potential groundwater recharge. Lower values represent areas with relatively lower potential water availability resulting from greater losses through evapotranspiration or increased surface runoff.

Figure 8

The long-term variability of the main hydrological components is presented in Figure 8. It should be noted that the figure shows the annual values of precipitation, actual evapotranspiration, surface runoff, and soil moisture for the period 2000–2025 rather than their aggregated long-term mean values. The time-series analysis indicates pronounced interannual variability in precipitation and surface runoff, whereas actual evapotranspiration and soil moisture exhibit relatively smaller fluctuations.

Thus, the hydrological assessment of the Byzhy River Basin indicates favorable conditions for potential groundwater recharge. The positive long-term mean potential water balance of 27.69 mm/year indicates that, after accounting for losses due to actual evapotranspiration and surface runoff, a certain amount of water remains that is potentially available for infiltration and groundwater recharge.

However, the potential water balance should not be interpreted as the actual rate of natural groundwater recharge. Rather, it represents a hydrological indicator of the area’s potential water availability. The final assessment of MAR feasibility also depends on the availability of river water, the infiltration capacity of the geological materials, the hydraulic properties of the aquifer, groundwater quality, and aquifer storage capacity, which are evaluated in the following sections.

LiDAR terrain analysis: A UAV-based LiDAR survey was carried out at the representative MAR site using a DJI Matrice 350 RTK with P-RTK 3 (Supplementary Figure S2). The acquired point cloud was processed to generate a Digital Elevation Model (DEM) with a spatial resolution of 5 cm. The high-resolution DEM enabled detailed characterization of the site microtopography (Supplementary Figures S3, S4) and formed the basis for the subsequent analysis of surface flow direction.

The surface flow direction map (Figure 9) was generated from a Digital Elevation Model (DEM) derived from high-resolution LiDAR survey data acquired over the study area. Data processing was performed using standard hydrological analysis tools.

Figure 9

In the first stage, the DEM was processed using the Fill tool to remove local depressions that could affect the accuracy of surface flow modeling. Subsequently, the Flow Direction tool based on the D8 algorithm was applied. This algorithm assigns one of eight possible flow directions (1, 2, 4, 8, 16, 32, 64, and 128) to each DEM cell according to the direction of the steepest downslope gradient.

The resulting flow direction raster was converted into a vector layer of arrows to provide a clearer visualization of surface water movement. To improve the visual representation of the map, the density of the arrows was optimized using a resampling method. The flow direction layer was then overlaid on the DEM. The color scale represents ground elevation, while the blue arrows indicate the predominant direction of surface runoff.

Analysis of the resulting map showed that surface runoff across the study area is controlled by the natural terrain and is generally directed toward the valleys of the Byzhy and Mukanshy river. Divergent flow patterns are observed along watershed divides, whereas the predominant surface runoff direction is toward the northeastern part of the study area. These patterns indicate the natural redistribution of surface water within the study area.

The site selected for Managed Aquifer Recharge (MAR) is in northeastern part of the study area, where surface runoff from the surrounding terrain converges toward the proposed recharge site. This topographic setting is favorable for MAR implementation because the natural land slope facilitates the conveyance of river water to the infiltration basin. During periods of high river discharge, these conditions promote infiltration and subsequent recharge of the underlying aquifer. Therefore, the identified surface flow characteristics confirm that the study area possesses favorable hydrogeomorphological conditions for MAR implementation.

3.3.2 Hydrogeological assessment: infiltration tests

To evaluate the infiltration characteristics of the surficial deposits for Managed Aquifer Recharge (MAR), three field infiltration tests were conducted using the double-ring infiltrometer method (Li et al., 2019, ASTM International, 2018). During each test, a constant water level was maintained in both the inner and outer rings, while the volume of water required to sustain the water level in the inner ring was recorded over time. The outer ring minimized lateral water movement, allowing predominantly vertical infiltration beneath the inner ring.

At the beginning of each test, relatively high infiltration rates were observed due to the initially unsaturated condition of the soil. As the pore spaces became progressively saturated, infiltration rates gradually declined and approached a quasi-steady-state condition. Under these conditions, the infiltration process became nearly constant, indicating that the infiltration capacity was primarily controlled by the hydraulic properties of the underlying deposits rather than by the initial soil moisture conditions.

The steady-state infiltration rate and the corresponding field-saturated hydraulic conductivity (Kfs) were calculated for each test location. The obtained hydraulic conductivity values ranged from 1.86 to 1.96 m·day−1, with a mean value of 1.92 m·day−1, indicating that the Quaternary alluvial deposits possess relatively high permeability and are favorable for infiltration-based MAR.

At Site TN-1 (Figure 10a), the steady-state infiltration rate corresponded to a field-saturated hydraulic conductivity of 1.96 m·day−1, indicating excellent infiltration capacity of the near-surface deposits.

Figure 10

At Site TN-2 (Figure 10b), the calculated hydraulic conductivity was 1.86 m·day−1. Although slightly lower than the values obtained at the other sites, the deposits still exhibited favorable infiltration characteristics suitable for MAR implementation.

At Site TN-3 (Figure 10c), the field-saturated hydraulic conductivity reached 1.93 m·day−1, confirming consistently high permeability across the investigated area.

Overall, the three infiltration tests produced highly consistent results. The relatively small variation in hydraulic conductivity (1.86–1.96 m·day−1) indicates that the infiltration properties of the surficial alluvial deposits are spatially uniform within the representative study area. These results demonstrate that the investigated deposits provide favorable conditions for infiltration-based Managed Aquifer Recharge (García-Menéndez et al., 2021) and can serve as reliable input parameters for subsequent hydrogeological analyses and MAR system design.

These field measurements provide an independent site-scale verification of the regional suitability assessment because infiltration capacity and field-saturated hydraulic conductivity were not used as input criteria in constructing the regional AHP model. The consistently high Kfs values measured at the selected MAR site (1.86–1.96 m·day−1; mean 1.92 m·day−1) therefore provide independent field evidence that the area identified as favorable by the regional GIS-based screening possesses suitable hydraulic properties for infiltration-based MAR. This field verification should be distinguished from the Monte Carlo sensitivity analysis and the AHP–Fuzzy Gamma model comparison, which evaluate model robustness and methodological uncertainty rather than empirical field performance.

In addition to the hydraulic characterization of the representative MAR site, field investigations included groundwater sampling to evaluate regional hydrochemical conditions and groundwater quality within the proposed recharge area. A total of 53 groundwater samples collected from observation wells, boreholes, and springs across the Almaty and Zhetysu regions were analyzed to characterize the regional hydrochemical setting, while additional groundwater samples collected within the Byzhy River Basin were used to evaluate local groundwater quality.

The regional hydrochemical composition was interpreted using Piper and Durov diagrams (Figures 11 and 12). Groundwater samples exhibited neutral to slightly alkaline conditions, with pH values ranging from 7.0 to 8.3. Total dissolved solids (TDS) varied considerably, from approximately 140 to 9,422 mg·L−1, indicating substantial spatial variability in groundwater mineralization.

Figure 11

Figure 12

The Piper diagram (Figure 11) shows that groundwater is predominantly characterized by Ca–Mg cations and HCO3–SO4 anions. Most samples belong to bicarbonate-rich and mixed bicarbonate–sulfate hydrochemical facies, whereas several samples exhibit increased chloride concentrations, reflecting differences in lithology, groundwater evolution, and water–rock interaction processes across the study area.

The Durov diagram (Figure 12) further confirms the predominance of Ca–Mg–HCO3–SO4 groundwater types while illustrating the wide variability of groundwater mineralization. Groundwater samples collected within the Almaty Region exhibit greater variability in TDS and major-ion composition than those from the Zhetysu Region, suggesting more heterogeneous hydrogeochemical conditions.

To evaluate groundwater quality, a spatial distribution map of groundwater mineralization was developed using hydrochemical data from 47 groundwater sampling points (Supplementary Table S1) located within the study area (Figure 13). All 47 sampling points were used for the spatial interpolation of groundwater mineralization, allowing its regional variability across the study area to be characterized. Of the total number of sampling points, 7 (Supplementary Table S1, Nos. 25, 28, 29, 30, 45, 46, 47) are located directly within the proposed MAR site and were used to characterize the local conditions of the selected area. The representative Byzhy River Basin, selected for detailed assessment of MAR suitability, is delineated in Figure 13 and shown at an enlarged scale to provide a clearer representation of the location of the MAR site relative to the groundwater sampling network and the main hydrogeological features. Groundwater mineralization ranges from 0.146 to 5.498 g·L−1 and was classified into five intervals: 0.146–0.860, 0.861–1.615, 1.616–2.392, 2.393–3.567, and 3.568–5.498 g·L−1. Fresh and slightly mineralized groundwater (0.146–1.615 g·L−1) predominates within the Byzhy River Basin, whereas more mineralized groundwater occurs mainly in the western part of the study area, outside the basin boundary.

Figure 13

The proposed MAR site, located in the northern and northeastern part of the basin, coincides with the zone of fresh groundwater where mineralization remains below 1 g·L−1. The combination of low groundwater mineralization, favorable infiltration characteristics, and suitable hydrogeological conditions indicates that the selected area is well suited for infiltration-based Managed Aquifer Recharge. Furthermore, the predominance of fresh groundwater reduces the potential risk of groundwater salinization during long-term recharge operations, which is particularly important under the semi-arid conditions of southeastern Kazakhstan.

Together with the infiltration tests and hydrogeological investigations, these results provide further evidence supporting the suitability of the selected site for long-term MAR implementation.

Hydrogeological conditions: The hydrogeological conditions of the study area are characterized based on a 1:200,000 hydrogeological map and the hydrogeological cross-section (Figure 14). The selected MAR site is located within the Upper Quaternary and Recent alluvial aquifer (aQIII-IV), which represents the principal aquifer in the Byzhy River Basin. The aquifer consists of boulders, boulder–pebble and gravel deposits, as well as sands with interbeds of sandy loam and loam. The aquifer thickness is approximately 50 m, providing substantial storage capacity for artificially recharged groundwater.

Figure 14

The hydrogeological cross-section shows that the Quaternary alluvial deposits are underlain by less permeable Pliocene formations, which limit the downward movement of infiltrated water. This geological setting promotes groundwater storage within the alluvial aquifer (Ministry of Geology of the USSR, 1977). The combination of the high permeability of the alluvial deposits and the considerable aquifer thickness indicates favorable hydrogeological conditions for the implementation of Managed Aquifer Recharge (MAR).

The results of the present study provide the basis for a conceptual Managed Aquifer Recharge (MAR) scheme for the Byzhy River Basin Under the semi-arid climatic conditions of southeastern Kazakhstan, surface water availability is characterized by pronounced seasonal variability, with relatively high river discharge during snowmelt and spring floods, followed by prolonged periods of low flow and water deficit. During periods of excess surface water availability, part of the river discharge may be diverted to infiltration facilities to replenish groundwater storage (Supplementary Figure S5). The recharged groundwater can subsequently serve as a strategic water reserve and be abstracted during dry years or drought periods to satisfy domestic, agricultural, and industrial water demands. Such cyclic storage and recovery improves water-use efficiency, reduces dependence on highly variable surface water resources, and increases the long-term resilience of regional water supply systems under semi-arid climatic conditions (Murray et al., 2007).

3.3.3 Quantification of water availability for managed aquifer recharge

To estimate the volume of river water available for Managed Aquifer Recharge (MAR), the long-term mean discharge of the Byzhy River was first determined. The calculation was based on daily river discharge records for the period 2005–2025, obtained from the Karymsak and Aynabulak hydrological gaging stations operated by Kazhydromet (2025).

The Karymsak gaging station is located in upper reach of the Byzhy River, whereas the Aynabulak station is situated in the middle reach of the basin. Since the proposed MAR site is located in downstream part of the basin, within its northern and northeastern sectors where no hydrological observations are available, the long-term mean river discharge was estimated using the average discharge recorded at these two upstream gaging stations. The calculated long-term mean discharge of the Byzhy River was 5.53 m3/s.

Currently, there are no internationally accepted standards defining a fixed proportion of river discharge that can be allocated to Managed Aquifer Recharge. A review of published MAR studies indicates that, for preliminary feasibility assessments, 10–30% of the available river flow is commonly considered, depending on local hydrological conditions, environmental flow requirements, and project objectives (Dillon et al., 2022; Sufyan et al., 2024; Ferrante and Gonwa, 2024; Ferrante, 2023; Dillon et al., 2009). Considering that the study area is located within a semi-arid climatic zone, is characterized by limited water resources and a relatively small catchment area, and taking into account the need to preserve the natural hydrological regime of the river, a conservative diversion coefficient of 10% of the long-term mean river discharge was adopted for the present study.

The discharge available for artificial recharge was calculated using the following Equation 13:

where QMAR is the discharge allocated for Managed Aquifer Recharge (m3/s), Qriver and is the long-term mean river discharge (m3/s).

The annual recharge volume was calculated as Equation 14:

where is the theoretical annual volume of river water potentially available for Managed Aquifer Recharge (m3/year); is the adopted fraction of the long-term mean river discharge potentially available for MAR (m3/s); and is the number of seconds in 1 year (31,536,000 s). The annualization of provides a volumetric estimate of the potentially available water resource and enables its comparison with the estimated groundwater storage capacity of the aquifer.

Based on this approach, the theoretical annual volume potentially available for MAR was estimated at 17.44 million m3/year, corresponding to 10% of the long-term mean river discharge. This volume was subsequently used to assess whether the identified aquifer has sufficient storage capacity to accommodate the estimated recharge potential.

To evaluate whether the aquifer is capable of accommodating the calculated recharge volume, the minimum aquifer area required for groundwater storage was estimated using the following Equation 15:

where S is the minimum aquifer area required for groundwater storage (m2); (VMAR) is the annual recharge volume (m3); is the aquifer thickness (m); and is the effective porosity.

An aquifer thickness of 50 m was determined from the hydrogeological cross-section of the study area. The specific yield (0.30) (Jabassov et al., 2026) was adopted from published data for Quaternary alluvial–proluvial aquifers with hydrogeological characteristics comparable to those of the study area.

The minimum aquifer area required to accommodate the estimated annual recharge volume was calculated to be 1.16 km2.

The prospective MAR site, located within the Quaternary aquifer in the northeastern part of the Byzhy River Basin (Figure 14), covers approximately 50 km2, which is more than 43 times larger than the minimum area required for groundwater recharge. This indicates that the available aquifer has sufficient storage capacity to accommodate the proposed recharge volume.

The potential aquifer storage capacity was estimated using the following Equation 16:

where is the potential aquifer storage capacity (m3); is the infiltration area (assumed equal to the water surface area) (km2); is the aquifer thickness (m); and is the effective porosity (Jabassov et al., 2026).

The Quaternary aquifer in the northeastern part of the Byzhy River Basin covers an area of approximately 50 km2 (50,000,000 m2). The aquifer thickness of 50 m was determined from the hydrogeological cross-section, whereas the specific yield of 0.30 was adopted from published literature. Based on these parameters, the potential aquifer storage capacity was estimated at 750 million m3.

To evaluate the natural groundwater regime of the aquifer, the groundwater discharge was estimated using a groundwater discharge modulus of 2 L·s−1·km−2, according to the Hydrogeological Atlas of Kazakhstan published. The groundwater discharge was calculated using the following Equation 17:

where is the groundwater discharge m3/s−1; is the groundwater discharge modulus (L·s−1·km−2); and is the infiltration area (km2).

The annual groundwater discharge volume was calculated using Equation 18:

where is the annual groundwater discharge volume (m3/year); is the groundwater discharge (m3/s); and t is the time period (s). The calculated annual groundwater discharge volume was 3.15 million m3/year.

The effective aquifer storage capacity was determined as the difference between the potential aquifer storage capacity and the annual groundwater discharge volume using Equation 19:

where is the effective aquifer storage capacity (m3); is the potential aquifer storage capacity (m3); and is the annual groundwater discharge volume (m3). The effective aquifer storage capacity was estimated at 746.85 million m3.

The proportion of annual Managed Aquifer Recharge relative to the effective aquifer storage capacity was calculated using the following Equation 20:

where is the proportion of the annual Managed Aquifer Recharge relative to the effective aquifer storage capacity (%); is the annual recharge volume (m3); and is the effective aquifer storage capacity (m3). The calculated recharge volume represents 2.33% of the effective aquifer storage capacity.

This result indicates that the proposed annual recharge volume occupies only a small fraction of the available aquifer storage capacity, demonstrating a substantial storage reserve. Therefore, the Quaternary aquifer in the northeastern part of the Byzhy River Basin is capable of accommodating the proposed recharge volume without significantly altering its natural groundwater regime, confirming its high suitability for Managed Aquifer Recharge (MAR).

4 Discussion

The obtained results are consistent with recent MAR studies. Martins et al. (2024) identified 22% of southern Portugal as having high MAR feasibility, while Chekirbane et al. (2025) reported about 19% of the Chiba watershed as highly to very highly feasible. In our study, 21.82% of the area was classified as Very Suitable for MAR, which is comparable with these results. However, unlike many regional-scale MAR assessments, our approach additionally provides site-scale validation of the selected Byzhy area using LiDAR, hydrochemical data, infiltration tests, water-balance analysis, and river-discharge data. This makes the assessment more practical because it confirms not only where MAR can be implemented, but also whether the selected site has suitable infiltration conditions and sufficient water available for recharge.

Despite these encouraging findings, several limitations should be acknowledged. Although field investigations in the Byzhy River Basin support the suitability of the selected MAR site, they were conducted within a representative area and therefore do not provide independent validation of the entire regional suitability map. The field investigations were conducted within the temporal and financial framework of a single research project and therefore provide only a limited representation of the hydrogeological system. Although historical river discharge records and field observations were available, continuous monitoring of groundwater dynamics within the representative MAR site was not available. Consequently, the hydraulic interaction between the Byzhy River and the underlying alluvial aquifer could not be quantified directly under different seasonal and hydrological conditions (Sallwey et al., 2019; Onglassynov et al., 2025). In addition, the aquifer storage assessment presented in this study is based on a static volumetric approach and therefore represents potential storage capacity rather than the transient hydraulic response of the aquifer to artificial recharge (Robertson et al., 2023).

Furthermore, the field infiltration tests performed in this study were intended to evaluate the infiltration characteristics of the surficial deposits rather than to design a full-scale recharge system. Although the double-ring infiltration (Li et al., 2019) tests demonstrated consistently high infiltration capacity, the design of recharge wells requires additional hydrogeological information that was not available within the present investigation (Guttman et al., 2017; Pettenati et al., 2012). Future engineering design should therefore include drilling dedicated test wells and conducting pumping and injection tests to determine aquifer transmissivity, storage properties, allowable injection pressures, well spacing, and long-term clogging (skin effect) (Jabassov et al., 2026; Casanova et al., 2016) under operational conditions. These investigations are essential for optimizing the configuration and performance of a full-scale MAR system and for parameterizing a reliable transient groundwater-flow model.

A more comprehensive understanding of river–aquifer interaction (Guo et al., 2023) would require the establishment of a dedicated hydrological monitoring station within the proposed MAR site together with a network of observation wells installed between the river and the recharge area (Maréchal et al., 2018). Continuous monitoring of groundwater levels, temperature, electrical conductivity, and river stage would enable direct quantification of recharge dynamics, river–aquifer exchange processes, groundwater travel times, and seasonal variations in infiltration. These observations would also provide the calibration and validation data required for transient numerical groundwater-flow modeling (Ringleb et al., 2016), such as MODFLOW, capable of simulating groundwater-level rise and decline, recharge-mound development and propagation, groundwater-flow redistribution, and aquifer recovery during and after managed recharge under different hydrological and operational scenarios.

In addition to hydrogeological investigations, future studies should include pilot-scale MAR implementation and techno-economic assessment (Zheng et al., 2021; Maréchal et al., 2020). Such analyses should compare the construction, operation, and maintenance costs of recharge facilities with the expected benefits of seasonal groundwater storage and recovery (Dillon et al., 2010), thereby providing the basis for selecting the most efficient MAR configuration under the semi-arid conditions of southeastern Kazakhstan.

The present study represents the first stage of MAR planning by identifying a hydrogeologically suitable recharge area and evaluating its recharge potential. The obtained results support a conceptual MAR strategy in which excess river water generated during spring snowmelt and high-flow periods is diverted to infiltration facilities and stored within the Quaternary alluvial aquifer (Jabassov et al., 2026; Gale, 2005; Murray et al., 2007). During subsequent dry years or periods of increased water demand, the stored groundwater may be recovered to supply municipal, agricultural, and industrial water users. Such seasonal storage and recovery can substantially improve water-use efficiency, reduce dependence on highly variable surface-water resources, and strengthen the resilience of regional water management under (Sallwey et al., 2019) the semi-arid climatic conditions of southeastern Kazakhstan.

Overall, this study demonstrates that reliable MAR planning requires considerably more than regional suitability mapping alone. Integrating GIS-based regional screening with hydrological assessment, hydrochemical characterization, LiDAR-based terrain analysis, field infiltration testing, hydrogeological investigations, and aquifer storage assessment substantially reduces uncertainty and provides a more robust basis for identifying technically feasible MAR sites. The proposed multi-scale framework therefore offers a transferable methodology for supporting sustainable groundwater management in Kazakhstan and other semi-arid regions facing increasing water scarcity and climate variability.

5 Conclusion

This study developed and applied a multi-scale framework for Managed Aquifer Recharge (MAR) assessment that links regional GIS-based suitability screening with independent site-scale field verification. The main contribution of the proposed approach is the transition from identifying potentially suitable areas through spatial multicriteria analysis to evaluating whether a selected site is supported by actual hydrogeological, hydrological, hydrochemical, and field evidence. This addresses an important limitation of conventional regional GIS-based MAR assessments, in which mapped suitability is often not verified under field conditions.

At the regional scale, integration of eight thematic criteria using the Analytic Hierarchy Process (AHP) identified 21.82% (58.41 km2) of the representative assessment area as Very Suitable for MAR. Monte Carlo simulation demonstrated relatively low sensitivity of the suitability model to variations in criterion weights and input raster values, while comparison with the Fuzzy Gamma model showed that the general spatial suitability pattern was maintained despite differences between the two multicriteria aggregation approaches. These analyses provide complementary information on parameter and methodological uncertainty and demonstrate the robustness of the regional suitability patterns.

Independent site-scale field verification in the Byzhy River Basin provided additional evidence supporting the regional screening results. Three double-ring infiltration tests, which were not used as input criteria in the regional AHP model, yielded field-saturated hydraulic conductivity values of 1.86–1.96 m·day−1 (mean 1.92 m·day−1), confirming favorable infiltration properties of the Quaternary alluvial deposits. The positive long-term mean potential water balance of 27.69 mm·year−1, groundwater mineralization below 1 g·L−1 within the proposed MAR area, and the presence of an approximately 50-m-thick Quaternary alluvial aquifer further supported the hydrogeological suitability of the selected site. UAV-based LiDAR surveying complemented these investigations by providing high-resolution terrain information for characterizing local microtopography and surface-flow directions at the proposed MAR site.

The water-resource and storage assessment further demonstrated the practical potential of the selected MAR area. Based on a conservative allocation equivalent to 10% of the long-term mean river discharge, the theoretical annual volume potentially available for MAR was estimated at 17.44 million m3·year−1. This volume represents only 2.33% of the estimated effective aquifer storage capacity of 746.8 million m3, indicating substantial potential for subsurface storage at the investigated site.

The novelty of this study therefore lies not in regional suitability mapping alone, but in linking GIS-based screening and uncertainty analysis with independent field-scale hydrogeological verification and quantitative assessment of water availability and aquifer storage potential, complemented by high-resolution UAV-LiDAR terrain characterization, within a single regional-to-site-scale framework. The results demonstrate that areas identified as favorable through regional spatial analysis can be systematically screened and subsequently verified using field observations before progressing toward pilot-scale MAR implementation. The proposed framework provides a transferable basis for MAR planning in Kazakhstan and other semi-arid regions where seasonal surface-water availability coincides with the need for long-term subsurface water storage.

Statements

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Author contributions

ZO: Visualization, Formal analysis, Methodology, Conceptualization, Investigation, Writing – review & editing, Data curation. SK: Writing – review & editing, Methodology, Visualization, Formal analysis, Data curation, Writing – original draft. RA: Formal analysis, Writing – original draft, Visualization, Data curation. AAb: Writing – original draft, Visualization, Data curation. AAl: Visualization, Data curation, Writing – original draft. IR: Investigation, Data curation, Writing – original draft. MA: Funding acquisition, Writing – original draft, Project administration.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This research has been funded by the Science Committee of the Ministry of Science and Higher Education of the Republic of Kazakhstan “Scientific and practical justification for the sustainable development of livestock farming based on groundwater irrigation of pastures”, grant no. BR24992885.

Acknowledgments

The authors gratefully acknowledge the Republican State Enterprise “Kazhydromet” for providing the hydrometeorological observation data used in this study. The authors also acknowledge the developers of the publicly available TerraClimate dataset and the Copernicus Climate Change Service (C3S), implemented by the European Center for Medium-Range Weather Forecasts (ECMWF), for providing the ERA5-Land dataset used in this research.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that Generative AI was used in the creation of this manuscript. Generative AI tools were used to support language editing and improve clarity of the manuscript. No AI tools were used for data analysis or generation of scientific content.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/frwa.2026.1932313/full#supplementary-material

References

  • 1

    AmanzholovaR. S.RakhmetovI. K.SaginZ. M.OnglassynovZ. A. (2024). Assessing groundwater for managed aquifer recharge in eastern Kazakhstan. J. Geogr. Environ. Manag.79:47. doi: 10.26577/JGEM20257947

  • 2

    ASTM International (2018). ASTM D3385-18: (2018). Standard Test Method for Infiltration Rate of Soils in Field Using Double-Ring Infiltrometer. West Conshohocken, PA: ASTM International.

  • 3

    BilalH.GovindanR.ZekriS.Al-MaktoumiA.RajabiM. M.TrikiC.et al. (2025). Identification of potential managed aquifer recharge sites in hyper-arid environment using GIS and analytical hierarchy process. Appl Water Sci15:99. doi: 10.1007/s13201-025-02431-3

  • 4

    BouwerH. (2002). Artificial recharge of groundwater: hydrogeology and engineering. Hydrogeol. J.10, 121–142. doi: 10.1007/s10040-001-0182-4

  • 5

    CarselR. F.ParrishR. S. (1988). Developing joint probability distributions of soil water retention characteristics. Water Resour. Res.24, 755–769. doi: 10.1029/WR024i005p00755

  • 6

    CasanovaJ.DevauN.PettenatiM. (2016). “Managed aquifer recharge: an overview of issues and options,” in Integrated Groundwater Management: Concepts, Approaches and Challenges, eds. JakemanA. J.BarreteauO.HuntR. J.RinaudoJ.-D.RossA. (Cham: Springer), 413–434.

  • 7

    ChekirbaneA.KhemiriK.MartinsT. N.StefanC.PanagiotouC. F. (2025). A participatory GIS-based multicriteria decision analysis approach to map the geospatial feasibility of managed aquifer recharge in a Tunisian coastal watershed. Environ. Process.12:28. doi: 10.1007/s40710-025-00769-z

  • 8

    DillonP.AlleyW.ZhengY.VanderzalmJ. (2022) Managed aquifer recharge: overview and governance. IAH special publication. International Association of Hydrogeologists: London, UK. Available online at: https://recharge.iah.org/files/2022/06/MAR-overview-and-governance-IAH-Special-Publication-18June2022.pdf (Accessed March 17, 2026)

  • 9

    DillonP.PavelicP.PageD.BeringenH.WardJ. (2009). Managed aquifer recharge: an introduction. Waterlines report series no. 13; National Water Commission: Canberra, Australia. Available online at: https://recharge.iah.org/monographs (Accessed May 3, 2026)

  • 10

    DillonP.StuyfzandP.GrischekT.LluriaM.PyneR. D. G.JainR. C.et al. (2019). Sixty years of global progress in managed aquifer recharge. Hydrogeol. J.27, 1–30. doi: 10.1007/s10040-018-1841-z

  • 11

    DillonP.TozeS.PageD.VanderzalmJ.BekeleE.SidhuJ.et al. (2010). Managed aquifer recharge: rediscovering nature as a leading-edge technology. Water Sci. Technol.62, 2338–2345. doi: 10.2166/wst.2010.444,

  • 12

    EbrahimG. Y.StefanC.SallweyJ.LautzeJ. (2024). Mapping the potential of managed aquifer recharge in Africa: GIS-based multi-criteria decision analysis approach. Groundw. Sustain. Dev.27:101374. doi: 10.1016/j.gsd.2024.101374

  • 13

    ECMWF. (2019). ERA5-land dataset covers the period from 1950 to the present. Copernicus climate change service (C3S), implemented by ECMWF. Available online at: https://cds.climate.copernicus.eu/datasets/reanalysis-era5-land (Accessed March 17, 2026).

  • 14

    Environment and Natural Resources Kazakhstan Institute of Geography (KIG). (2022). National atlas of Kazakhstan. Available online at: https://ingeo.kz/ (Accessed March 17, 2026).

  • 15

    Esri Living Atlas. (2021). European Space Agency WorldCover 2021 Land Cover. ArcGIS Living Atlas of the World. Available at: ArcGIS Living Atlas (Accessed February 2, 2026)

  • 16

    FamigliettiJ. S. (2014). The global groundwater crisis. Nat. Clim. Chang.4, 945–948. doi: 10.1038/nclimate2425

  • 17

    FAO/UNESCO. (2003). Digital soil map of the world and derived soil properties, version 3.6; Available online at: https://cds.climate.copernicus.eu/datasets/reanalysis-era5-land (Accessed March 17, 2026)

  • 18

    FerenczS. B.MangelA.Day-LewisF. (2024). Managed aquifer recharge as a strategy to redistribute excess surface flow to baseflow in snowmelt hydrologic regimes. Front. Water6:1375523. doi: 10.3389/frwa.2024.1375523

  • 19

    FerranteJ.A. (2023) Feasibility of managed aquifer recharge (MAR) for southwest Kansas. Available online at: https://msoe.tind.io/record/898 (Accessed May 3, 2026)

  • 20

    FerranteJ. A.GonwaW. S. (2024). Feasibility of managed aquifer recharge in southwest Kansas. J. Irrig. Drain. Eng.150:04024036. doi: 10.1061/JIDEDH.IRENG-10211

  • 21

    GaleI. (2005). Strategies for managed aquifer recharge (MAR) in semi-arid areas. Available online at: https://recharge.iah.org/files/2017/01/Gale-Strategies-for-MAR-in-semiarid-areas.pdf (Accessed April 4, 2026)

  • 22

    García-MenéndezO.Renau-PruñonosaA.MorellI.BallesterosB. J.EstellerM. V. (2021). Hydrogeochemical changes during managed aquifer recharge (MAR) in a salinised coastal aquifer. Appl. Geochem.126:104866. doi: 10.1016/j.apgeochem.2020.104866

  • 23

    GleesonT.WadaY.BierkensM. F. P.van BeekL. P. H. (2012). Water balance of global aquifers revealed by groundwater footprint. Nature488, 197–200. doi: 10.1038/nature11295,

  • 24

    GreenT. R.TaniguchiM.KooiH.GurdakJ. J.AllenD. M.HiscockK. M.et al. (2011). Beneath the surface of global change: impacts of climate change on groundwater. J. Hydrol.405, 532–560. doi: 10.1016/j.jhydrol.2011.05.002

  • 25

    GuoZ.ChenK.YiS.ZhengC. (2023). Response of groundwater quality to river-aquifer interactions during managed aquifer recharge: a reactive transport modeling analysis. J. Hydrol.616:128847. doi: 10.1016/j.jhydrol.2022/.128847

  • 26

    GuttmanJ.NegevI.RubinG. (2017). Design and testing of recharge wells in a coastal aquifer: summary of field scale pilot tests. Water9:53. doi: 10.3390/w9010053

  • 27

    IssaldayevaS.AlimkulovS.RaimbekovaZ.BekseitovaR.KaratayevM. (2023). The climatic and river runoff trends in central Asia: the case of the Zhetysu Alatau region, the south-eastern part of Kazakhstan. Heliyon9:e17897. doi: 10.1016/j.heliyon.2023.e17897

  • 28

    JabassovA.OnglassynovZ.AlimgazinaA.ErmenbayA.EreevD.AbyshevA.et al. (2026). Identification of suitable managed aquifer recharge sites using GIS-AHP and field-based evaluation of aquifer storage capacity in central Kazakhstan. Water18:1410. doi: 10.3390/w18121410

  • 29

    KaldzbazevA.ChenY.IssanovaG.WangH.MahmudovaL. (2016). Runoff response to the glacier shrinkage in the Karatal river basin, Kazakhstan. Arabian J.Geosci.9, 1–8. doi: 10.1007/s12517-015-2106-y,

  • 30

    KartheD.ChalovS.BorchardtD. (2015). Water resources and their management in central Asia in the early twenty-first century: status, challenges and future prospects. Environ. Earth Sci.73, 487–499. doi: 10.1007/s12665-014-3789-1

  • 31

    KazakisN.KougiasI.PatsialisT. (2015). Assessment of flood hazard areas at a regional scale using an index-based approach and analytical hierarchy process: application in Rhodope–Evros region, Greece. Sci. Total Environ.538, 555–563. doi: 10.1016/j.scitotenv.2015.08.055,

  • 32

    KazakisN.VoudourisK.PavlouA. (2018). GIS-based groundwater recharge potential assessment in Mediterranean regions. Environ. Monit. Assess.190:52. doi: 10.1007/s10661-017-6403-3

  • 33

    KazemM.FadlelsaidA. M.AbazarM. A. D.BudayT. (2026). Uncertainty-aware groundwater potential mapping in arid basement terrain using AHP and Dirichlet-based Monte Carlo simulation: evidence from the Sudanese Nubian shield. Water18:901. doi: 10.3390/w18080901

  • 34

    Kazhydromet. (2025). Annual hydrometeorological bulletin of Kazakhstan; Available online at: https://www.kazhydromet.kz/en/klimat/ezhegodnyy-byulleten-monitoringa-sostoyaniya-i-izmeneniya-klimata-kazahstana (Accessed March 17, 2026)

  • 35

    KhanM. Y. A.El KashoutyM.ZaidiF. K.EgbueriJ. C. (2023). Mapping aquifer recharge potential zones (ARPZ) using integrated geospatial and analytic hierarchy process (AHP) in an arid region of Saudi Arabia. Remote Sens.15:2567. doi: 10.3390/rs15102567

  • 36

    LiM.LiuT.DuanL.LuoY.MaL.ZhangJ.et al. (2019). The scale effect of double-ring infiltration and soil infiltration zoning in a semi-arid steppe. Water11:1457. doi: 10.3390/w11071457

  • 37

    LiuY.EckertC. M.EarlC. (2020). A review of fuzzy AHP methods for decision-making with subjective judgements. Expert Syst. Appl.161:113738. doi: 10.1016/j.eswa.2020.113738

  • 38

    MachiwalD.JhaM. K.MalB. C. (2011). Assessment of groundwater potential in arid regions using remote sensing, GIS and multicriteria decision analysis techniques. Hydrogeol. J.19, 919–929. doi: 10.1007/s10040-011-0732-5

  • 39

    MaréchalJ.-C.BouzitM.RinaudoJ.-D.MoirouxF.DespratsJ.-F.CaballeroY. (2020). Mapping economic feasibility of managed aquifer recharge. Water12:680. doi: 10.3390/w12030680

  • 40

    MaréchalJ.-C.SellesA.DewandelB.BoissonA.PerrinJ.AhmedS. (2018). An observatory of groundwater in crystalline rock aquifers exposed to a changing environment: Hyderabad, India. Vadose Zone J.17:180076. doi: 10.2136/vzj2018.04.0076

  • 41

    MartinsT. N.LeitãoT. E.OliveiraM. M.PanagiotouC. F.StefanC.ChkirbeneA.et al. (2024). Proposal for a managed aquifer recharge feasibility index for southern Portugal using multi-criteria decision analysis. Groundw. Sustain. Dev.26:101280. doi: 10.1016/j.gsd.2024.101280

  • 42

    MicklinP. (2007). The Aral sea disaster. Annu. Rev. Earth Planet. Sci.35, 47–72. doi: 10.1146/annurev.earth.35.031306.140120

  • 43

    Ministry of Geology of the USSR. (1977). Hydrogeological map of the USSR, scale 1:200,000; L-44-XXX. Available online at: https://www.geokniga.org/topolists/19200 (Accessed April 1, 2026)

  • 44

    MurrayR.TredouxG.RavenscroftP.BothaF. (2007). Artificial recharge strategy: version 1.3; Available online at: https://www.artificialrecharge.co.za/strategydocument/ARStrategySectA.pdf (Accessed March 17, 2026)

  • 45

    NASA SRTM DEM database (2024). NASA Shuttle Radar Topography Mission (SRTM) Global 1 Arc Second Digital Elevation Model (SRTMGL1). NASA Land Processes Distributed Active Archive Center (LP DAAC). Available at: NASA SRTM Global 1 Arc Second V003 (Accessed January 29, 2026).

  • 46

    NeshatA.PradhanB.DadrasM. (2014). Groundwater vulnerability assessment using Monte Carlo simulation and GIS-based models. Environ. Earth Sci.71, 349–364. doi: 10.1007/s12665-013-2430-8

  • 47

    NogueiraG.E.H. (2017) Tracing the hydrochemical water types and salinization mechanisms in the great Maputo area as a function of groundwater recharge, hydrogeological properties and human activities. Available online at: https://scholar.tecnico.ulisboa.pt/records/U2J_J_GaW_4j_ZgzYha0H4LldA9-Lp5grJxQ?lang (Accessed May 22, 2026)

  • 48

    OnglassynovZ.BerndtssonR.RakhimovaV.RakhimovT.JabassovA.RakhmetovI.et al. (2025). GIS-based multi-criteria assessment of managed aquifer recharge (MAR) zones using the analytic hierarchy process (AHP) method in southern Kazakhstan. Water17:2774. doi: 10.3390/w17182774

  • 49

    PanagiotouC. F.EisenreichS.BaroutaO. T.ChekirbaneA.MartinsT.NeophytidesS.et al. (2024). Identification of feasible regions for managed aquifer recharge in the Republic of Cyprus using a co-participative multi-criteria decision analysis. Groundw. Sustain. Dev.27:101323. doi: 10.1016/j.gsd.2024.101323

  • 50

    PettenatiM.CroisetN.Picot-ColbeauxG.CasanovaJ.AzaroualM.BesnardK.et al. (2012). Optimisation of wastewater treatments through combined geomaterials and natural soil filter: modelling tools. J. Water Reuse Desalination2, 185–193. doi: 10.2166/wrd.2012.023

  • 51

    RinglebJ.SallweyJ.StefanC. (2016). Assessment of managed aquifer recharge through modeling—a review. Water8:579. doi: 10.3390/w8120579

  • 52

    RobertsonA. J.KennedyJ. R.WildermuthL. M.BellM. T.FuchsE. H.RinehartA.et al. (2023). Determining seasonal recharge, storage changes, and specific yield using repeat microgravity and water-level measurements in the Mesilla Basin alluvial aquifer, New Mexico, 2016–2018. J. Appl. Geophys.209:104916. doi: 10.1016/j.jappgeo.2022.104916

  • 53

    RussoT. A.FisherA. T.LockwoodB. S. (2014). Assessment of managed aquifer recharge potential using GIS and hydrogeological modeling in California. J. Hydrol.519, 3290–3304. doi: 10.1016/j.jhydrol.2014.10.025

  • 54

    SaatyT. L. (1980). The Analytic Hierarchy Process: Planning, Priority Setting, Resource Allocation. New York, NY: McGraw-Hill.

  • 55

    SallweyJ.Bonilla ValverdeJ. P.Vásquez LópezF.JunghannsR.StefanC. (2019). Suitability maps for managed aquifer recharge: a review of multi-criteria decision analysis studies. Environ. Rev.27, 138–150. doi: 10.1139/er-2018-0069

  • 56

    SallweyJ.OngdasN.Al-HosbanM.StefanC. (2024). Mapping the potential for managed aquifer recharge in Kazakhstan. Cent. Asian J. Water Res.10, 95–116. doi: 10.29258/CAJWR/2024-R1.v10-2/95-116.eng

  • 57

    ScanlonB. R.HealyR. W.CookP. G. (2002). Choosing appropriate techniques for quantifying groundwater recharge. Hydrogeol. J.10, 18–39. doi: 10.1007/s10040-001-0176-2

  • 58

    ScanlonB. R.ReedyR. C.FauntC. C.PoolD.UhlmanK. (2016). Enhancing drought resilience with conjunctive use and managed aquifer recharge in California and Arizona. Environ. Res. Lett.11:035013. doi: 10.1088/1748-9326/11/3/035013

  • 59

    SeifA. K.MasriaA.GhareebM.SalehA. A.SolimanK.AmmarA. I. (2024). Identifying managed aquifer recharge and rain water harvesting sites and structures for storing non-conventional water using GIS-based multi-criteria decision analysis approach. Appl Water Sci14:181. doi: 10.1007/s13201-024-02246-8

  • 60

    SmolyarV. A.MustafayevS. T. (2007). Hydrogeology of the Lake Balkhash Basin. Almaty: Gylym.

  • 61

    SotnikovE.OnglassynovZ.RakhimovaV.KanafinK.MiroshnichenkoO.BerndtssonR.et al. (2025). Identifying fresh groundwater potential in unconfined aquifers in arid Central Asia: a remote sensing and geo-information modeling approach. Water17:2985. doi: 10.3390/w17202985

  • 62

    StefanC.AnsemsN. (2018). Web-based global inventory of managed aquifer recharge applications. Sustain. Water Resour. Manag.4, 153–162. doi: 10.1007/s40899-017-0212-6

  • 63

    SufyanM.MartelliG.TeatiniP.CherubiniC.GoiD. (2024). Managed aquifer recharge for sustainable groundwater management: new developments, challenges, and future prospects. Water16:3216. doi: 10.3390/w16223216

  • 64

    TaylorR. G.ScanlonB.DöllP.RodellM.van BeekR.WadaY.et al. (2013). Ground water and climate change. Nat. Clim. Chang.3, 322–329. doi: 10.1038/nclimate1744

  • 65

    TuinhofA.HeederikJ.P. (2003) Netherlands National Committee for the IAH. Available online at: https://www.hydrology.nl/images/docs/iah/publications/4_Management_of_Aquifer_Recharge_and_Subsurface_Storage.pdf (Accessed March 17, 2026)

  • 66

    TurnerB.L.SkoleD.SandersonS.FischerG.FrescoL.LeemansR. (1995). Land-use and land-cover change: science/research plan. Available online at: https://pure.iiasa.ac.at/4402/ (Accessed March 17, 2026)

  • 67

    U.M. Akhmedsafin Institute of Hydrogeology and Environmental Geosciences. (2022). Atlas of hydrogeological maps of the Republic of Kazakhstan. Available online at: https://science.satbayev.university/ru/atlas (Accessed March 17, 2026)

  • 68

    University of Idaho. (2026). TerraClimate. Available online at: https://www.climatologylab.org/terraclimate.html (Accessed March 17, 2026)

  • 69

    VeselovV.V. (2002) Hydrogeological zoning and regional assessment of groundwater resources of Kazakhstan: analysis of research results for 1961–2002. Available online at: https://www.geokniga.org/books/31655 (Accessed March 17, 2026)

  • 70

    WadaY.van BeekL. P. H.BierkensM. F. P. (2010). Global depletion of groundwater resources. Geophys. Res. Lett.37:L20402. doi: 10.1029/2010GL044571,

  • 71

    Yazdani MinaP.van ThienenS. F.StofbergM. H. J.van HuijgevoortM. H. J.BartholomeusR. P. (2026). Managed aquifer recharge: modeling approaches to integrated assessment of groundwater interventions. Groundwater Sustain. Dev.32:101552. doi: 10.1016/j.gsd.2025.101552,

  • 72

    ZhengY.RossA.VillholthK.G.DillonP. (2021) Managing aquifer recharge: a showcase for resilience and sustainability. Available online at: https://unesdoc.unesco.org/ark:/48223/pf0000379962 (Accessed March 17, 2026)

Summary

Keywords

analytical hierarchy process, Byzhy river basin, GIS, groundwater management, Kazakhstan, managed aquifer recharge, site suitability assessment

Citation

Onglassynov Z, Kulbekova S, Amanzholova R, Abyshev A, Alimgazina A, Rakhmetov I and Absametov M (2026) A multi-scale framework for identifying and validating managed aquifer recharge sites in semi-arid Kazakhstan. Front. Water 8:1932313. doi: 10.3389/frwa.2026.1932313

Received

08 July 2026

Revised

16 September 2026

Accepted

30 September 2026

Published

16 October 2026

Volume

8 - 2026

Edited by

Naveen Joseph, Radford University Artis College of Science and Technology, United States

Reviewed by

Zaidoon T. Abdulrazzaq, Scientific Research Commission, Iraq

Mezali Farouk, University of M’sila, Algeria

Updates

Copyright

*Correspondence: Sholpan Kulbekova, ; Raushan Amanzholova,

Disclaimer

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Outline

Figures

Cite article

Copy to clipboard


Export citation file


Share article

Article metrics