Sediment management strategy prioritization in hydropower reservoirs under data scarcity

1 Introduction

Hydropower is one of the world’s major renewable electricity sources (Tefera and Kasiviswanathan, 2022). Its long-term sustainability depends on maintaining reservoir storage capacity, which is increasingly threatened by reservoir sedimentation (Gaudard and Romerio, 2014). Reservoir sedimentation is a major challenge in many regions of the world and has serious consequences for water resources management, flood control, and energy production (Hauer et al., 2018). Excessive sediment deposition reduces reservoir storage capacity, decreases hydropower generation, increases water turbidity, lowers flood conveyance capacity, and deteriorates water quality (Adeogun et al., 2018; Anzani et al., 2026). Therefore, effective sediment management is essential for maintaining the economic and sustainable operation of hydropower reservoirs. These challenges become even more critical in reservoirs where hydrological and sediment monitoring data are limited, preventing the application of physically based sediment management models.

Sediment management aims to maintain reservoir functionality by reducing sediment inflow, routing sediment through or around reservoirs, or removing deposited sediment. Sediment accumulation rates are influenced not only by reservoir characteristics but also by catchment-scale factors such as land use changes, vegetation loss, and sediment supply conditions (Lo et al., 2022). According to Morris (2020), sediment management strategies can be broadly classified into active and adaptive approaches. Active strategies include watershed erosion control, sediment routing, and sediment removal, whereas adaptive strategies seek to reduce the impacts of storage loss without directly restoring the sediment balance. Various techniques, such as watershed conservation practices, sediment bypass tunnels, flushing, sluicing, and dredging, have been applied to reduce the adverse effects of sedimentation (Owens et al., 2005). No single sediment management strategy is universally applicable because reservoir characteristics, sediment conditions, and operational objectives differ among hydropower reservoirs (Anzani et al., 2026). Sediment transport processes are influenced by interactions among hydrological conditions, sediment sources, and catchment characteristics, which can create considerable uncertainty in sediment management decisions (Vaheddoost et al., 2022). Changing hydroclimatic conditions can affect water availability and increase drought risk, creating additional challenges for water-resource management (Ozcelik et al., 2024). Extreme events such as floods and droughts can also affect sediment transport and deposition in reservoirs, increasing the complexity of sediment management (Landwehr et al., 2022).

Selecting an appropriate sediment management strategy is therefore a complex decision-making problem. Beyond their effectiveness in controlling sedimentation, sediment management strategies also differ in their technical, economic, environmental, and social implications (Quaranta et al., 2023). This challenge is particularly significant in data-scarce reservoirs, where consistent records of water level, discharge, and sediment concentration are often unavailable or incomplete (Rautela et al., 2025). Under such conditions, expert knowledge becomes an important source of information for evaluating sediment management strategies and supporting decision-making. Experienced reservoir managers and researchers often possess operational knowledge that cannot be fully captured by limited monitoring records. Consequently, expert-based evaluation provides a practical alternative when quantitative datasets are incomplete or unavailable.

Because sediment management strategies influence technical, economic, environmental, and social aspects simultaneously, their evaluation requires balancing multiple and often conflicting criteria (Alvarez-Guerra et al., 2009; Quaranta et al., 2023). Considering these sustainability dimensions together with different stakeholder perspectives is essential for effective planning and long-term management (Yilmaz, 2025; Yilmaz and Karadayi-Usta, 2026). Recent studies have also highlighted the need for decision-support frameworks that integrate sustainability assessment with Multi-Criteria Decision-Making (MCDM) methods for reservoir sediment management. Although MCDM methods have been successfully applied to several sediment-related decision problems, their application to sediment management strategy prioritization in hydropower reservoirs remains limited (Otálvaro Barco et al., 2025). Since the evaluation of sediment management strategies involves multiple criteria, MCDM methods provide an appropriate basis for comparing alternatives and supporting decision-making.

Among the available MCDM methods, the Best-Worst Method (BWM) requires fewer pairwise comparisons and can improve the consistency of criteria weighting compared with many conventional weighting approaches (Rezaei, 2016). For ranking alternatives, the Combined Compromise for Ideal Solution (CoCoFISo) method integrates multiple compromise measures to improve the discrimination and stability of ranking results (Rasoanaivo et al., 2024). Therefore, this study proposes an integrated BWM-CoCoFISo framework for prioritizing sediment management strategies in hydropower reservoirs under data-scarce conditions. The main contributions of this study are threefold. First, it addresses sediment management decisions in data-scarce reservoirs by incorporating expert judgments into the evaluation process. Second, it provides a comprehensive sustainability assessment by integrating technical, economic, environmental, and social criteria. Third, it establishes a consistent prioritization process for evaluating a broad range of sediment management strategies using optimization-based weighting and compromise-based ranking.

This paper is organized into six sections. Section 2 reviews the literature on sediment management in hydropower reservoirs and MCDM applications in sediment management and highlights the research gaps motivating this study. Section 3 describes the proposed methodology. Section 4 presents the results, including sensitivity and comparative analyses. Section 5 discusses the findings and their practical implications. Finally, Section 6 concludes the paper and suggests directions for future research.

2 Literature review

This section reviews the existing literature on sediment management, sustainability assessment, and MCDM approaches. Based on the reviewed studies, the main research gaps that motivate the proposed framework are identified.

2.1 Sediment management in hydropower reservoirs

A systematic literature review was conducted using the Scopus database to identify studies on sediment management in hydropower reservoirs. On 8 June 2026, the database was searched using the query (“reservoir sediment” OR “sediment management”) AND “hydropower*”, with the search restricted to English-language journal articles. A total of 189 publications were retrieved and analyzed using VOSviewer. A minimum keyword occurrence threshold of four was applied to construct the keyword co-occurrence network shown in Figure 1.

Figure 1 reveals six major thematic groups within the literature. The dominant research themes include watershed sediment generation and transport, sediment management strategies for sustainable hydropower, reservoir operation and sedimentation processes, environmental impacts of reservoir sedimentation, hydraulic and experimental sediment removal techniques, and numerical simulation of sediment dynamics. These themes indicate that sediment management in hydropower reservoirs has been investigated from hydrological, hydraulic, environmental, and operational perspectives, reflecting the multidisciplinary nature of the field. Among these themes, sediment management strategies constitute one of the most extensively studied research areas. These strategies are commonly classified into active and adaptive approaches.

Morris (2020) described active sediment management as an approach that preserves sediment balance by controlling sediment inflow, routing sediment around or through reservoirs, and managing deposited sediments. In contrast, adaptive strategies consist of measures that accommodate the loss of reservoir storage capacity without directly addressing the sediment balance. Studies focusing on reducing sediment yield from the watershed have primarily evaluated erosion control measures. Kaura et al. (2019) investigated forest conservation and showed that reducing deforestation effectively decreased reservoir sedimentation while improving long-term hydropower performance and economic outcomes. Adeogun et al. (2020) evaluated vegetative filter strips and stone bunds and reported substantial reductions in sediment yield and sediment concentration in erosion-prone areas. Similarly, Adeogun et al. (2018) compared reforestation, vegetative filter strips, and stone bunds, finding that vegetative filter strips and reforestation were the most effective in reducing sediment yield, while reforestation was the most cost-effective strategy. Beyond watershed-based strategies, reservoir-based sediment routing strategies have also been widely investigated. Dahal et al. (2025) demonstrated that sediment bypass tunnels effectively reduced reservoir sedimentation while emphasizing the importance of optimizing reservoir operation to balance sediment management and hydropower generation.

Several studies have integrated multiple active sediment management strategies. Shrestha et al. (2021) compared catchment-level strategies (terracing, vegetative filter strips, and no-tillage) with the reservoir-level strategy of flushing and highlighted the benefits of integrating watershed- and reservoir-based strategies to reduce sedimentation and associated costs. Abdul Razad et al. (2018) evaluated check dams, settling basins, dredging, and bottom-outlet flushing, demonstrating that combining watershed erosion control with reservoir sediment removal can increase reservoir storage capacity. Glas et al. (2025) investigated drawdown sluicing, drawdown flushing, and dredging, showing that combining sediment routing and sediment removal strategies can support sustainable reservoir sediment management while minimizing downstream impacts. Similarly, Invernizzi et al. (2025) optimized sluicing and flushing through coordinated reservoir operation and demonstrated that integrating sediment routing and sediment removal strategies can simultaneously improve sediment connectivity and hydropower performance. In contrast, adaptive sediment management strategies include focusing or redistributing sediment, increasing storage capacity, improving operational efficiency, modifying infrastructure, and adjusting to reduced reservoir benefits. Rather than directly controlling sediment, these approaches aim to adapt reservoir systems to the long-term impacts of sedimentation. Compared with active sediment management, however, adaptive strategies have received relatively less attention in the recent hydropower literature (Morris, 2020). Anzani et al. (2026) noted that sediment management strategies can be classified into preventive measures, such as watershed management and sediment trapping, and corrective measures, including flushing, bypass tunnels, dredging, and density current venting. The review further highlighted that selecting appropriate strategies requires consideration of reservoir-specific characteristics and operational conditions, reinforcing the need for integrated sediment management approaches.

Although previous studies have investigated a wide range of sediment management strategies, most have primarily evaluated their effectiveness in reducing sedimentation, maintaining reservoir storage capacity, or improving hydropower performance. However, these strategies also affect economic, environmental, technical, and social aspects of hydropower systems. To better understand these broader impacts, previous hydropower sustainability studies were classified according to the sustainability dimensions they considered, with the results summarized in Table 1.

TABLE 1. Sustainability dimensions addressed in previous hydropower studies.

As shown in Table 1, previous studies have evaluated hydropower sustainability from different perspectives. The economic dimension focuses on implementation costs, cost-effectiveness, and broader economic benefits. The environmental dimension considers impacts on ecosystems, water quality, biodiversity, and other ecological conditions. The technical dimension addresses reservoir operation, hydropower generation performance, infrastructure, and the technical feasibility of management strategies. The social dimension encompasses stakeholder interests, community wellbeing, and other societal impacts. While many studies focused on one or several sustainability dimensions, comprehensive frameworks integrating all four dimensions remain limited. This limitation is particularly important because sediment management strategies differ in their technical feasibility, implementation costs, environmental impacts, and social implications. Evaluating these strategies therefore requires balancing trade-offs across multiple sustainability dimensions. However, comprehensive sustainability assessments specifically designed for evaluating sediment management strategies in hydropower reservoirs are still scarce. Therefore, this study integrates economic, environmental, technical, and social dimensions into a unified evaluation framework for assessing sediment management strategies.

2.2 Applications of multi-criteria decision-making methods in sediment management

Sediment management requires complex decision-making processes due to the integration of multiple technical, environmental, economic, and social criteria. MCDM methods have been widely applied to support the evaluation, prioritization, and selection of sediment management alternatives. Previous studies have employed various MCDM approaches, including weighting methods, compromise-based ranking techniques, outranking methods, and hybrid frameworks, to address various sediment-related decision problems such as contaminated sediment remediation, sediment reuse, risk assessment, and management alternative selection. Table 2 summarizes representative MCDM applications in sediment management.

TABLE 2. Summary of MCDM applications in sediment management.

AHP: Analytic Hierarchy Process, BWM: Best-Worst Method, CoCoFISo: Combined Compromise for Ideal Solution, FUCOM: Full Consistency Method, SAW: Simple Additive Weighting, SMAA: Stochastic Multicriteria Acceptability Analysis, TOPSIS: Technique for Order Preference by Similarity to Ideal Solution, VIKOR: VIseKriterijumska Optimizacija i KOmpromisno Resenje.

The reviewed studies indicate that MCDM approaches have been mainly employed for addressing sediment-related decision problems involving contaminated sediment remediation, risk assessment, and the prioritization of management alternatives. Among the applied methodologies, hybrid frameworks combining weighting and ranking techniques have been frequently preferred, as they enable the integration of diverse criteria and expert judgments within complex decision environments. In particular, AHP- and VIKOR-based hybrid models have been widely adopted for remediation strategy selection, whereas SAW, TOPSIS, PROMETHEE, and related compromise-based approaches have been utilized for ranking sediment management strategies and evaluating environmental risks. The literature shows that MCDM provides a flexible decision-support framework for sediment management by accommodating multiple criteria and conflicting objectives.

2.3 Research gaps and motivation

Sections 2.1 and 2.2 demonstrate that previous studies have extensively investigated sediment management strategies in hydropower reservoirs and the application of MCDM methods to sediment-related decision problems. However, the integration of these two research domains remains limited. Existing sediment management studies have primarily focused on hydraulic performance, sediment transport processes, and technical optimization, while comprehensive sustainability-based evaluations considering economic, environmental, technical, and social dimensions remain insufficient. Furthermore, although MCDM approaches have been increasingly applied in sediment-related decision problems, their application to the prioritization of sediment management strategies in hydropower reservoirs is still scarce. Existing applications often rely on site-specific quantitative data, which may limit their applicability in data-scarce reservoir systems where expert knowledge and qualitative assessments play an important role. In addition, the potential of combining optimization-based weighting methods with compromise-based ranking approaches has not been sufficiently explored for sustainable sediment management.

To address these gaps, this study proposes an integrated BWM-CoCoFISo framework for prioritizing sediment management strategies in hydropower reservoirs under data-scarce conditions. Specifically, the proposed framework integrates sustainability dimensions into sediment management evaluation, enables decision-making under data-scarce conditions through expert-based assessments, combines BWM-based criteria weighting with CoCoFISo-based alternative ranking, and evaluates a broader range of sediment management strategies within a unified sustainability framework.

3 Research framework and methodology

3.1 Expert group and judgment elicitation

Expert judgments were used to evaluate the criteria and sediment management strategies under the data-scarce conditions considered in this study. An expert group was formed to provide informed judgments for BWM-based criterion weighting and CoCoFISo evaluation. Unlike data-based methodologies, expert-based approaches focus on informed judgments from knowledgeable participants and can therefore be conducted with relatively small expert groups (Rezaei et al., 2012; Moktadir et al., 2020). The group size was considered appropriate for structured expert elicitation, where the emphasis is placed on the relevance and diversity of expert knowledge rather than statistical representativeness. This approach is also consistent with previous expert-based MCDM studies that have used relatively small expert groups for structured decision-making (Otay and Kahraman, 2026; Akram et al., 2026).

The expert group consisted of ten experts from Türkiye and represented three stakeholder groups: academia, the public sector, and the private sector. The academic group included four researchers with expertise in hydrology, sediment management, and MCDM. The public-sector group consisted of two engineers and one technician. The private-sector group included three engineers. Expert selection considered experience in sediment management and reservoir systems, hydrology and water resources, and MCDM and decision analysis. The experts had different levels of professional experience. Two experts had up to 10 years of experience, three had 11–15 years, two had 16–20 years, and three had more than 20 years of experience.

Before the evaluation, each expert received the same briefing document. The document described the study objectives, evaluation criteria, sediment management strategies, and evaluation procedure. Definitions of the criteria and strategies were also provided. The experts completed the evaluations independently.

Two different uses of the 1–9 scale were applied in the framework. In the BWM, the standard nine-point comparison scale was used to assess the relative importance of the criteria, ranging from 1 (equally important) to 9 (extremely more important). In the CoCoFISo evaluation, a nine-point numerical rating scale was used to assess the relative performance of each sediment management strategy for each criterion. For benefit-oriented criteria, a score of 1 represented the least favorable performance and a score of 9 represented the most favorable performance. For cost-related criteria, lower-cost strategies received higher scores, while higher-cost strategies received lower scores. Therefore, for investment cost (C4) and operation and maintenance cost (C5), higher scores indicate greater cost favorability and a relatively lower cost burden. These scores represent relative expert assessments and not observed monetary values.

The individual expert judgments obtained through this procedure were used in the BWM and CoCoFISo analyses described in Section 3.2. Equal importance was assigned to all experts in the baseline analysis. Assigning different weights according to professional experience or stakeholder group would introduce an additional subjective judgment regarding the relative influence of individual experts. Therefore, equal expert weighting was used in the analysis.

3.2 Methodology

In this study, the linear BWM, proposed by Rezaei (2016), was employed to determine the weights of the evaluation criteria. Compared with the original non-linear BWM formulation (Rezaei, 2015), the linear model guarantees a unique global optimal solution, thereby eliminating the possibility of convergence to local optima caused by sensitivity to initial values. Furthermore, the linear formulation reduces computational complexity and provides a more reliable assessment of the consistency of expert judgments.

To rank the sediment management strategies, the CoCoFISo method was employed. The combined compromise solution (CoCoSo) method is based on the simple additive weighting and exponentially weighted product methods. It provides a complete ranking of alternatives by combining additive and multiplicative aggregation mechanisms with a comparability sequence (Yazdani et al., 2019). Despite these advantages, CoCoSo has certain computational limitations. When an alternative has the worst performance for all evaluation criteria, the additive aggregation step cannot be computed because of denominator-related problems. To address these limitations, Rasoanaivo et al. (2024) proposed the CoCoFISo method. In CoCoFISo, the normalization procedure of the technique for order preference by similarity to ideal solution (Hwang and Yoon, 1981) is adopted to eliminate denominator-related problems in the CoCoSo algorithm. In addition, a revised formulation is introduced to prevent zero-denominator cases in the calculation of the k i b index, thereby improving the numerical stability of the method.

The methodological framework adopted in this study consists of a preparation stage, two main stages, and a validation stage, as illustrated in Figure 2. The preparation stage includes problem definition and a literature review to establish the decision framework. The first main stage (Stage 1) determines the importance of the evaluation criteria using the linear BWM. The second main stage (Stage 2) prioritizes the sediment management strategies using the CoCoFISo method based on the obtained criterion weights. Finally, the stability of the results is examined through sensitivity analysis, and the results are further compared with other MCDM methods.

The detailed procedures of Stage 1 are presented in the following steps.

Step 1.1. Identification of the evaluation criteria. A comprehensive set of decision criteria C 1 , C 2 , . . . , C n is established based on literature reviews, expert opinions, or the specific requirements of the research problem. A structured survey is then designed to collect evaluations from the experts.

Step 1.2. Selection of the best and the worst criteria. Based on the established criteria set, the expert identifies the most important criterion as the best (C B) and the least important/preferable criterion as the worst (C W).

Step 1.3. Determination of the Best-to-Others (BO) comparison scores. The expert determines the preference of the selected best criterion (C B) over all other criteria (C j) using the standard nine-point comparison scale from 1 (equally important) to 9 (extremely more important). The resulting BO vector is given in Equation 1:where a B j indicates the preference of the best criterion B over criterion j, and a B B = 1.

A B = a B 1 , a B 2 , . . . , a B n (1)

Step 1.4. Determination of the Others-to-Worst (OW) comparison scores. The expert evaluates the preference of all other criteria (C j) over the selected worst criterion (C W) using the same 1–9 scale. The resulting OW vector is given in Equation 2:where a j W denotes the preference score of the criterion C j over the worst criterion C W, and a W W = 1.

A W = a 1 W , a 2 W , . . . , a n W T (2)

Step 1.5. Calculation of the optimal criteria weights. The final weights w 1 , w 2 , . . . , w n and the optimal consistency indicator (ξ) are determined by solving the linear programming model defined in Equations 3–5, formulated by Rezaei (2016):

subject to:

min ⁡ ξ L

w B − a B j w j ≤ ξ L , ∀ j (3)

w j − a j W w W ≤ ξ L , ∀ j (4)

∑ j w j = 1 , w j ≥ 0 , ∀ j (5)

Upon solving the optimization model, the optimal criterion weights and the optimal value of ξ are obtained. In the linear BWM formulation, ξ directly reflects the consistency of the pairwise comparisons. A value of ξ closer to zero indicates higher consistency among the expert judgments.

Based on the obtained criterion weights, the framework proceeds to Stage 2. The detailed procedures of Stage 2 are presented in the following steps.

Step 2.1. Construction of the decision matrix. The decision matrix representing the performance of strategies with respect to the criteria is presented in Table 3.

TABLE 3. Decision matrix.

Let S = S 1 , S 2 , . . . , S m denote the set of strategies and C = C 1 , C 2 , . . . , C n denote the set of criteria. The decision matrix is represented as X = x i j m x n, where x i j indicates the performance value of the strategy S i with respect to the criterion C j.

Step 2.2. Normalization of the decision matrix. The decision matrix is normalized to ensure comparability among criteria using Equation 6.

where r i j is the normalized value of x i j.

r i j = x i j ∑ i = 1 m x i j 2 (6)

Step 2.3. Computation of weighted comparability sequences. Two aggregated measures are calculated for each strategy. The weighted comparability sequence (S i) and the power-weighted comparability sequence (P i) for each strategy are calculated as Equations 7, 8: where w j is the weight assigned to j-th criterion.

S i = ∑ j = 1 n w j r i j (7)

P i = ∑ j = 1 n r i j w j (8)

Step 2.4. Calculation of the appraisal scores. The scores of the strategies are obtained through the following three aggregation strategies using Equations 9–11.

k i a = P i + S i ∑ i = 1 m P i + S i , i = 1 , 2 , … , m (9)

k i b = P i + S i 1 + P i 1 + P i + S i 1 + S i , i = 1 , 2 , … , m (10)

where 0 ≤ λ ≤ 1 and λ is equal to 0.5 by default. Thus, it is adopted to assign equal importance to additive and multiplicative aggregations. The parameter λ can be defined with different values depending on the decision context.

k i c = λ S i + 1 − λ P i λ max i S i + 1 − λ max i P i , i = 1 , 2 , … , m (11)

Step 2.5. Determination of the final score. The final score (k i for each strategy is calculated using Equation 12.

k i = k i a k i b k i c 1 3 + k i a + k i b + k i c 3 (12)

Step 2.6. Ranking of strategies. Strategies are ranked in descending order based on k i. The strategy with the highest value is considered the most preferable option.

4 Results

This section presents the results of the proposed framework following the methodological stages described in Section 3. First, the problem context and decision framework are introduced. Then, the criterion weights and strategy prioritization results are presented. Finally, the stability of the obtained rankings is examined through sensitivity and comparative analyses.

4.1 Problem definition

Sediment management decision-making in hydropower reservoirs is often constrained by the limited availability of reliable, long-term, and spatially representative hydrological and sediment data. Incomplete monitoring records can limit the characterization of sediment transport processes and the quantitative assessment of sediment management strategies. Although field observations and process-based numerical models provide essential information for sediment management, their application may be constrained when reliable, long-term, and spatially representative hydrological and sediment monitoring data are unavailable or insufficient for data-intensive analyses. Under such conditions, a structured expert-based MCDM approach can provide a systematic basis for evaluating and prioritizing sediment management strategies across multiple dimensions. In this study, data scarcity refers to the limited availability of site-specific quantitative hydrological, sediment, financial, and operational data.

In data-scarce settings, expert-based prioritization can provide an initial decision basis when consistent site-specific quantitative information is limited. Expert judgments are not intended to replace quantitative data or process-based modeling results; rather, they provide a structured basis for prioritization when such information is unavailable or insufficient. The resulting priorities can subsequently be refined as site-specific monitoring data, financial information, and process-based modeling results become available. The framework therefore addresses the prioritization of sediment management strategies under data-scarce conditions while providing a basis for the subsequent integration of site-specific quantitative information. The decision hierarchy in this study consists of four dimensions, twelve evaluation criteria, and ten sediment management strategies, as illustrated in Figure 3.

To evaluate the sustainability of sediment management strategies, twelve evaluation criteria were selected based on a comprehensive review of the literature and grouped into four sustainability dimensions: environmental, economic, technical, and social (Adeogun et al., 2018; Dahal et al., 2025; Anzani et al., 2026). The environmental dimension includes biodiversity impact, water and soil quality, and sediment connectivity, reflecting the ecological impacts of sediment management strategies. The economic dimension comprises investment cost, operation and maintenance cost, and the impact on hydropower generation, representing the financial implications of each strategy. The technical dimension consists of sediment control effectiveness, technical feasibility, and climate resilience, describing the engineering performance and long-term applicability of the alternatives. Finally, the social dimension includes social acceptance, stakeholder participation, and institutional feasibility, reflecting the societal and governance aspects of sediment management.

The sediment management strategies considered in this study are summarized in Table 4. The selected strategies represent a broad range of watershed-based and reservoir-based sediment management strategies, including erosion control measures, sediment routing techniques, and sediment removal methods that have been widely reported in the literature. The resulting decision hierarchy and strategies define the general decision problem addressed by the proposed framework. The framework is subsequently applied to a data-scarce expert-based prioritization setting, with the BWM, and CoCoFISo, procedures described in Section 3 and the corresponding results presented in Section 4.2.

TABLE 4. Sediment management strategies considered in this study.

4.2 Findings of integrated BWM-CoCoFISo framework

The expert judgments obtained following the procedure described in Section 3.1 were used to determine the criterion weights through BWM and to construct the CoCoFISo decision matrix. Each expert independently identified the best and worst criteria and completed the corresponding Best-to-Others (BO) and Others-to-Worst (OW) comparison vectors. Since the criteria were structured hierarchically, the BWM was applied at two levels. First, each expert determined the weights of the four main dimensions. Second, separate BWM analyses were conducted for the criteria within each dimension. Thus, each expert completed five independent BWM evaluations, consisting of one evaluation for the main dimensions and four evaluations for the criteria within each dimension. Consequently, five optimization models were solved for each expert. A total of fifty BWM optimization models were therefore formulated and solved for the ten experts. The individual BO and OW comparison inputs for the four main dimensions are presented in Table 5. The corresponding criterion-level comparison inputs are provided in Table 6.

TABLE 5. Individual BWM comparison inputs for the four sustainability dimensions.

D1: environmental, D2: economic, D3: technical, and D4: social. The BO and OW vectors are reported in the order D1-D4. The ξ value represents the consistency indicator obtained from the optimization model.

TABLE 6. Individual BWM comparison inputs for the evaluation criteria.

The criterion codes follow the decision hierarchy presented in Figure 3. C1-C3: environmental, C4–C6: economic, C7-C9: technical, and C10-C12: social. BO and OW values are reported in the order of the criteria within each dimension.

The individual dimension and local criterion weights obtained from the BWM models were used to calculate the global criterion weights. For group decision making, both arithmetic- and geometric-mean aggregation have been used to combine individual weights (Mohammadi and Rezaei, 2020). In this study, the arithmetic mean was adopted as the baseline aggregation method under the equal-importance assumption for the experts, as also used in previous BWM applications (Štreimikienė et al., 2025). The influence of the aggregation method was further examined using geometric-mean aggregation in the sensitivity analysis. The resulting dimension weights, local criterion weights, global criterion weights, and criterion rankings are presented in Table 7.

TABLE 7. Final local and global criterion weights obtained using the BWM.

The results show that the technical dimension had the highest weight (0.365), followed by the economic (0.301), environmental (0.245), and social (0.089) dimensions. Among all criteria, sediment control effectiveness (C7: 0.212) received the highest global weight, indicating its dominant role in the evaluation process. The next most influential criteria were impact on hydropower generation (C6: 0.151) and sediment connectivity (C3: 0.147), ranked second and third, respectively. Following the determination of the criterion weights, the sediment management strategies were evaluated using the CoCoFISo method. First, the decision matrix was constructed based on the expert evaluations of the ten sediment management strategies with respect to the twelve sustainability criteria. The individual expert evaluations were aggregated using the arithmetic mean to obtain a single group decision matrix. The resulting aggregate decision matrix is presented in Table 8.

TABLE 8. Aggregate decision matrix used in the CoCoFISo analysis.

To assess the level of agreement among the ten experts, Kendall’s coefficient of concordance (W) was calculated separately for each evaluation criterion using the individual expert evaluation data. For each criterion, the ten sediment management strategies were ranked separately by each expert. Tied values were assigned average ranks (Amerise and Tarsitano, 2015). Kendall’s W was calculated with correction for tied ranks following the standard formulation of the statistic (Legendre, 2005). Kendall’s W ranges from 0 to 1, with higher values indicating greater agreement among experts. The criterion-specific Kendall’s W coefficients are presented in Table 9. The results indicate that the level of concordance varied across the twelve criteria, suggesting that the degree of agreement among experts differed depending on the evaluation criterion. The Kendall’s W coefficients ranged from 0.504 to 0.839. The highest concordance was observed for C5, followed by C1 and C4. In contrast, the lowest concordance was obtained for C7, followed by C10 and C11. All Kendall’s W coefficients were statistically significant (p < 0.001). This indicates statistically significant agreement among the experts for all evaluation criteria.

TABLE 9. Kendall’s coefficient of concordance analysis for the evaluation criteria.

The chi-square approximation was evaluated using n – 1 = 9 degrees of freedom, where n = 10 is the number of evaluated strategies.

The decision matrix was normalized according to the procedure described in Section 3.2. Subsequently, the normalized decision matrix and the BWM-derived criterion weights were used to calculate the weighted and power-weighted comparability sequences for each strategy. Based on these aggregation measures, three appraisal scores were obtained using different aggregation approaches. The final CoCoFISo score was then calculated by combining these appraisal scores. Finally, the sediment management strategies were ranked according to their final scores. The calculated CoCoFISo parameters, appraisal scores, final scores, and rankings are presented in Table 10. S1 ranked first, followed by S2, S4, and S3 in second, third, and fourth positions, respectively, while S10 ranked 10th. Differences among the appraisal approaches were mainly observed for the middle-ranked strategies. In particular, S5, S6, S7, S8, and S9 changed relative positions across the appraisal approaches. These strategies also had closely spaced final CoCoFISo, scores, indicating limited separation among the middle-ranked strategies.

TABLE 10. Ranking of sediment management strategies.

4.3 Sensitivity analysis

Sensitivity analysis was conducted to examine the stability of the proposed BWM-CoCoFISo framework under alternative decision assumptions. Two sensitivity analyses were performed to examine the effects of criterion weights and aggregation methods. Criterion-weight sensitivity was examined by changing one criterion weight at a time. For each criterion, its weight was increased and decreased by 30% in separate scenarios, while the weights of the other criteria were proportionally adjusted so that their sum remained equal to one. With twelve criteria, this resulted in twenty-four sensitivity scenarios. The ranking results are presented in Figure 4.

Figure 4 shows a high degree of consistency with the baseline ranking across the twenty-four scenarios. Spearman’s rho values ranged from 0.891 to 1.000, with a mean value of 0.985. The largest change occurred when the weight of sediment control effectiveness (C7) was increased by 30%, with a Spearman’s rho of 0.891. Rank changes were mainly observed among the middle-ranked strategies. Reforestation (S1) remained first, and dredging (S10) remained 10th in all scenarios. These results indicate substantial agreement with the baseline ranking, while the relative positions of some middle-ranked strategies remain sensitive to changes in criterion weights.

Sensitivity to the aggregation method for BWM-derived weights was examined by comparing arithmetic-mean and geometric-mean aggregation of the individual BWM weights. The baseline model uses the arithmetic mean to obtain the group criterion weights. The geometric mean was used as an alternative while keeping all other elements of the framework unchanged. The resulting criterion weights are presented in Figure 5.

The geometric-mean aggregation led to minor changes in the global criterion weights. The largest change occurred for C4, whose weight decreased from 0.093 to 0.075, followed by C3, whose weight increased from 0.147 to 0.156. The weights obtained using geometric-mean aggregation were then used in the CoCoFISo analysis. Compared with the baseline ranking, only S6 and S8 exchanged positions, moving from sixth and seventh to seventh and sixth, respectively, while all other strategies retained their original ranks. Spearman’s rho between the baseline and geometric-mean rankings was 0.988, indicating a very high level of agreement between the two rankings.

4.4 Comparative analysis

To further evaluate the reliability of the proposed BWM-CoCoFISo framework, a comparative analysis was conducted using the Measurement Alternatives and Ranking according to the Compromise Solution (MARCOS) method (Stević et al., 2020) and the Weighted Aggregated Sum Product Assessment (WASPAS) method (Zavadskas et al., 2012). These methods were selected because they are well-established MCDM methods employing different ranking mechanisms and have been widely applied in sustainability assessment and decision-making. The resulting rankings are presented in Figure 6.

The comparison results indicate a consistent ranking pattern among the three methods. Reforestation (S1), vegetative filter strips (S2), check dams (S4), settling basins (S5), density current venting (S9), and dredging (S10) maintained identical rankings across all methods. Minor ranking differences were observed only for the middle-ranked strategies (S3, S6, S7, and S8), reflecting differences in the aggregation mechanisms employed by the ranking methods. The comparison showed substantial agreement between the proposed BWM-CoCoFISo framework and the MARCOS and WASPAS methods, while some middle-ranked strategies differed across the methods.

5 Discussion

This section discusses the interpretation of the prioritized strategies and their implications for reservoir sediment management.

5.1 Interpretation of the prioritized strategies

The ranking results indicate that watershed-based strategies generally performed better than reservoir-based strategies. This finding is consistent with the distinction reported by Morris (2020) and Anzani et al. (2026), where watershed-based measures can reduce sediment generation, runoff-driven erosion, or sediment delivery before sediment reaches the reservoir, while reservoir-based strategies mainly address sediment after deposition or during transport.

Reforestation, vegetative filter strips, and check dams received the highest overall priority rankings. Their higher rankings may be attributed to their ability to reduce sediment generation and transport within the watershed before sediment enters the reservoir. However, the practical implementation of reforestation may depend on land availability, vegetation establishment time, and local climatic conditions. Similar observations have been reported for watershed erosion-control measures, including vegetation-based practices and sediment retention structures (Adeogun et al., 2018; Anzani et al., 2026).

Reservoir-based strategies showed intermediate rankings due to their different operational requirements and implementation conditions. Sediment bypass tunnels reduce sediment deposition by diverting sediment-laden flows around reservoirs, but their application depends on site-specific conditions (Dahal et al., 2025). Similarly, sluicing and flushing can support sediment management by routing or removing sediments within reservoirs; however, their effectiveness is influenced by reservoir operation conditions (Glas et al., 2025; Invernizzi et al., 2025). These factors may explain the lower rankings of these strategies compared with watershed-based measures.

Dredging received the lowest overall ranking. Although it can remove accumulated sediments and restore reservoir storage capacity, it addresses sedimentation after deposition rather than reducing sediment input. In addition, it requires repeated interventions for continued sediment removal (Morris, 2020; Glas et al., 2025).

5.2 Implications for reservoir management

The findings of this study have implications for sediment management planning by supporting the comparison and prioritization of different management alternatives. From a policy perspective, the results highlight the importance of considering sediment control strategies according to reservoir characteristics, management objectives, and available resources rather than relying on a single intervention type.

From a methodological perspective, sediment management decisions often involve uncertainties related to sediment sources, transport processes, and the long-term performance of control measures. In this context, expert-based multi-criteria approaches can provide a structured framework for evaluating alternative strategies when complete monitoring information is not available. Such approaches can help integrate technical knowledge with management priorities during the decision-making process.

For reservoir managers, the results indicate that preventive measures and reservoir-based strategies should be evaluated together, considering their different roles in sediment management. This approach can be particularly relevant for small and medium-sized reservoirs, where the selection of suitable strategies may depend on site conditions, operational constraints, and available implementation capacity.

5.3 Limitations and further studies

The proposed framework provides a structured basis for prioritizing sediment management strategies when site-specific quantitative information is incomplete or unavailable. Its application therefore depends on the quality and scope of the available expert knowledge. In particular, the evaluation of economic criteria is affected by the absence of site-specific financial data. The scores represent relative expert assessments rather than observed monetary values.

The composition and experience of the expert group may influence the resulting priorities. Geographic and professional backgrounds may shape how experts perceive the effectiveness of different sediment management strategies. For example, experts with greater watershed-management experience may place more emphasis on upstream measures such as reforestation and erosion control. In contrast, experts with stronger reservoir-operation or sediment-management experience may favor reservoir-based strategies such as dredging or sediment flushing. The social dimension received the lowest weight in the analysis. This low weight may also have been influenced by the composition and expertise of the expert group, which included experts from engineering, hydrology, sediment management, and MCDM fields. These considerations highlight the context-dependent nature of expert-based prioritization.

The present study provides a general prioritization framework rather than a site-specific reservoir assessment. The results therefore do not account for local sediment characteristics, hydrological conditions, reservoir geometry, financial conditions, or institutional constraints. The results also show limited separation among several middle-ranked strategies. Small differences in their final scores indicate that their relative positions may be sensitive to changes in expert judgments, criterion weights, and aggregation assumptions.

Future studies should apply the framework to selected reservoirs using site-specific expert judgments and observed data. Financial data, monitoring data, and operational information can be incorporated to refine the criteria assessments. Different expert-weighting approaches that account for differences in professional experience and expertise can also be investigated. Broader stakeholder participation, including local communities and water users, can provide additional perspectives on the social dimension. Probabilistic, interval-based, or fuzzy MCDM approaches can be used to represent uncertainty in expert assessments and criterion weights. Scenario-based analyses can examine how changes in reservoir conditions, management objectives, and data availability affect the resulting priorities. A further direction is the integration of expert-based MCDM with quantitative sediment and hydrological information. Monitoring data can be used to refine criteria related to sediment connectivity, sediment control effectiveness, and hydropower impacts. Process-based numerical models can provide site-specific estimates of sediment transport, deposition, and management performance. These quantitative outputs can be incorporated into the MCDM framework together with expert judgments. Such integration can complement expert-based assessments with site-specific quantitative evidence and support adaptive sediment management planning as new information becomes available.

6 Conclusion

This study proposed an integrated BWM-CoCoFISo framework for prioritizing sediment management strategies in hydropower reservoirs under data-scarce conditions. This approach enabled the evaluation of sediment management strategies by considering technical, economic, environmental, and social aspects within a unified decision framework. Under the evaluated criteria and expert judgments, watershed-based strategies, particularly reforestation, vegetative filter strips, and check dams, received the most favorable overall rankings. The sensitivity and comparative analyses showed substantial agreement in the resulting rankings, while some middle-ranked strategies remained sensitive to changes in decision assumptions. The study shows that expert-based MCDM approaches can provide structured decision-support frameworks for sediment management planning.

Despite its contributions, this study has several limitations. First, the evaluation process was based on expert judgments because of the limited availability of quantitative sediment monitoring data. Although expert-based assessments provide useful information under data-scarce conditions, the results may still reflect the experience and perspectives of the selected experts. Second, the evaluation was based on a predefined set of criteria and strategies. Future applications should apply the framework to specific reservoirs using site-specific monitoring, financial, and operational data. Process-based numerical model outputs can be incorporated to complement expert judgments and refine the prioritization of sediment management strategies. Broader stakeholder participation can provide additional perspectives, particularly for the social dimension.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

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.

Footnotes

Edited by: Babak Vaheddoost, Bursa Technical University, Türkiye

Reviewed by: Egemen Aras, Bursa Technical University, Türkiye

Andi Muhammad Yasser Hakim, National Research and Innovation Agency (BRIN), Indonesia