Prioritizing sub-watersheds for soil and water resource conservation using multi-analytical approaches: a study of Shilabati River Basin, W.B., India

Watershed prioritization has become a pivotal framework for identifying regions susceptible to environmental degradation and formulating effective conservation strategies. This process facilitates the efficient allocation of limited resources to address soil erosion, sedimentation, and hydrological challenges, which are critical for sustainable natural resource management (Shrimali et al. 2001; Vittala et al. 2004). The strategic significance of this methodology lies in its ability to integrate geomorphological, hydrological, and land use parameters to prioritize areas for soil and water conservation interventions (Mishra & Nagarajan 2010). Effective management of watersheds is essential for mitigating environmental degradation, promoting balanced development, and ensuring long-term ecosystem health (Kudnar & Rajasekhar 2020).

The morphometry of a watershed is an essential component of hydrological analysis, as it provides a quantitative understanding of topographic relief, drainage patterns, and sediment transport dynamics (Das & Das 2024a, 2024b). Morphometric parameters, such as drainage density, bifurcation ratio, and stream frequency, play a critical role in assessing flood susceptibility and soil erosion potential (Sreedevi et al. 2009; Maddamsetty & Pawar 2023). Recent advancements in remote sensing (RS) and geographic information systems (GIS) have revolutionized watershed analysis by enabling the rapid and accurate assessment of morphometric attributes across diverse landscapes (Grohmann 2004; Kouli et al. 2007; Kaya et al. 2019; Yadav et al. 2024). These tools have become indispensable for delineating watershed boundaries, extracting drainage networks, and prioritizing sub-watersheds based on their vulnerability to environmental stress (Patel et al. 2015; Dey et al. 2025).

The Shilabati River Basin (SRB), located in West Bengal, India, exemplifies a region facing acute environmental and socio-economic challenges. Spanning approximately 3,500 km², the basin is characterized by a complex geomorphology, varying from the undulating terrain of the Chotanagpur Plateau to the alluvial plains of its lower reaches (Shit & Maiti 2012). The basin experiences significant soil erosion, resulting in an annual sediment load of approximately 2.5 million tons, which adversely affects water quality and agricultural productivity. Agricultural runoff exacerbates water pollution, with nearly 70% of the river exceeding permissible pollution thresholds. Frequent flooding events further compound the region's vulnerabilities, displacing millions of people and causing extensive damage to infrastructure and livelihoods.

Despite the pressing environmental concerns in the SRB, existing studies have largely focused on singular approaches to watershed prioritization, often relying on standard geomorphological or hydrological assessments (Godif & Manjunatha 2022). These studies fail to capture the intricate interplay of factors influencing soil erosion and flood susceptibility, particularly in ungauged sub-basins where data scarcity limits the application of traditional models (Das & Das 2024a, 2024b). Hybrid methodologies, such as the Nash-GIUH model, have demonstrated potential for addressing these challenges by integrating geomorphological and hydrological parameters to enhance flood risk assessment (Das & Das 2024a, 2024b). However, the application of such comprehensive approaches remains limited in the context of the SRB.

This study addresses these gaps by adopting a multi-analytical framework that combines morphometric analysis (MA), principal component analysis (PCA), the sub-watershed prioritization tool (SWPT), land use/land cover (LULC) analysis, and hypsometric analysis (HA). By integrating these methodologies, the research aims to provide a holistic understanding of the hydro-geomorphological dynamics of the SRB and prioritize its sub-watersheds based on their susceptibility to erosion and flood risks. MA serves as the foundation for quantifying watershed characteristics, while PCA reduces data redundancy and identifies critical parameters influencing prioritization (Shekar et al. 2023). The SWPT tool automates the prioritization process by synthesizing morphometric and hydrological data, ensuring accuracy and efficiency (Rahmati et al. 2019). LULC analysis captures the spatial dynamics of land use changes and their implications for soil and water conservation (Biswas & Chakraborty 2016), while HA provides insights into the erosion stages and landscape evolution of the sub-watersheds (Pike & Wilson 1971; Willgoose & Hancock 1998).

The novelty of this study lies in its synergistic integration of multiple analytical approaches, which enables a comprehensive and data-driven prioritization of sub-watersheds in the SRB. The study also emphasizes the importance of incorporating real-time climate change scenarios, socio-economic factors, and community participation into watershed management frameworks to ensure long-term sustainability. By leveraging the strengths of RS and GIS technologies, the research provides a replicable model for prioritizing sub-watersheds in other regions facing similar challenges.

This study seeks to revolutionize the watershed prioritization process by combining advanced analytical techniques to address the unique environmental challenges of the SRB. The findings are expected to contribute significantly to the field of watershed management, offering practical insights for policymakers, planners, and conservationists in their efforts to mitigate environmental degradation and promote sustainable resource management.

The SRB is a distinct hydrological unit, functioning as a sub-basin of the Rupnarayan River, which ultimately drains into the Ganges River (Shit & Maiti 2012). The SRB's geographical characteristics are marked by undulating topography, transitioning to the Chotanagpur plateau fringe in the lower basin, with elevations ranging from −4 m to 230 m. The basin's terrain is complex, with significant spatial differences in meteorological and hydrological elements. The Shilabati River flows eastward, and its left tributaries are predominantly first-order streams trending southeast. The SRB's drainage pattern ranges from dendritic to sub-dendritic, influenced by lithology and regional land slope.

The SRB is characterized by a sub-humid tropical climate, with average annual rainfall ranging from 100 to 150 cm and typical high temperatures ranging from 32° to 37 °C. The area is covered with tropical dry deciduous forests, and flood-related economic damage and destruction are prevalent, notably in the regions of Ghatal, Banka, and Khirpai (Mukherjee et al. 2024). Land degradation varies in intensity, impacting isolated lateritic tracts throughout the study area. The geological composition varies, with gneiss and schist dominating the upper basin and younger alluvium prevalent in the lower basin (Bera et al. 2020). Gongoni Danga, located on the right bank of the Shilai River near Garbeta in West Medinipur, exhibits typical lateritic mesoscale Badlands (Bandyopadhyay 1988). Population pressure, LULC changes, and development initiatives are exerting major effects on the basin's geoenvironmental landscape. More than a million people are supported by agriculture in the middle and lower parts of the watershed's floodplain, which comprises 23 CD blocks spanning four districts.

The study utilized a multi-analytical framework to prioritize 23 sub-watersheds within the SRB based on morphometric, hypsometric, hydrological, and land use characteristics. Each method addressed specific objectives, ensuring a holistic evaluation of the basin's conservation needs.

PCA was employed in this study to reduce data redundancy and identify the major factors affecting watershed processes in the SRB. By analyzing a set of correlated variables, including stream length, slope gradient, and drainage intensity, PCA identified a smaller set of uncorrelated variables, called principal components. Using XLSTAT software, 12 morphometric parameters were reduced to four key components, with each component strongly linked to one highly correlated parameter (Shekar et al. 2023). This statistical approach eliminated redundancy while retaining critical information for prioritization. The PCA results were then used to rank sub-watersheds based on relief and linear parameters, with the highest scores ranked first (Supplementary Figure 1). Conversely, sub-watersheds with the lowest shape parameter scores were ranked first. Finally, they were grouped into high, medium, and low susceptibility categories based on their C_p results, with those having higher PCA scores, indicating higher soil erosion risk, prioritized for conservation.

SWPT offers efficient and accessible analysis of data, avoiding shortcomings of the traditional methods. Sub-watersheds are ranked by C_p values, with lower values indicating higher susceptibility (Rahmati et al. 2019). This data can be used with machine learning algorithms for prioritization and sustainable management practices.

This study investigates the impact of LULC on the soil and water resources of the SRB and proposes prioritization strategies. LULC analysis was conducted using Landsat-9 imagery, and processed through a supervised classification method to derive six distinct land use classes: agricultural land, barren land, plantation, mixed grazing land, built-up areas, and water bodies. The analysis revealed that the region is predominantly forested, followed by agricultural land. Sub-watersheds dominated by built-up areas were ranked as a higher priority due to increased runoff and reduced infiltration, while forested areas received lower ranks for their conservation benefits. The analysis allocated a rank of 1 to sub-watersheds with the highest built-up area and similarly ranked other parameters, with the highest values receiving rank 1 for built-up areas and the lowest values receiving rank 1 for trees, crops, and rangeland parameters (Biswas & Chakraborty 2016). Compound parameter (C_p) values obtained by the above process were used to categorize sub-watersheds into three susceptibility groups: high, medium, and low.

The elevation-relief ratio method was applied to compute the hypsometric integral (HI) values, derived from SRTM-DEM data, to determine the stage of landform evolution (Pike & Wilson 1971). Using the hypsometric curve, the watershed is categorized into three phases: old, mature, and young. The maximum HI value is ranked as 1, and so on. Sub-watersheds were classified as: high, medium, and low categories based on the C_p (compound parameter) values.

The research employed a robust validation process to assess the accuracy of predicted risk zones. Sampling sites were systematically selected from the entire basin, informed by historical flood data and concurrent erosivity evidence. The prioritization results were then validated using receiver operating characteristic (ROC) curves, comparing field-verified 2024 flood data with predicted priority zones. The area under the curve (AUC) score of 0.797 confirmed the reliability of the framework. Furthermore, a sensitivity analysis was conducted to test the influence of parameter weights, ensuring robustness and minimizing biases in the prioritization process.

TPR signifies correctly predicted attribute descriptions, while FPR indicates the percentage of misclassified positive cases (middle/high attributes). The value of AUC lies between 0 and 1. Similarly, the values of (0.9–1), (0.8–0.9), (0.7–0.8), (0.6–0.7), and (0.5–0.6) AUC represent perfect, very good, good, moderate, and weak respectively for model validation.

The multi-method approach employed in this study effectively addressed the research objectives. MA and PCA provided quantitative insights into erosion and hydrological dynamics, identifying high-risk sub-watersheds. The SWPT automated prioritization, enhancing efficiency in conservation planning. LULC analysis and HA offered valuable perspectives on human-environment interactions and landscape evolution. Finally, validation of the results ensured credibility, bridging field data with analytical outputs. This integrated methodology facilitated a comprehensive understanding of the SRB's dynamics, providing a robust foundation for subsequent analysis and interpretation.

In the present research work, MA was performed on linear, areal, relief, and shape parameters using mathematical modelling.

SRB was assessed using Strahler's (1964) stream ordering methodology, classifying it as a fifth-order stream divided into 23 sub-watersheds. Stream order, a primary stage in quantitative drainage watershed assessment, reveals the hierarchical structure of the drainage network. The SRB's stream ordering follows Horton's (1945) laws of stream numbers, where the stream number (Nu) represents the total number of streams in each order. The study found that Salbani has the maximum stream number (76), while Daspur-I has the minimum (7) (Table 2). Watersheds with more streams experience higher runoff and peak flow rates. The sub-watersheds exhibit varying stream numbers and lengths, with Salbani having the longest streams and Daspur-I the shortest (Figure 3). The bifurcation ratio (R_b), a relief and dissection index, varies across regions, with Garbeta-III exhibiting the highest Rb, indicating a highly complex and branched stream network. In contrast, Daspur-II and Lakshmisagar have the lowest Rb, indicating a simpler and less hierarchical stream network (Table 2). High-priority sub-watersheds, such as Salbani, exhibit high bifurcation ratios (R_b = 4.67) and stream frequencies (F_s = 5.91 km²), reflecting a branched drainage network and rapid runoff potential. These characteristics indicate high vulnerability to soil erosion and sediment transport. Conversely, low-priority sub-watersheds, such as Daspur II (R_b = 2.13, F_s = 2.71 km²), are less prone to hydrological stress. Practical management interventions include checking dams and vegetative barriers in high R_b areas to reduce peak discharge and sediment transport.

The areal aspect of the SRB watersheds was analysed using various parameters, including the constant of channel maintenance (C_cm), drainage density (D_d), drainage intensity (D_i), drainage texture (D_t), length of overland flow (L_o), and texture ratio (T_r). The C_cm, defined as the inverse of D_d (Strahler 1964), varies significantly within the studied watersheds, with Keshpur exhibiting the largest C_cm and Daspur-I the smallest, indicating a more dispersed drainage network in Keshpur compared to Daspur-I (Table 3). Drainage density (D_d), the ratio of the total length of streams of all orders to the watershed area (Horton 1945), ranged from 0.42 km/km² in Daspur II to 2.21 km/km² in Salbani. Higher D_d values, such as in Salbani, indicate low infiltration capacity and high surface runoff, necessitating rainwater harvesting structures. Conversely, low D_d areas, like Daspur-I, are suitable for groundwater recharge zones. Drainage intensity (D_i), the ratio of stream frequency (F_s) to D_d (Faniran 1968), is highest in Barisol, suggesting a higher potential for flooding and erosion due to a higher stream frequency relative to its drainage density. The texture ratio (T_r), which reflects the overall drainage network texture, is highest in Salbani and lowest in Chandrakona-II, highlighting variations in channel density and spacing across the watersheds.

The relief aspect of the SRB watersheds was analysed using various parameters, including relative relief (H), drainage density (D_d), ruggedness number (Rn), dissection index, relief ratio (Rh), and gradient ratio (Gr). Relative relief, which measures the difference between the maximum and minimum elevations within a watershed (Schumm 1956), is highest in Khatra and lowest in Daspur-II (Table 3). Watersheds with higher relative relief, such as Khatra, exhibit greater runoff potential. The ruggedness number, the product of maximum watershed relief and drainage density (Strahler 1964), indicates a more structurally complex and erosion-prone terrain. Khatra, with the highest Rn, suggests a highly complex terrain susceptible to erosion, whereas Daspur-I, with the lowest Rn, indicates a less complex terrain with lower erosion potential. Sub-watersheds like Garbeta III show high ruggedness numbers (Rn = 4.11) and relief ratios (Rh = 0.07), indicating steep, erosion-prone terrain. The relief ratio, which measures the ratio of a watershed's total relief to its longest dimension parallel to the main drainage line (Schumm 1956), is highest in Barisol and lowest in Salbani. Higher Rh values indicate increased erosion intensity (Aher et al. 2014), making sub-watersheds with high Rh values more susceptible to erosion. Slope stabilization techniques such as contour bunding and afforestation are recommended for these areas.

The circulatory ratio (R_c) measures a watershed's circularity, with higher values indicating a greater flood hazard (Miller 1953). The study found that Lakshmisagar has the highest R_c (more susceptible to flooding), while Chandrakona – I have the lowest R_c (lower flood risk). The elongation ratio (R_e), which compares the diameter of a circle with the same area as the watershed to its maximum length (Schumm 1956), categorizes watersheds into various shapes. Lakshmisagar has a higher R_e, indicating a more circular shape, whereas Chandrakona – II has a lower R_e, indicating a more elongated shape (Table 3). The form factor (F_f), the ratio of watershed area to the square of its length, indicates potential peak flow. A higher F_f means quicker peak stream flow (Horton 1945); Daspur – I has the highest F_f, suggesting quicker peak flows, while Salbani has the lowest F_f. The compactness coefficient (C_c), the ratio of the watershed's perimeter to the circumference of an equivalent circular area (Horton 1945), indicates erosion potential. Chandrakona – I have the highest C_c, suggesting higher erosion potential, while Lakshmisagar, with the lowest C_c, indicates greater elongation and reduced erosion risk. SRB's landscapes with values close to 0 for F_f, R_e, R_c are more elongated, while those close to 1 are more circular (Miller 1953; Schumm 1956; Strahler 1964). Circular terrains have low infiltration capacity, high runoff, and are susceptible to soil erosion when compared to elongated terrains. The compound parameters (C_p) result was calculated for sub-watershed prioritization, as represented in Table 4. The sub-watersheds were then divided into three categories: low, medium, and high. Ghatal, Keshpur, Chandrakona II, Salbani, Garbeta – III, Vishnupur, Amdangra, Khatra, and Puncha are high-priority sub-watersheds, Daspur – I, Chandrakona Town, Chandrakona – I, Arenga, Barisol, Simplapal, Taldangra, Onda, Indpur and Hirabandh are medium-priority sub-watersheds, and Daspur-II, Garbeta-I, Garbeta-II, and Lakshmisagar are low-priority sub-watersheds. There has been a direct correlation between shape aspects and flood potentiality, as higher priority zones are associated with increased erodibility and vice versa. Therefore, in high-priority sub-watersheds, appropriate practices should be adopted for soil and water conservation (Das & Bandyopadhyay 2015).

PCA was employed to prioritize sub-watersheds in the SRB based on their susceptibility to flooding. Initially, 25 morphometric parameters were reduced to five key components using XLSTAT software, and then 12 morphometric parameters were reduced to four principal components, explaining 67% of the total variance (Table 5). The PCA results revealed strong correlations between morphometric parameters, indicating their significant role in flood susceptibility. Specifically, parameters such as basin length (L), dissection index (D_i), relief ratio (R_r), mean stream length ratio (R_l), and texture ratio (R_t) were found to influence flood susceptibility. High-priority sub-watersheds, such as Chandrakona II, exhibited strong correlations with parameters such as drainage density (D_d) and form factor (R_f). A Pearson correlation coefficient (r = 0.82, p < 0.01) confirmed the statistical significance of relationships between stream order (u) and stream length (L_u), indicating the influence of topographic relief on hydrological responses. The basin length and dissection index were strongly correlated (>0.64) with area and perimeter, respectively, suggesting that larger and more dissected sub-watersheds are more prone to flooding. Similarly, the mean stream length ratio, relief ratio, and texture ratio were correlated with basin length, drainage stream frequency, and form factor, indicating that longer streams and steeper slopes increase flood susceptibility. The PCA-derived parameters enabled the classification of sub-watersheds into three groups: high (C_p ≤ 8), medium (C_p 8–15), and low (C_p ≥ 15) susceptibility. Sub-watersheds with the lowest compound parameter (C_p) values were ranked highest for susceptibility, while those with higher C_p values ranked lower (Aher et al. 2014). The first two factors captured 60% of the variability in data, enabling efficient prioritization and targeted interventions for better watershed management. The management implications of this study highlight the importance of targeted interventions, such as promoting erosion-resistant vegetation in sub-watersheds with high R_f values.

This study evaluated flood susceptibility in the SRB by analyzing hydro-geomorphologic aspects, specifically examining morphometric parameters to understand their impact on flood patterns and soil erosion. The automated SWPT was employed to prioritize 23 sub-watersheds based on their erosion susceptibility. Figure 4 illustrates the correlation matrix derived from the auto WSA, highlighting the relationships between various morphometric characteristics and the hydrological behaviour of the sub-watersheds. The analysis revealed complex relationships between morphometric parameters and flood susceptibility. Stream frequency (F_s) is positively linked to form factor (R_f), elongation ratio (R_e), drainage density (D_d), and texture ratio (R_t), indicating a relationship between stream number and watershed morphology. However, STI decreases with increasing F_s, R_f, R_e, D_d, and R_t, indicating lower sediment transport efficiency in areas with higher stream frequency and drainage density. In comparison to LULC analysis, it is evident that sub-watersheds with higher proportions of dense forests, plantations, and mixed grazing lands have lower flood susceptibility, whereas those with extensive built-up areas and barren lands are more prone to flooding (Table 6).

The SWPT integrated morphometric and hydrological indices to automate prioritization. Sub-watersheds like Chandrakona I, with a high sediment transport index (STI = 0.84), were identified as erosion hotspots. Practical recommendations include sediment retention structures and buffer zones along riparian areas. The automated prioritization of sub-watersheds used constant compound parameter values (C_p), as detailed in Table 7. SWPT-derived C_p values have an inverse relationship with erodibility. Watersheds with the highest C_p are of low priority, while those with the lowest C_p are of high priority. Lower C_p values signify higher erosion risk and hence, higher priority. Garbeta-III, with a C_p value of −301.75, was ranked as the highest priority sub-watershed due to its high erosion risk. In contrast, Daspur-I, with a C_p value of 36.28, was determined to have the lowest priority due to its lower erosion susceptibility. The SWPT analysis-based prioritization is presented in Table 7.

An HA was conducted to evaluate the erosion and deposition stages of the SRB. The HI values for all sub-watersheds were calculated using the elevation-relief method, ranging from 0.48 to 0.54. This range suggests that most sub-watersheds are relatively mature, with stabilizing erosional processes. Notably, the HI values indicate varying degrees of erosion susceptibility, with higher values signifying increased susceptibility to erosion and sedimentation. Specifically, sub-watersheds with HI values closer to 0.54 are more prone to erosion, whereas those with lower HI values (near 0.48) exhibit reduced erosion susceptibility.

Different methods may yield varying priority levels in the sub-watershed priority assessment in SRB. However, if three out of the five methods consistently assign a similar priority to a particular sub-watershed, it is reasonable to consider that priority. For example, if Khatra is consistently rated as a high priority by three methods, it can be reliably deemed a high-priority area. By identifying common results and considering consistency among the methods, a more reliable assessment of priority sub-watersheds can be achieved. The results are shown in Table 8 and the final common priority rank (CPR) map of SRB is shown in Figure 7(f).

To ensure long-term sustainability, ecosystem health, and community well-being, the prioritization strategy plan targets high-, medium-, and low-priority sub-watersheds with specific actions and policies. It presents a methodical strategy that integrates ecosystem-based management, climate resilience, and sustainable development to rank sub-watersheds in the SRB in order of importance for soil and water conservation (Table 9).

Small check dams, mulching, and the planting of erosion-resistant grasses are necessary in the Gangani area to prevent the significant erosion and soil degradation caused by the River Shilabati. Flood management in the Ghatal region required geomorphological and hydrological techniques with strong regional participation, moving beyond reliance on the eagerly anticipated Ghatal Master Plan. By addressing both short- and long-term issues with targeted, implementable solutions combined with comprehensive, root-cause-focused methodologies, the Ghatal Master Plan integration with river basin planning will improve flood resilience.

Effective river basin management and soil and water conservation depend on understanding the interconnectedness of natural systems, human activities, and sustainability considerations. The SRB is vulnerable to soil erosion, groundwater depletion, and monsoonal flooding. To combat these challenges, strategic prioritization of sub-watersheds for targeted interventions is essential. While morphometric analyses and LULC studies have been explored in isolation, few studies have combined advanced techniques such as RS-GIS, PCA, and HA in a cohesive framework to assess the geomorphological, hydrological, and environmental characteristics of watersheds. The prioritization and ranking of 23 sub-watersheds were done by assigning ranks to individual indicators (MA, Arc-SWPT, PCA, LULC, and HA) and deriving a C_p value from it. The ranking of the sub-watersheds was truly determined by the relationship with erodibility, and flood probabilities, by assigning the highest priority/rank based on the highest C_p value in the case of HA and the lowest value in the case of SWPT, MA, PCA, and LULC analysis, thereby ensuring a data-driven approach to watershed management. The MA, PCA, HA, and ARC-SWPT techniques were employed to assess basin-wide soil and water conservation, with a focus on flood susceptibility, erosion, sedimentation, and resistivity to prioritize watersheds (Figure 7). While LULC plays a crucial role that basin hydro geomorphic processes influenced by humanistic perspectives, too. Prioritizing watersheds for soil and water resource conservation is a multidimensional approach that involves a comprehensive analysis of various morphometric characteristics. In this study, 12 essential morphometric parameters were examined to propose priorities for conservation efforts, considering both flood potential and erosivity for conceptual clarity. Each parameter was evaluated, with rankings and scores assigned based on their relationships between linear and relief parameters and soil erosion susceptibility. Linear parameters (drainage density, stream frequency, bifurcation ratio, drainage texture, and length of overland flow) demonstrated a positive correlation with soil erodibility and increased flood potential. During the sub-watershed prioritization process, the sub-watersheds were ranked according to their linear parameter values, with the highest values receiving a rank of 1, the second-highest a rank of 2, and so forth, with the lowest values ranked last. In contrast, shape parameters displayed a negative correlation with soil erosion. Therefore, sub-watershed prioritization was based on high linear parameter values indicating high erosion risk and high shape parameter values indicating low erosion risk. The results generated from ArcSWPT provide an auto-generated prioritization ranking. PCA identified five key components from the initial 25 morphometric parameters, collectively explaining 60% of the dataset's variability. Using a maximum likelihood classifier, a supervised image classification approach was applied for LULC analysis to investigate the effects of LULC on soil and water resources in the SRB. The results guided the identification of priority areas for sustainable management and conservation. HA provides valuable insights into a watershed's topography, influencing hydrology, ecology, and land use. Examining elevation distribution informs hydrological processes, erosion potential, and land suitability, essential for effective watershed management. The current research utilizes hydro-geomorphological approaches to evaluate flood susceptibility within the SRB. The Shilabati River's eastward flow generates unique hydrological patterns, characterized by left tributaries being susceptible to flooding, right tributaries experiencing inundation and significant runoff, and sediment deposition following the flow regime. Consequently, this underscores the need for a revised flood management strategy that prioritizes a comprehensive understanding of the interconnected hydrology between both riverbanks. The methods employed in this research provided an integrated framework for watershed prioritization, aligning with the objectives of this research. However, it has been acknowledged that more advanced methodologies, such as multi-criteria decision-making (MCDM) techniques like analytic hierarchy process (AHP), Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), QGIS Soil and Water Assessment Tool (QSWAT) analysis, synthetic unit hydrograph (SUH), Geomorphic Instantaneous Unit Hydrograph (GIUH), Instantaneous Unit Hydrograph (IUH), and semi-empirical methods like Revised Universal Soil Loss Equation (RUSLE), and machine learning approaches, could offer improved precision, especially for predictive studies (Bhagwat et al. 2011; Akay & Koçyigit 2020; Singh & Kansal 2023; Das & Das 2024a, 2024b). These methods have been shown to excel in flood hazard assessment and watershed management through pixel-to-pixel raster-based analysis, providing more refined spatial decision-making capabilities. Although this study did not employ MCDM or machine learning techniques due to its specific focus on watershed prioritization rather than flood prediction, these methods are highlighted as a potential direction for future research. The findings emphasize the need for holistic, data-driven watershed management in the SRB, integrating hydrological, geomorphological, and ecological perspectives to mitigate flood risks and promote sustainable land use.

This study highlights the critical importance of prioritizing sub-watersheds within the SRB for efficient management and conservation of land and water resources. By integrating advanced techniques such as RS and GIS, MA, PCA, HA, and LULC analysis, we provided a comprehensive evaluation of the basin's geomorphological, geohydrological, and geoenvironmental characteristics. Apart from these techniques, there are other methods such as the SUH and GIUH model, hazard degree, El-Shamy's approach, AHP, TOPSIS, and other machine learning techniques for basin prioritization that can be used further for future research. Our multi-analytical approach allowed for the precise identification and extraction of drainage basins, enhancing the accuracy of our prioritization process. The findings revealed significant variations in priority across the sub-watersheds. High priority was consistently assigned to Ghatal, Chandrakona I and II, Salbani, Garbeta I and III, Vishnupur, and Amdangra based on the combined results of all analytical methods used. Medium priority was given to sub-watersheds such as Daspur – I, Keshpur, Chandrakona Town, Garbeta – II, Sarenga, Barisol, Simplapal, Onda, Taldangra, Indpur, Hirabandh, and Puncha. Meanwhile, low priority was assigned to Daspur – II, Garbeta – I, Lakshmisagar, and Khatra. The drawbacks of the research work are the scarce accessibility of data, which confines the application of deep learning models. Future research work should expand the data collection process to include more micro-level watersheds, enabling the application of deep learning models for more effective flood management strategies. Ultimately, the aforementioned study presents insightful information for environmental researchers and land use planners. Sub-watershed prioritization allows stakeholders to develop programs for soil and water conservation, which ensures sustainable development and protects the watershed from erosion and flooding in the SRB and any other watershed.

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

The authors declare there is no conflict.