Utilizing statistical and MCDM techniques in indexing morphometric parameters towards improved watershed management in the Nandhour-Kalish drainage system

Watersheds play a crucial role in supporting ecosystems, agriculture, and anthropogeny. This makes the prioritizing of watersheds an essential undertaking for the effective and sustainable management of drainage systems and the allocation of their natural resources (Singh et al. 2021). This prioritization becomes vital for nations experiencing challenges like water scarcity, climate change impacts, and escalating demand for natural resources. The significance of prioritizing watersheds lies in the efficient allocation of limited water resources, concentrating efforts on areas pivotal to overall watershed health and functionality (Randhir et al. 2001; Alilou et al. 2019).

The hydrological behaviour of a river basin is intricately associated with the physiographic characteristics of the drainage basin, including factors like its size, shape, length, slope, and drainage density, among several others (Shekar et al. 2023). Notably, morphometric analysis is an essential process for gaining critical insights into the dynamics of a watershed. Morphometry also plays a vital role in comprehending how drainage characteristics interact with various hydrological processes, such as infiltration, runoff, erosion, sediment transport. Analyzing the morphometrics of river basins is a fundamental approach to studying how these basins respond to their topographical features (Sharma et al. 2023). Statistical methods are commonly employed in hydrological research to investigate data for understanding the relationship between different hydro-meteorological parameters (Mahmoodzada et al. 2023). Watershed prioritization involves the systematic ranking of sub-watersheds within a larger drainage system depending on the need for prioritization like implementation of soil and water conservation measures, etc. (Gaikwad & Bhagat 2017). Various methods utilize specific criteria to determine the priority of different watersheds, considering factors such as hydrological characteristics, land use, soil erosion potential, water quality, biodiversity, and socio-economic aspects (Siddiqui et al. 2020). In the past, a multitude of studies have investigated morphometry-based prioritization of watersheds using Remote Sensing (RS), Geographical Information System (GIS), and topographic data (Deal et al. 1997; Chopra et al. 2005; Mishra et al. 2017; Arshed et al. 2023; Shekar et al. 2023).

In India, the escalating trends of population growth, urbanization, industrial expansion, agricultural activities, etc. (Varade et al. 2023) have resulted in a surging demand for water resources (Tiwari & Kushwaha 2021). In India, the world's most populated country, the state of Uttarakhand holds immense significance, particularly due to its geography hosting some of the major glaciers in the Indian Himalayas, which contribute to some of the major river channels in the country. Studies have indicated the declining status of the water resources in the state of Uttarakhand. Kumar et al. (2021) assessed the soil erosion status and watershed prioritization in the Nainital district of Uttarakhand, India, based on the Revised Universal Soil Loss Equation integrated with RS data. Sarkar et al. (2022) studied the Pindar River watershed based on morphometric analysis of 13 parameters using various Multi-Criteria Decision-Making (MCDM) techniques. According to Singh et al. 2012, the Kumaun and Garhwal regions of the state are facing critical challenges associated with drinking water supply in quality and quantity. Singh & Kansal (2023) evaluated various morphometric, semi-empirical methods and hydrologic methods for extreme events of soil erosion in the Alaknanda basin of Uttarakhand. They investigated 13 morphometric parameters for 12 sub-watersheds in the region. The state of Uttarakhand is historically known to be prone to extreme rainfall events (Singh et al. 2023b).

Evidently, as advocated in previous studies, with the inclusion of statistical methods and MCDM approaches within a GIS framework, the process of prioritization has become increasingly simplified. Such platforms now offer and account for the ascendancy of one morphometric parameter over the other depending upon the characteristics of the study area (Ahmad 2018; Ahmad & Verma 2018; Sangma & Guru 2020; Sarkar et al. 2022). Many studies have demonstrated the competence of statistical approaches like correlation analysis, Principal Component Analysis (PCA), etc., when used in the context of parameter selection (Siddiqui et al. 2020; Das et al. 2021). Traditional methods for assessing morphological parameters such as field reconnaissance and hydrological surveys/channel measurements are often time-consuming, costly, and manpower-intensive. In contrast, the emergence of RS and GIS technologies has significantly simplified, reduced costs, and accelerated these assessments, making them substantially more efficient and accessible (Aher et al. 2014). The RS data provide means to evaluate the land use surrounding the hydrological channels and satellite-based elevation products to simplify the understanding of the drainage channels and flow characteristics (Chitra et al. 2011; Kumar et al. 2021). GIS tools based on such RS products have thus been widely used with MCDM techniques for watershed management and prioritization (Chopra et al. 2005; Chowdary et al. 2013; Ahmad & Verma 2018).

The assignment of comparative weights to input parameters for watershed prioritization has been carried out using various MCDM techniques (Jaiswal et al. 2015). These techniques include methods such as the Simple Additive Weighting (SAW), VIekriterijumsko KOmpromisno Rangiranje (VIKOR), Best–Worst Method (BWM), Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), Game Theory Algorithm (GTA), and many more (Sarkar et al. 2022; Nasiri Khiavi et al. 2023). This diverse set of MCDM techniques ensures a comprehensive and well-informed approach to prioritizing watersheds (Akdogan & Guven 2023). The integration of the diverse criteria establishes a robust framework for watershed prioritization, contributing to holistic watershed management practices. This approach ensures a thorough consideration of various factors, leading to effective and informed decision-making (Chowdary et al. 2013). However, due to the availability of a large number of morphometric parameters in decision-making, the computational procedure becomes complex and expensive. Subsequently, it is necessary to initially determine the most influencing parameters for a target drainage system and then apply a suitable MCDM technique for addressing the relative importance among the identified parameters. In this context, the use of statistical methods can prove advantageous, especially in input data reduction for the MCDMs. Additionally, the absence of any standard procedure in the watershed prioritization process makes it relatively difficult for the end users over which statistical and MCDM method to choose for the available inventory.

The Nandhour-Kalish watershed is one of the relatively ungauged watersheds in the state of Uttarakhand, with limited ground stations for ground water monitoring, particularly in the high-altitude regions. The river contributes to the Nanakmatta dam near Sitarganj, Udhamsingh Nagar district; the dam outlet connects to the Deoha river in Pilibhit, Uttar Pradesh, which further connects with the Ghaghra River, a tributary of the great Ganga River. Subsequently, the watershed holds critical significance, being one of the primary water resources in the districts of Champawat, Udhamsingh Nagar, Nainital, and others. Considering that limited studies exist in the literature describing the status of the Nandhour-Kalish watershed, this study makes an attempt towards the development of a canonical procedure for sub-watershed prioritization of this watershed in a GIS-assisted framework using both statistical and MCDM techniques. The structure of this paper unfolds in two main sections: first, it strives to develop a Standard Operating Procedure (SOP) for selecting most influencing morphometric parameters or criteria parameters in regards to generation of surface runoff and sediment yield through the use of a data reduction technique called PCA. Second, the paper explores the application of Criteria Importance for Intercriteria Correlation (CRITIC) MCDMs to assign the relative priority among identified criteria indices and overall computation of priority weights for the developed sub-watersheds. This categorization ensures a lucid and organized presentation, centring on SOP development and the effective use of CRITIC MCDM for watershed prioritization.

The headwaters of the Nandhour-Kalish watershed can be traced to the Kundal River Forest located in the Champawat district of Uttarakhand. The spatial extent of the watershed ranges from latitude 28°53ʹN to 29°23ʹN and longitude 78°45ʹE to 80°08ʹE. The topography of the watershed varies significantly, ranging from 132 m above mean sea level in the flat foothill areas to 2,084 m in high mountains. The watershed areal coverage is approximately 474.09 km², with a periphery perimeter of 276.21 km. The average annual rainfall of the study area is 1,433.4 mm, with minimum temperatures ranging from 3 to 5 °C and maximum temperatures ranging from 39 to 41 °C, respectively (Gururani et al. 2023).

The study employs the Advanced Spaceborne Thermal Emission and Reflection Radiometer Global Digital Elevation Model (ASTER GDEM) version 3.0 and Survey of India (SoI) toposheets (Toposheet no. 53O11, 53O12, 53O15, 53O16, 53P9, and 53P13 of scale 1:50,000) datasets for the analysis and estimation in the context of morphometry-based prioritization. Figure 1 shows the topography of the Nandhour-Kalish watershed based on ASTER GDEM. The DEM product is used in further morphometric analysis by extensively utilizing the ArcGIS hydrology tools in conjunction with the Soil and Water Assessment Tool (SWAT) 2012. SWAT was only used in the development of the sub-watershed using the ‘watershed delineator’ tab. Only the watershed data were taken from the SWAT interface towards the estimation of morphometric parameters as the sub-watershed level, as in previous studies (Rani et al. 2022; Gururani et al. 2023). Additionally, PCA was used for the selection of the most influential morphometric parameter and was executed using the IBM SPSS Statistics 19 tool. The advanced automated statistical analysis features of the software were leveraged for a detailed exploration of data patterns and structures, contributing towards improving the overall understanding of the dataset, while effectively excluding any potential outliers or irregularities.

The linear parameters included the following:

• 1. Basin length (L_b),

• 2. Stream number (N_u),

• 3. Average stream length (L̅_u),

• 4. Bifurcation ratio (R_b),

• 5. Stream length ratio (R_l).

The areal indices included the following:

• 6. Watershed area (A),

• 7. Drainage density (D_d),

• 8. Stream frequency (F_s),

• 9. Texture ratio (R_T),

• 10. Elongation ratio (R_e),

• 11. Form factor (F_f),

• 12. Drainage texture (T),

• 13. Length of overland flow (L_g)

Relief parameters included the following:

• 14. Relief ratio (RR)

• 15. Ruggedness no. (R_n)

• 16. Hypsometric Integral (HI)

• 17. Slope (S)

PCA is a widely employed dimensionality reduction method in statistics and machine learning (Maćkiewicz & Ratajczak 1993). PCA has been widely used for handling high-dimensional datasets, aiming to condense the data while preserving the majority of its variability (Meshram & Sharma 2017; Siddiqui et al. 2020; Tiwari & Kushwaha 2021; Kumar et al. 2022; Sharma et al. 2023). PCA, also commonly referred to as the Principal Component Transform, reduces dimensionality or selects the most significant features by transforming the original variables into a set of uncorrelated variables referred to as principal components (PCs) (Wold et al. 1987). These components are linear combinations of the initial variables and are organized based on the variance they individually define. In essence, PCA provides a streamlined representation of the data, enabling a more efficient analysis while retaining essential information (Maćkiewicz & Ratajczak 1993). In this study, the estimated morphometric parameters were imported to IBM SPSS Statistics 19 for the identification of the most influencing parameters (later referred to as criteria indices) using PCA.

The IBM SPSS Statistics 19 tool executes PCA in a semi-automated manner, enabling users to interpret extensive datasets into a concise set of components (Kassambara 2017). By developing a covariance matrix, PCA quantifies the relationship between input variables. While the number of development PCs are equal to the number of input variables, not all of the developed PCs are essential for data interpretation and subsequent reduction. To determine the most informative PCs, a scree plot is developed, showing eigenvalues against the PCs. A significant change in slope suggests the break towards the identification of candidate PCs. These PCs cumulatively account for explaining 70–90% of the variance in the input dataset (Jolliffe 2002). Following this, the pattern matrix, which is an oblique rotated factor solution, filters criteria parameters. It assesses how well inputs load on developed PCs. Morphometric indices with loading scores ≥ 0.8 are selected from legitimate PCs, resulting in the selection of 10 morphometric parameters for subsequent subjection to CRITIC MCDM towards the determination of their relative weights.

CRITIC is a decision-making and multi-criteria analysis method designed to assess and prioritize alternatives using a predefined set of criteria. This approach aids in determining the significance of each criterion and gaining insights into the interrelationships or correlations among them (Diakoulaki et al. 1995). By evaluating both the importance and interactions of criteria, CRITIC provides a comprehensive framework for decision-makers to prioritize alternatives effectively in complex decision scenarios for the assessment of relative weights.

The individual SM values, i.e., SM_i values corresponding to each parameter, were then added to estimate Symmetric Matrix Values (SMV_i), which mathematically multiplied with SD provide the Criteria Information (C_j) of the parameter. The various C_j values are then normalized to get the relative ratings, R_j, of each parameter. Using these ratings and the normalized values of criteria indices, a score, sub-watershed scores (W_j), was generated for each sub-watershed. These sub-watershed scores were then categorized into four classes using a quartile-based approach (as follows) to derive the final relative priority map using the ArcGIS Interface:

• 1. Less than the first quartile (Q₁) was labeled as low priority,

• 2. Greater than or equal to Q₁ but less than the second quartile (Q₂) was placed under moderate,

• 3. Greater than or equal to Q₂ but less than the third quartile (Q₃) was categorized as high,

• 4. Greater than Q₃ was tallied under very high classes of relative priority.

The Nandhour-Kalish watershed is determined to be a fifth-order watershed, with a total stream length of 833.06 km, a basin length of 43.43 km, and a basin width of 10.92 km with a dendritic and sub-dendritic type drainage pattern. Of the drainage pattern of the identified 12 sub-watersheds, SW11 has the largest geographic area (76.06 km²), while SW1 has the least (29.09 km²). The numeric values of morphometric parameters are presented in Table 1.

The drainage density in a way defines the degree of closeness among the stream channel and affects the infiltration, time to peak, as well as the volume of runoff at the outlet of any drainage system. Typically, large values of D_d correspond to a lower infiltration (Choudhari et al. 2018). The numeric value of D_d for the Nandhour-Kalish watershed is observed to be 1.76 km/km², placing the watershed under the class of a very coarse drainage texture (Smith 1950). For the sub-watersheds, the same ranges from 1.45 km/km² (SW2) to 2.14 km/km² (SW9) are observed.

Length of overland flow (L_g), which is half of the reciprocal of drainage density, acts as an index indicative of the estimated distance traversed by sheet flow water before transitioning into channelized flow. It mostly depends on the surface slope (Horton 1945). Its lower values reflect more sheet erosion, while higher values correspond to greater channel erosion (Das et al. 2021). For the study area, its value was estimated at 0.285 km, while for the sub-watersheds, the same fluctuated between 0.234 km (SW9) and 0.346 km (SW2).

The values of F_s and R_T for the study area are estimated at 3.12/km² and 5.49/km, respectively. At the sub-watershed scale, these parameters range from 2.45 (SW6) to 3.68 (SW11)/km² and 1.45 (SW9) to 2.95 (SW5)/km, respectively. Usually, large values of these indices suggest high slope and low infiltration characteristics (Farhan et al. 2017).

R_e and F_f for the Nandhour-Kailash watershed were estimated at 0.566 and 0.995, respectively. According to the Strahler (1952) classification, the Nandhour-Kalish watershed is observed to have an elongated shape (with a R_e value less than 0.7), which indicates that the peak flows shall be relatively small for rainfall events of a duration less than the time of concentration. In regards to the sub-watersheds, R_e and F_f are in the range from 0.641 (SW11) to 0.684 (SW1) and 0.681 (SW1) to 0.776 (SW11), respectively.

The drainage texture, T, also referred to as the infiltration number, reflects on the relative channel spacing in a fluvial dissected terrain. The greater the value of T, the greater the runoff (Smith 1950). Moreover, it is a function of climate, land use, soil, relief, and developmental stage of the drainage area. For the Nandhour-Kalish watershed, T is estimated to be 5.49/km³, while for the sub-watersheds, it varies from 3.86 (SW6) to 7.54 (SW11)/km³.

The basin relief, which is the difference between its highest and lowest elevations, of the Nandhour-Kailash watershed is 1,951 m, while for the sub-watersheds, it digresses from 1,478 m (SW4) to 27 m (SW12). As a reflection of denudational characteristics, large relief with small basin length represents higher gradients (Sharma et al. 2015). The RR is a measure of the overall steepness and an indicator of the intensity of the erosion processes; the RR for the Nandhour-Kailash watershed was estimated at 0.706 km/km², while for the sub-watersheds, it was approximated from 0.003 (SW12) to 0.144 (SW2), respectively. Slope expresses the average weighted gradient and is critical in governing the infiltration as it directly influences the contact time between the overland flow and the soil. Furthermore, it plays a key role in the selection of appropriate water and soil conservation measures (Gaikwad & Bhagat 2017). For the sub-watershed, the slope range is 57.10% (SW2) to 4.63% (SW12).

The ruggedness number, a dimensionless parameter, asserts the cumulative effect of length and gradient attributes (Aher et al. 2014). Larger values of R_n suggest more potentiality towards soil erosion, whereas lower values articulate structural complexity. For the study area, R_n was estimated at 0.044, while for the sub-watersheds, the observed range was 0.001 (SW10, SW11, and SW12) to 0.039 (SW1), respectively. The HI parameter represents the development stage of the watershed. Additionally, HI also gives acumen to the activity strength in the geologic structures of a drainage system. Strahler (1964) defined a classification for the watershed development stage into three categories based on their HI values. For the Nandhour-Kailash watershed, HI was estimated at 0.481, representing an equilibrium or mature stage. For the sub-watersheds, the same wavered from 0.188 (SW8), representing an old/Monadnock stage, to 0.458 (SW10), suggesting a mature/equilibrium stage of development.

To identify criteria indices from the PCA, the pattern matrix with parameter loadings as depicted in Table 2 was utilized. Figure 4(b) illustrates the stacked column chart of parameter loadings. A threshold absolute loading value greater than 0.8 is used for the selection of criteria indices. From Table 2, it is evident that parameters D_d and L̅_u load quite well in PC1, while loadings of L_g, S, and RR are negative. This suggests that the set of these parameters tends to change together, though it is not always statistically correct. Similarly, A, N_u, and L_b have a positive loading in PC2, while F_f and R_e displayed negative loadings. Subsequently, in PC3, HI discerns a positive loading value. This analysis resulted in the selection of 11 key input parameters as criteria indices, namely, D_d, L̅_u, L_g, S, RR, A, N_u, L_b, F_f, R_e, and HI. A 35.29% reduction in the number of input parameters is obtained via the application of PCA.

Following the identification of criteria indices, the contention lies in assigning relative weights to the criteria indices with respect to each other. For this purpose, an MCDM approach named CRITIC is employed. Under CRITIC, Contrast Intensity (CI) and the Measurement of Conflict (MOC) are estimated. The computation of CI is based on the SD, with larger values corresponding to greater relative importance, while ‘r_jk’ is used for the determination of a conflict created, with lower values of ‘r_jk’ citing higher conflict and vice versa. Combinedly, CI and MOC are used to estimate the R_j of the input parameters. Supplementary material Tables S3 illustrates the NDM from CRITIC, while Supplementary material Table S4 illustrates the correlation matrix. From the criteria indices, for most of them, higher values represent larger potentialities towards the generation of surface runoff and erosion, except for L_g for which the case is vice versa. Supplementary material Table S5 depicts the SMV_i, C_j, and R_j values for the input variables. Following the computation, it was observed that RR is assigned the greater R_j (13.3%), followed by S (12.93%), while the least was determined for N_u (7.5%).

The minimum value of W_i was observed to be 6.2632 for SW9, and the maximum value was observed to be 9.8186 for SW2. Subsequently, these two sub-watersheds represent the regions requiring least and immediate water conservation practices in their observed priority classes for the Nandhour-Kalish watershed. Interestingly, each of the four priority classes is observed to include three sub-watersheds. The entire watershed is covered by three primary regions of the Champawat district to the east, Nainital to the north-west, and Udham Singh Nagar to the south. The watershed receives high overall mean annual precipitation in excess of 1,300 mm, as per the rainfall data of the Indian Meteorological Department (IMD). We analyzed the precipitation trends for these regions, as shown in Figure 5. Due to the coarser spatial resolution (0.25 × 0.25°) of the IMD rainfall data, in some cases the sub-watersheds were merged in the assessment of the precipitation trends. Overall, it is observed that the sub-watersheds in the ‘very high’ and ‘high’ priority classes show a relatively significant declining trend of precipitation compared to the sub-watersheds in the ‘moderate’ and ‘low’ priority classes. The combined precipitation trend for SW4 and SW8 shows a significantly declining trend. Notably, here it indicates the dominance of the SW4 (‘very high’ priority) over SW8 (‘moderate’ priority). Subsequent trends for the other sub-watersheds, namely, SW10-12 (combined trend for SW10 and SW12), SW12, and SW11, show increasing trends of precipitation, indicating a relatively higher water availability for sub-watersheds in ‘low’ and ‘moderate’ priorities.

In a study by Chuphal et al. (2024), the drought conditions in India were investigated based on the z-scores corresponding to the standardized precipitation and evapotranspiration index. The seasonal state and district wise time series of the drought categories with these z-scores are available at https://indiadroughtatlas.in/. We analyzed these scores for the regions in the sub-watersheds that were available in this dataset, as shown in Figure 5. The approximate mean z-score for the regions in the sub-watersheds corresponding to the ‘very high’ and ‘high’ priorities in the Champawat and Nainital districts showed, in general, relatively much higher values of the mean z-score when compared with those in the Udham Singh Nagar district (regions corresponding to ‘low’ and ‘moderate’ priority classes). The excessive ground water extraction compared to recharge in the Nainital regions (Haldwani) in contrast to the Udham Singh Nagar (Sitarganj) also shows some agreement with the relative sub-watershed priorities (see Supplementary material, Figure S1).

One of the primary challenges in effectively implementing and developing efficient water management practices is soil erosion. Uttarakhand (study area) is broadly divided into three physiographic regions: Tarai, Bhabhar (foothills), and the Shivalik's (outer Himalayas). According to George et al. (2021), in the Tarai region of Uttarakhand (which includes SW10, SW11, and SW12), the mean soil erosion rate was 7.45 t ha⁻¹ yr⁻¹, resulting in a total soil loss of 2.94 Mt yr⁻¹. In the Bhabhar (foothills) region (which includes SW8, SW4, and SW6) of Uttarakhand, the mean soil erosion rate was 9.70 t ha⁻¹ yr⁻¹, leading to a total soil loss of 5.35 Mt yr⁻¹. In the Shivalik's (outer Himalayas) region (which includes SW1, SW2, SW3, and SW5) of Uttarakhand, the mean soil erosion rate was 17.13 t ha⁻¹ yr⁻¹, resulting in a total soil loss of 7.48 Mt yr⁻¹. It was observed that the overall mean erosion rate and total soil loss of the Ganga River basin were 22.22 t ha⁻¹ yr⁻¹ and 15.69 Mt yr⁻¹, respectively. The comparatively higher mean soil loss rates observed in Shivaliks can be attributed to the higher rainfall rates, resulting in higher rainfall erosivity values. Altogether, more than 50% of the total soil loss from the entire study area is contributed by hilly physiographic regions (categorized in very high and high priority sub-watersheds). These observations agree with the relative watershed priority map for the Nadhour-Kalish River watershed.

Soil conservation practices take into consideration factors such as soil erodibility, commonly used to identify the ability of soils to resist erosion, based on their physical characteristics (Mahapatra et al. 2018). The topographic factor determines the influence of topography on soil and is computed using the gradient and length of slope at each grid observation location. Generally, the steeper the slope of the land, the greater the amount of soil loss by water. Soil erosion by water accelerates as the slope length increases due to a greater accumulation of runoff (Kumar et al. 2021). According to Mahapatra et al. (2018), the values of the K-factor and topographic factor observed in the Champawat region (covering SW1 and SW5), the Nainital region (covering SW2–SW4, SW6–SW8), and the Udham Singh Nagar region (covering SW9–SW12) were 3.22, 7.22, and 5.71%, respectively. In general, runoff occurs when there is an excess amount of water on a slope that cannot be absorbed into the soil or trapped on the surface. Low infiltration due to soil compaction, crusting, or freezing accelerates the amount of runoff over a terrain. The rainfall erosivity factor was observed to be 16.15% in the Champawat region, 9.36% in the Nainital region, and 5.78% in the Udham Singh Nagar region. It was noted that 9.36 and 5.78% of the area are affected by very slight and slight soil loss, respectively, mainly covering the Nainital and the Udham Singh Nagar districts. However, some portions of the moderately severe and moderate erosion classes occur in Nainital, Champawat, combining 7.37% for the hilly region and the Udham Singh Nagar region and 6.71% for the tarai region, respectively. Theoretically, these observations are consistent with the developed relative priority map developed in this study for the Nandhour-Kalish watershed, which also indicates relatively higher priorities for sub-watersheds in the hilly regions as compared to the tarai regions.

Watershed prioritization using MCDM techniques is pivotal in resource management and environmental conservation efforts. However, prioritizing sub-watersheds using morphometric indices has been challenging due to the problem of selection from a wide availability of a large number of morphometric parameters. Additionally, the absence of standard operating procedures often leaves investigators uncertain about which specific procedure to follow in watershed priority assessment. This uncertainty can be addressed by employing PCA and CRITIC in grouping morphometric parameters for watershed management, specifically focusing on the Nandhour-Kalish River watershed. Through MCDM techniques, which are widely used for watershed prioritization, decision makers can weigh the importance of each criterion and assess their impacts on watershed health and future sustainability. Subsequently, prioritization facilitates targeted interventions and investments in areas most vulnerable to degradation or in need of restoration. Moreover, MCDM empowers stakeholders to make informed decisions, fostering consensus-building and equitable distribution of resources among competing interests, especially in conditions where on ground hydro-meteorological data such as water yield and soil loss are unavailable, limiting applicability of hydrologic modeling. Ultimately, watershed prioritization using MCDM enhances the efficacy of conservation strategies, promotes resilience against environmental stressors, and supports long-term sustainability goals.

This study is an attempt to resolve this issue and develop a nimble procedure for sub-watershed prioritization using morphometric indices only. The presented study uses a statistical method (PCA) for identifying criteria morphometric indices from the set and an MDCM (CRITIC) for assessment of relative weights of the identified criteria variables. The results from the study suggest that morphometrically, approximately 47.49% of the geographic area of the watershed is under very high and high classes of relative priority in the context of runoff generation and soil erosion. The outputs from the deliberated study shall aid in the determination of regions with high erosion and runoff susceptibilities.

The importance of the integrated PCA-CRITIC approach lies in the sense that a user may input any number of morphometric indices, select priority parameters using the PCA, and then assess the relative weights of priority parameters using CRITIC. The conferred study is a pre-requisite before the implementation of any catchment conservation programme as it identifies sub-watersheds that require utmost attention within the study area. Subsequent facilitation of the optimal allocation of conservation and management interventions, especially in a resource-limited environment, is feasible.

The authors report no funding for this work.

All relevant data are available from an online repository or repositories. The 30 m ASTER DEM is freely available from the USGS at https://earthexplorer.usgs.gov/. The toposheets for the Nandhour-Kalish Watershed are available freely for Indian Citizens from the Survey of India at https://onlinemaps.surveyofindia.gov.in/FreeMapSpecification.aspx

The authors declare there is no conflict.