Prioritization and identification of vulnerable sub-watersheds using morphometric analysis and an integrated AHP-VIKOR method Open Access

The existence of water and soil are the most indispensable resources on earth for the sustenance of living beings. Climate change and increasing population have an adverse effect on the environment, leading to soil erosion that reduces land productivity and the capacity of rivers. The surface characteristics of a catchment have a direct relationship with runoff, and morphometric analysis can be a key aspect of understanding the characteristics of a watershed (Tripathi et al. 2005). Various morphometric parameters can be quantitatively used to examine a watershed and its stream network pattern (Horton 1945). The analysis of channel networks and drainage basins is important in order to understand a catchment's geo-hydrological behaviour. Drainage basins are a distinct morphologic region with special implications for drainage patterns and geomorphology (Strahler 1957). A key part of any hydrological investigation is a morphometric analysis, as it helps in the management and development of drainage basins (Kumar et al. 2015). The lithology, bedrock, and geological structures of a watershed significantly influence morphometric parameters. As a result, data on geomorphology, hydrology, geology, and land-use patterns are extremely useful for conducting a reliable study of the drainage network of a watershed (Binjolkar & Keshari 2007).

Unsustainable usage of natural resources has a negative impact, which eventually leads to environmental degradation. The need for precise data is a major challenge for micro-level sustainable natural resource planning and management. Consequently, micro-level hydrological unit sub-watersheds are carefully selected to improve planning and management approaches by addressing severe problems like soil erosion, soil degradation, droughts, floods, and excessive runoff (Aher et al. 2014). In comparison to traditional data processing methods, fast-emerging geospatial technology, remote sensing (RS), and geographic information systems (GISs) are effective techniques for overcoming most land and water-related problems for sustainable planning and management (Rao et al. 2010; Singh et al. 2021). With the emergence of satellite technologies, various aspects of watershed-level planning can now be executed spatially across a large area. To understand aspects of the drainage system of the catchment of a watershed, RS and GIS can be utilized for quantitative drainage analysis (Lakshminarayana et al. 2022). Prioritization of watersheds is imperative for preparing a holistic approach for natural resources management and conservation. Watershed management practices, on the other hand, are impractical to execute throughout a watershed. Thus, they should be initiated within the sub-watershed that is most sensitive to vulnerability.

Prioritization involves assessing various erosive hazard parameters, which generates ambiguous issues when ranking watersheds. To address this issue, a plethora of multi-criteria decision making (MCDM) methods is used to provide and elucidate an optimum solution for watershed prioritization. The analytical hierarchy process (AHP) is an MCDM method that uses a pairwise comparison procedure to determine a preference scale between alternatives (Saaty 1980). According to that review, various studies have suggested that AHP provides a viable decision among different alternatives in watershed prioritization by reducing bias among the parameters (Kundu et al. 2017; Gaikwad & Bhagat 2018; Balasubramani et al. 2019; Bera & Banik 2019; Kumar & Sarkar 2022; Shivhare et al. 2022). Alternatively, a few researchers have claimed that the VIKOR (VIseKriterijumska Optimizacija I Kompromisno Resenje) method has higher accuracy for prioritization (Amiri et al. 2018; Arab et al. 2018; Bhattacharya et al. 2020; Meshram et al. 2020; Chae et al. 2022).

In Jharkhand, India, the tolerance limit of soil loss has exceeded its threshold limit, which needs proper conservation planning to increase agricultural production (Sahoo et al. 2014). Moreover, Lal et al. (2019) used a soil erosion model, the Revised Universal Soil Loss Equation (RUSLE), which identified that the Ranchi district soil erosion ranges from 2 to 30 tons/ha/year, with the majority of the soil erosion affected areas being located within the vicinity of the current study area. Therefore, ranking the different sub-watersheds in these areas will serve a crucial role in the sustainable development of watersheds. Hence, keeping the above perspective in view, the present work aims to combine the AHP and VIKOR methods to prioritize various sub-watersheds, along with a sensitivity analysis of the model.

The delineation of Rarhu watershed was conducted using the Pour Point feature in ArcGIS from NASADEM. The Soil and Land Use Survey of India (SLUSI) has provided specific codification for each sub-watershed, and Rarhu watershed comes under watershed code 4H3C7. Sub-watersheds were prepared according to SLUSI datasets.

A quantitative morphometric study of watersheds is the best approach for assessing the relationship between various aspects of an area (Sukristiyanti et al. 2018). The study used linear, areal, and relief parameters such as bifurcation ratio (R_b), drainage density (D_d), elongation rate (R_e), compactness coefficient (C_c), form factor (R_f), stream frequency (F_s), drainage texture (D_t), circulatory ratio, (R_c), basin shape (B_s), basin relief (R), and relief ratio (R_r) which are shown in Table 1. The morphometric data used in the analysis was prepared using ArcGIS 10.8 software.

MCDM methods are effective in complex decision-making because they provide the best solutions and reduce biases within different alternatives. In the current research, two MCDM methods, AHP and VIKOR, were integrated. Weights were calculated using the AHP method, which was integrated into the VIKOR method to assign ranking of sub-watersheds. Further, eleven different parameters were selected for the construction of the AHP matrix in this study.

An AHP is a framework for dealing with complex decision-making issues. It has a multi-level system with a variety of decision criteria, sub-criteria, objectives, and alternatives. The AHP method uses a pairwise comparison procedure to arrive at a scale of preferences between different alternatives (Saaty 1980). AHP is a robust and versatile decision-making approach that can be applied to various multi-criteria issues, such as assessing the risk of soil erosion (Jaiswal et al. 2014). It also proposes to measure the validity and consistency of the pairwise comparison matrix to prevent any subjectivity in decision-making. AHP analysis leads to less repetition of data and avoids ambiguous calculations.

A random consistency index (RI), as described by Saaty (1980), was calculated once the reciprocal matrix of different sizes was obtained. The comparison matrix in the current study has a total of eleven elements. Table 2 shows the random consistency index values.

Among the various MCDM approaches, the VIKOR method is an effective method for addressing issues related to incompatible and compatible standards (Opricovic & Tzeng 2004). A decision matrix is created, and normalization is computed using a linear method. Criteria weight values obtained through the AHP method were employed to calculate a VIKOR weighted normalized matrix.

In sensitivity analysis, the selection of alternatives varies as the relative importance of specific criteria in the selection process changes (Muñoz et al. 2016). It assesses the accuracy and reliability of a decision-making method and may also be used to validate the level of uncertainty among the values. In this study, the ‘v’ values of the VIKOR method were altered from 0 to 1 to evaluate the extent of variation in sub-watershed ranking. Values with low sensitivity are advantageous, while values with high sensitivity are disadvantageous, influencing the ranking of alternatives.

Morphometry is a quantitative approach to study the three-dimensional features of drainage basins, which incorporates linear, areal, and relief parameters. Horton (1945) developed a set of quantitative methods for studying the drainage features in different basins. In the current research, the linear, areal, and relief parameters were calculated for the prioritization of 21 sub-watersheds (named SW-1 through to SW-21), as shown in Table 3.

Areal parameters directly influence the runoff and hydrograph of a rainfall event. The drainage density (D_d) illustrates the close connection of the stream channel spacing (Horton 1932). The higher drainage density values indicate more runoff with meagre vegetation and hilly terrain. However, the lower the drainage density, the lower will be the relief with high recharge or potential groundwater zones with porous subsoil. There are five drainage density classifications which range from extremely coarse (<2), coarse (2–4), moderate (4–6), fine (6–8), to very fine (>8) (Chandrashekar et al. 2015). Rarhu watershed drainage density values vary from 1.330 to 2.042 km/km². The elongation ratio is a dimensionless value used to describe the shape of a watershed. In general, the elongation ratio values are categorized into two types: low values, which represent an extended watershed, and high values that define a circular watershed. According to Strahler (1957), if an area has high relief and a steep land slope, the values range between 0.6 and 0.9, and SW-3, SW-4, SW-8, SW-9, SW-15, and SW-18 in the Rarhu watershed fall into this category. A higher elongation ratio indicates a good infiltration rate with low runoff, while a lower number indicates a low infiltration rate with high runoff. The compactness coefficient defines the relationship between a hydrologic basin and a circular basin of same area. A circular basin is detrimental to drainage because it generates a shorter time of concentration before a basin reaches its peak flow (Altaf et al. 2013). Furthermore, C_c is independent of watershed area and solely depends on slope, and it has an inverse relationship with erodibility. Across the watershed, C_c values were highest for SW-1, SW-5, SW-10, SW-11, and SW-16, indicating less susceptibility to erosion, and lowest for SW-4, SW-8, SW-9, and SW-20, suggesting more susceptibility to erosion.

The form factor is defined as watershed area divided by the square of the watershed's axial length, and the values are less than 1. A lower form factor value indicates a more elongated shape with a lower peak flow for a longer duration. In contrast, a higher form factor value shows a more circular shape with a higher peak flow for a shorter duration. In this research, sub-watersheds SW-1, SW-5, SW-6, SW-10, SW-11, SW-13, and SW-16 have the lowest R_f value with maximum elongation shape where low peak flow occurs for a longer duration, unlike SW-3, SW-15, SW-18, and SW-20 which have R_f values suggesting a circular shape. SW-4 and SW-21 have the highest stream frequency (F_s) values of 1.727 and 1.690, respectively, indicating that both watersheds have mountainous relief and uneven topography. Drainage texture measures a physical characteristic of a watershed by dividing the total number of stream segments of all orders within the watershed by the watershed's perimeter (Horton 1945). SW-3, SW-4, SW-9, SW-14, SW-18, and SW-21 have coarse to moderate D_t with high relief, impermeable soil surface, low infiltration capacity, and excessive runoff.

The circulatory ratio, as described by Miller (1953), is the ratio of a watershed area to the area of a circle with the same perimeter as the watershed. The circulatory ratio in the basin varies from 0.4 to 0.5, indicating highly elongated and extremely permeable homogenous geologic materials. The circulatory ratio value of a watershed is high if it has a circular shape, a moderate to high elevation, and a high permeable surface. A sub-watershed with a low circulatory ratio value is typically characterized by its low elevation, elongated shape, and highly impermeable surface. In addition, R_c is controlled by the watershed's length and frequency of streams, geological features, relief and slope steepness, climate, and land cover (Miller 1953). The calculated values for R_c ranged from a minimum value observed in SW-10 of 0.189 to a maximum value in SW-20 of 0.779. The basin shape parameter (Bs) and erodibility have an inverse relationship: the lower the value, the higher the erodibility. In Rarhu watershed, SW-20 has the lowest value of 1.570, while the SW-10 has the highest value of 13.091.

In the current research, two relief parameters were determined for the prioritization of sub-watersheds. Basin relief is the vertical difference between the highest and lowest points. The basin relief determines the stream gradient, as well as flood patterns and sediment transport volumes (Hadley & Schumm 1961). SW-11 has the highest relief value of 416 m, showing a steep slope and more erodibility due to runoff, whereas SW-20 has the lowest relief value of 146 m (Table 5). The relief ratio is a measure of a basin's overall relief, defined as the elevation difference between the basin's lowest and highest points along the basin's longest dimension parallel to the main drainage line. The SW-15 relief ratio was calculated to be 0.047, and it has been observed that higher relief ratio values describe areas with high reliefs and steep terrain. Furthermore, SW- 5 had the lowest value of 0.012, which might be attributed to the drainage basin's resistant underlying rocks and low slope (Schumm 1956). A high relief ratio value distinguishes hilly terrain. High relief areas were observed in the west part of the watershed.

Prioritization is a vital step in any watershed management approach for attaining beneficial outcomes and identifying problematic areas in order to find an appropriate solution with various soil and water conservation measures. The use of compound parameter values for estimating watersheds is not as ambiguous as it may sound. However, compound parameters weigh all parameters equally, despite the fact that certain features are more or less relevant in identifying soil erosion risk than others. Therefore, ranking watersheds based on a compound parameter can be misleading. In this research, the AHP method was used to calculate the weight of each morphometric parameter. As seen in Table 6, AHP entails objectively constructing a matrix and then comparing each potential pair. Additionally, the pairwise comparison matrix must be checked for validity and consistency, eliminating decision-making subjectivity.

It is important to determine the robustness and reliability of the decision-making process (Malekian & Azarnivand 2015). Therefore, a sensitivity analysis was conducted to ascertain how the values of ‘v’ influence the ranking of various sub-watersheds when altered in the VIKOR method. The v value is crucial in ranking the alternatives (i.e., the sub-watersheds). Similar analyses were conducted by Suh et al. (2019) and Ramavandi et al. (2021) to examine the model's applicability for decision making.

The sustainability of a biological and ecological system primarily depends on the availability and use of soil and water resources. Therefore, developing a long-term watershed management strategy necessitates watershed prioritization. A morphometric analysis of the different features is required to understand a watershed's hydrological behaviour dynamics. Remote sensing and geographic information systems are the most advanced approaches for assessing watershed priority spatially for sustainable development and management. According to the findings of this study, sub-watersheds SW-4, SW-9, SW-14, SW-18, and SW-21, with a total area of 227.88 km² (36.17%), are highly vulnerable and need immediate action with suitable soil and water conservation practices. Further, it was found that, 147.47 km² (23.40%) belong to high, 135.57 km² (21.51%) are moderate, and 123.04 km² (19.53%) are of a low priority class. Thus, the study demonstrated that prioritizing sub-watersheds primarily on morphometric parameters and using the integrated AHP-VIKOR method could be efficient and time-saving in watershed management and planning. This integrated approach showed that more accurate results could be achieved using the VIKOR method by considering maximum group utility and minimum individual regret. The findings of the study may be utilized by decision-makers, resource planners, and watershed development projects to identify priority sub-watersheds within the Rarhu watershed that need immediate adaptation with proper conservation and land management measures.

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

The authors declare there is no conflict.