Watershed Prioritisation of Drainage Basins Based on Geomorphometric Parameters, Neyyar Watershed, India

Figure 2 illustrates the process over DEM. Using ArcGIS 10.4.1, sub-watersheds (SW1– SW11) are categorised. Linear, shape and relief parameters were the three types of morphometric features analysed and classified. Table 1 shows the many empirical approaches used to determine these characteristics. The Neyyar River basin linear parameters were estimated and reported in Table 2. The highest value is ranked as 1, for the relief and linear features, and so on, while the lowest value is ranked as 1, for the areal or shape feature, and so on. The compound parameter (Cp) value was obtained by adding all the ranks in SW1, and the resultant sum was divided by the number of features. This was repeated for the remaining SWs. Based on their Cp values, the SWs were divided into three groups: high, medium and low.

Fig. 2.

PCA is performed to reduce many variables into a small number of features while preserving the original information’s relationships. In other words, one of the crucial morphometric features for prioritising watersheds based on features that are strongly connected with components was assessed utilising PCA. In the current study, 15 morphometric factors were reduced to three crucial elements using Statistical Package for Social Sciences (SPSS) software version 22. Each component accounts for one highly associated feature, as seen by the rotated component matrix. Nookaratnam et al. (2005) stated that soil erosion has a direct impact on linear and relief features as well as an indirect impact on shape features. After obtaining three features, the sub-watershed with the highest value of relief and linear features has been ranked first, followed by the sub-watershed with the second highest value in the second rank, the third highest value in the third rank, and so on. The sub-watershed with the lowest value in the shape feature has been ranked first, followed by the sub-watershed with the second lowest value in the second rank, the third lowest value in the third rank, and so on (Nookaratnam et al. 2005). The Cp value was obtained by adding all the ranks in SW1, and the resultant sum was divided by the number of features. This was repeated for the remaining SWs. Based on their Cp values, the SWs were divided into three groups: high, medium and low. However, the PCA method of data handling has its requirements and limitations, and these have to be carefully observed.

The hypsometric curve and HI for the sub-watershed were also determined using the SRTM-DEM. The elevation-to-relief ratio technique was used in this study to estimate HI values. The HIs in this case study range from 0.146 to 0.388. After obtaining the HI values, the next step is to divide the HI values into three equal intervals in order to assign ranks as the SWs were classified into three groups: high, medium and low. The maximum interval values are treated as high priority, the next interval values are treated as medium priority and the minimum interval values are treated as low priority.

Table 3 displays the descriptive statistical analysis of all selected sub-watershed morphometric properties. Minimum, maximum, sum, mean, standard error, variance, standard deviation, median, 25th percentile, 75th percentile, skewness and kurtosis were calculated. The R_bm and L_o parameters have a left-skewed distribution, as shown in Table 3, R_lm, D_d, R_hp, R_c, R_e, F_f, F_s, D_t, C_c, D_i, B_h and R_n parameters have a right-skewed distribution. With morphometric parameters’ kurtosis values evaluated, it becomes evident that 10 of the parameters, namely, R_bm, R_lm, D_d, L_o, R_hp, R_c, R_e, F_f and C_c, exhibit platykurtic distribution. It indicates that the estimated kurtosis results were flat with thin tails, resulting in negative kurtosis results. Other morphometric features have positive kurtosis results, indicating that the distribution for these groupings of features is peaked and has thick tails, indicating that they are in the leptokurtic distribution (Redvan, Mustafa 2021). The whiskers box plots for the stated parameters are shown in Figure 10.

Fig. 10.

To examine the features of the drainage system, morphometric analysis was performed for the SWs. To prioritise SWs for soil conservation, 15 morphometric factors representing linear, shape and relief features were used. Linear, shape and relief features are three different morphometric features that have been used to rank SWs in order of risk because they are either directly or inversely related to the risk of soil erosion, runoff and peak discharge. The relief and linear features are directly associated with risk of soil erosion, runoff and peak discharge, while the shape features are inversely associated with risk of soil erosion, runoff and peak discharge (Nookaratnam et al. 2005).

The highest value is ranked first for relief and linear features, and so on, while the lowest value is ranked first for the shape feature, and so on. Table 4 displays the ranking results for all 11 SWs. The compound parameter (Cp) value (4.47) was obtained by adding all the ranks in SW1 (67), and the resultant sum was divided by the number of features (15). This was repeated for the remaining SWs.

The SWs were classified into three classes after the compound value calculation: high (≥ 4.45 to < 7.33), medium (≥ 5.33 to < 6.20) and low (≥ 6.75 to < 7.07). SW1 is a high-priority sub-watershed, while SW2, SW4, SW6, SW3, SW7 and SW8 are medium-priority SWs, and SW5, SW9, SW10 and SW11 are low-priority SWs. The final sub-watershed priority map for the Neyyar watershed is shown in Figure 11.

Fig. 11.

A statistical technique known as PCA can be used to find hidden elements that determine the pattern of correlations among several observable variables while preserving the original, accurate data. The correlation between each morphometric feature was identified using correlation analysis in SPSS version 22. It was evident from this investigation that F_f is very strongly correlated with Re. There was a strong correlation between D_d and F_s; between L_o, F_s and D_d; between D_i and F_s; between R_hp and B_h; between R_n, B_h and R_hp; and between C_c and R_c. On the contrary, D_t and R_lm have a very weak correlation. Similarly, there is a weak correlation between Fs and R_lm and D_d and R_lm. In the given correlogram, as shown in Table 5, the blue cell with a value of 1 and the red cell with value of −1 represent the strongest positive (1) and negative correlation (−1) between two characteristics, respectively. In Table 6, the component loading matrix is displayed. The top three components, all of which have eigenvalues of more than 1, account for 91.53 percent of the overall variance in the actual information and are clearly important. According to the rotational component matrix, component 1 is strongly correlated with drainage density (as seen by the bold font, which indicates the number is very close to either +1 or −1), component 2 is strongly correlated with relief (as seen by the bold font) and component 3 is strongly correlated with the RHO coefficient (Table 7) (Meshram, Sharma 2017, Shekar, Mathew 2022c). The three important morphometric features deduced from PCA are drainage density, relief and the RHO coefficient. Therefore, the 11 SWs of the Neyyar watershed are prioritised using these characteristics.

According to the three morphometric features (drainage density, relief and RHO coefficient) obtained using the PCA technique, Table 8 displays the final priority ranks and Cp values. SWs were categorised as high (Cp values within ≥ 2 to < 4), medium (Cp values within ≥ 4 to < 6) and low (Cp values within 6 and 8). Among the 11 SWs, SW1 falls under a high priority; SW2, SW3 and SW7 fall under a medium priority; and SW4, SW5, SW6, SW8, SW9, SW10 and SW11 fall under a low priority. Figure 11 depicts the final sub-watershed priority map for the Neyyar watershed.

The hypsometric curves capture the relief ratio and watershed volume, which are useful in providing excess rainfall and other hydrological processes (Keller, Pinter 1996, Vivoni et al. 2008). These graphs show not only the basin’s erosion status but also the tectonic, climatic and lithological elements that influence it (Sarp et al. 2011). Roy (2002) claimed that as the aspect ratio decreases, the river system becomes more branched, resulting in a greater bifurcation ratio. The drainage network and basin geometry have a great impact on hypsometry.

The HI was determined using the elevation-to-relief ratio approach proposed by Pike and Wilson (1971). In the current research, HI varies from 0.146 to 0.388 (Fig. 12). Low-priority (≥ 0.146 to < 0.226), medium-priority (≥ 0.226 to < 0.307) and high-priority (≥ 0.307 to < 0.388) SWs were identified (Farhan et al. 2016). SW1, SW7, SW8, SW9, SW10 and SW11 are high-priority SWs, SW2 and SW4 are medium-priority SWs, while SW3, SW5 and SW6 are low-priority SWs, as shown in Table 9. The final sub-watershed priority map for the Neyyar basin is shown in Figure 11.

Fig. 12.

In the present study, compared with three methods, some SWs have a common priority such as SW1, SW2 and SW5, and there is good correlation among them. Other SWs like SW3, SW4, SW6, SW7, SW9, SW10 and SW11 have a minimum of two methods with the same priority. For example, SW3 is having medium, medium and low here; this could be suggested as medium because the majority of the methods show similar priorities. Finally, SW8 is different in all its methods because there is no correlation among them, as shown in Table 10.

Traditionally, topographic maps or field surveys are used to determine the morphometric parameters. These variables are essential for drainage analysis since they are readily available, straightforward and affordable. However, the extraction of channel networks and the delineation of watersheds from topographic maps are not only time-consuming but also use non-digital methods. So, in order to merge with RS and geographic information system data, these data are digitalised. For computerised mapping and spatial analysis, GIS is a powerful resource. GIS can be used to establish watershed priorities in a less expensive, less time-consuming and less labour-intensive manner. In this current research, three methods have been used for the Neyyar River basin prioritisation, wherein novel strategies are introduced to rank the priority of SWs based on PCA and hypsometric analysis. The results obtained from three methods indicate that SW1 is a common high priority area with a significant risk of soil erosion, runoff and peak discharge. Hence, high-priority SWs need immediate attention for soil and water conservation measures.