Assessment of Drought vulnerability through an integrated approach using AHP and Geoinformatics in the Kangsabati River Basin

3.1 Analytical hierarchy process

The AHP method was used to generate a pairwise matrix of conditional variables. The AHP method is a well-known method for solving complex problems related to drought vulnerability (Palchaudhuri and Biswas, 2016; Ekrami et al.,2016; Sivakumar et al.,2021). Different researchers have used pairwise rating methods in flood risk studies to capture the significance and contributions of each conditional element in flood mapping (Vojtek & Vojteková, 2019). All the parameters were ordered hierarchically in this approach to allow pairwise comparison and their relative significance was ascertained using a comparative scale suggested by Saaty (1980) consisting of integer numbers ranging from 1 to 9 (1 indicating equal importance and 9 refers to extreme importance). A 11 × 11 pair wise reference matrix of diagonals, it compares the value of each unit, elements are equal to 1. The remaining values in each row reflect the relative significance of the remaining variables. Rank of each parameter was determined using pair wise comparison matrix, the relative weight matrix, and the normalized principal Eigen vector (Table 1).

The estimation of the normalized principal eigen vector based on the comparison matrix was calculated by dividing the summation of column values with the relative weight matrix. The normalized principal eigen vector was obtained by averaging the rows, and then the effect percentages for each thematic layer based on the normalized principal eigen vector were calculated. The consistency ratio (CR) was used to check the accuracy of the relation, where less than or equal to 0.1 implies an appropriate reciprocal matrix while a ratio greater than 0.1 indicates that the matrix should be modified (Sani Yahaya et al. 2010). It is determined as:

3.2 Weighted overlay analysis

The weighted overlay analysis is an easy, simple, and sufficient approach for evaluating possible drought hazard areas. Eleven environmental factor maps were used to prepare a drought risk zonation map of the study area. In order of significance, those factors were given a numerical proportional rating on a 1–5 scale. The classes of the factor were also assigned weights and grades, with higher weights and ranks showing a greater impact on drought occurrence. To create the drought hazard zone map, these factors were overlaid as thematic layers in GIS using the weighted overlay method (WOM). Finally, the drought map was prepared by adding all the parameters after multiplication to their weights (Eq. (3)). Drought map was categorized into five zones: very high, high, medium, low and no drought.

4.1 Elevation, slope, aspect and LULC

The availability of water is also heavily influenced by the plain's elevation. The digital elevation model (DEM) represents landscapes and ground surface topography and is critical for overcoming problems such as climate change impacts, emergency management, environmental management, and water security. This study has utilized a 30-meter-resolution SRTM DEM dataset (USGS). High elevations are observed in the north region, while the area in the south region is mostly a low-elevation zone (Fig. 3a). The descriptive statistics shows the mean sea level ranges from −1 to 659 m, having mean and standard deviation as 126.69 and 107.83 m respectively. Slope is another significant drought risk parameter that quantifies the inclination of the ground surface from the horizontal. When compared to the nearby ground surface, water runoff is significantly higher over steeper terrain. As a result, as compared to steep plains, terrain regions with lower slopes are less sensitive to droughts (Jain et al. 2015). The descriptive statistics shows the slope in degree ranges from 0 to 89.99, having mean and standard deviation as 80.11 and 28.11 respectively. Fig. 3b shows the spatial variation of the slope percentage. Aspect refers to the slope direction, and different aspects have varying degrees of sun radiation which influence the hydrology of the study area (Fig. 3c). This parameter is influenced by hydrology. The research area's aspect map was separated into several classifications (Bennie et al. 2008). A value was determined for each class based on the influence of temperature, humidity, infiltration, runoff, and soil evaporation. Using relevant literature and local specialized knowledge, vulnerability classifications for various aspect ranges were developed.

Close

Fig. 3

Thematic layers for drought analysis: a) elevation, b) slope, c) aspect and d) LULC.

Export to PPT

Drought susceptibility can be seen dynamically as a function of land use and management, involving government farming methods and societal issues. Due to its dynamic character, it is regarded as an exposure variable for the study. The land use data were collected from the Esri 2020 Land cover (Karra et al., 2021). The dominant land use in the study area was identified by crop (63.74 %) followed by trees (18.62 %), built area (9.45 %), scrub/shrub (4.25 %), water (3.37 %), grass (0.35 %), barren land (about 0.16 %) and flooded vegetation (0.06 %) as shown in Fig. 3d. Because of their dependence on water for living, agricultural (urban use) is implicitly the first (second) victim due to water scarcity and drought among the eight stated land-use groups. As a result, the agricultural category has the largest number weight value since it is thought to be more sensitive to drought.

4.2 Population density, LST, NDVI and NDWI

Population density has an impact on water consumption as well. Water use and demand are both high in places with dense populations. As a result, regions with a higher population are thought to be more sensitive to drought than those with a lower population. Population density is classified as an exposed factor in this study because while the population expands, so does the demand for water and the strain on scarce water supplies. The constant rise in population density will have an influence on water supply in any given location, increasing the risk of drought. The population density estimates for 2020 were acquired from NASA Socio Economic Data and Application Centre (SEDAC) (CIESIN, 2022) (Socioeconomic Data and Application Centre) (Fig. 4a). In contrast, the population density in the southeast study region is quite high. This suggests that the region's high population density and rapid population growth rates increase the likelihood of water shortage or water stress conditions. Land surface temperature (LST) produced from TIR band data has been discovered to give critical and valuable information on the status of the land surface, and it is commonly used in calculating water and energy budgets at the surface-atmosphere interface (Gutman, 1990). LST (Fig. 4b and Table 2) is used as a proxy in this context to quantify evapotranspiration, vegetation water stress, soil moisture, and thermal inertia (Moran et al., 1994).

Close

Fig. 4

Thematic layers for drought analysis: a) population density, b) LST, c) NDVI and d) NDWI.

Export to PPT

NDVI can be utilized as a response variable in semiarid and arid environments to identify and quantify drought disturbance, with low values correlating to stressed plants (Tucker & Choudhury, 1987). NDVI index and a Landsat 8 ETM + image was used to map the plant cover in this study. The Normalized Difference Vegetation Indicator (NDVI) is a vegetation index that measures the amount of photosynthesis occurring in plants or their greenness (Fig. 4c). This vegetation index is used rather commonly. Greenness can be calculated by turning a remotely sensed image's red and near-infrared bands into a vegetation index and dividing by the resulting number. NDVI is computed for every image, whether it has a red and near-infrared wavelength band or not (Table 2). NDVI images vary from −1 to + 1, with −1 represents saturated water and + 1 corresponds to healthy vegetation (Myneni et al., 1995). When making this inquiry, just the density of vegetation cover would be considered (Myneni et al., 1995). Water stress can influence the vegetative canopy during a drought episode. This can have a significant influence on plant development in general, resulting in crop failure or poorer crop yield in agricultural regions. Early detection of plant water stress can help to avoid such outcomes. The Normalized Difference Water Index (NDWI) has been shown to be highly linked to plant water content (Fig. 4d). As a result, it is an excellent predictor of plant water stress. Several studies have proved its utility for drought monitoring and early warning (Gu et al., 2008). It is calculated using near infrared (NIR) and short-wave infrared (SWIR) reflectance, making it sensitive to changes in liquid water content and spongy mesophyll in plant canopies (Gao, 1996; Perveen et al., 2016).

4.3 NDDI, VCI and SMI

Normalized Difference Drought Index (NDDI) can provide an appropriate measure of the dryness of a specific area, because it combines information on both vegetation and water (Fig. 5a and Table 2). NDDI combines information from the NDWI (Normalized Difference Water Index) and NDVI (Normalized Difference Vegetation Index) data. The first is an index generated from the visible red and near infrared channels, evaluating variations in chlorophyll absorption and reflection in the vegetative canopy's spongy mesophyll that are reflected in these bands. (Renza et al., 2010) used Landsat to apply the index to diverse vegetation in Spain and found that it performed better than other indices when sensing drought conditions, such as NDVI or NDWI alone. NDDI takes advantage of the fact that NDVI detects plant matter while NDWI detects plant wetness. The vegetation condition index (VCI) established by Kogan (1997) normalizes NDVI pixel by pixel, scaling between minimum and maximum NDVI values (NDVI min and NDVI max, respectively), as recorded at each pixel over a lengthy temporal record since 1981. NDVI′ is the average NDVI for a composite time period (which can be a decade, growing season, Week, month, or a year). The normalization emphasizes relative changes in the local NDVI signal across time while minimizing the impact of spatial and temporal variability in NDVI phenology between various land cover types and meteorological circumstances. This measure is used to estimate vegetative health and monitor drought (Fig. 5b). The SMI is used to assess soil moisture conditions during a drought (Fig. 5c). It is an excellent predictor of drought. The dataset was obtained from the Meteorological & Oceanographic Satellite Data Archival Centre (MOSDAC) having a spatial resolution of 40 km. The SMI results are obtained Pre-monsoon of the year 2018 from the SMAP l-band radiometer.

Close

Fig. 5

Thematic layers for drought analysis: a) NDDI, b) VCI and c) SMI.

Export to PPT

4.4 Spatial distribution of drought vulnerability in the Kangsabati river basin

Drought vulnerability analysis revealed that most of the basin, nearly 28.5 % of its area, falls under the medium drought category, followed by the high (21.1 %), no drought (20 %), low (19.5 %), and very high (10.6 %) categories. High NDVI, high to moderate land surface temperature, moderate elevation, and a steep slope have resulted in high to very high drought vulnerability. These areas are also characterized by built-up areas and moderate population density. Catchment-wise areas under different drought categories are presented in Table 3. The maximum areas under high and very high drought risk can be identified in catchment 2 (9.6 %), catchment 3 (8.9 %) and catchment 1 (5.5 %) illustrated in Fig. 6. Similar study was also carried out over few parts of the study area (Palchaudhuri and Biswas, 2016; Ghosh et al.,2020).

Close

Fig. 6

Spatial distribution of drought in the study area.

Export to PPT

Apart from drought risk, some critical problems can be identified, such as soil dryness, ground water table decline, water scarcity, drought, soil erosion, deforestation, crop failure, and lastly, environmental imbalance in the upper catchment area. The basin is situated over a transition zone between the Chotanagpur plateau and the river plain. This peculiar lithological setting may be responsible for these problems. The upper catchment area is situated in Purulia and Bankura districts, which are famous for dry and extreme climatic zones and have high potential for water scarcity. The undulating hard rock terrain has infertile lateritic soil with a low water holding capacity. Kangsabati itself is an intermittent river which propagates dryness and aridity in the region. Due to the unsupported circumstances, land erosion and soil fertilization dominate the basin. In these unfavourable physical conditions, agricultural practices and other primary activities get hampered, and people are facing meteorological and agricultural drought almost every year. The gully erosion accelerates with the arrival of the monsoon, and the loose materials deposit in the river bed. Consequently, the lower catchment is severely flooded with a huge economic damage. Kangsabati River basin is such a tropical river that fosters both drought and flood every year. To protect the basin, the government took initiative by making a barrage over Kangsabati near Mukutmanipur. It was planned to supply the reserve water to the upper catchment area in the dry season by the canal system, which would help for both agricultural use and domestic use. Likewise, it may control the lower catchment flood by holding the extra water during the rainy season. But recent studies have mentioned that the barrage failed to prevent the hydro-meteorological hazards, and the Kangsabati command area is at great risk both environmentally and economically.

Various plans, strategies, and techniques have already been applied, but they failed to mitigate these problems. In this circumstance, a nature-based solution (NBS) is the best alternative to mitigate HMH as a long-term solution by developing the adaptive capacity of nature. Two types of problems have been recognized: 1) agriculture and ecosystem-related, and 2) soil and land degradation as a result of groundwater decline. Long-term solutions can be found by using a pair of NBS in the Disaster Risk Reduction (DRR) strategy, such as an ecosystem-based approach (EBA) and blue infrastructure for drought risk reduction. To mitigate the vegetal degradation and prevent the ecosystem extinction, ‘Climate Smart Agriculture’ can be adopted which includes rain fed agriculture and contour trenching for restoration of agricultural land and riparian forest ecosystem and reforestation as a safeguard of Ecosystem and habitat protection. On the other hand, blue infrastructure may be obtained for groundwater table recharge and land reformation which includes Groundwater Based Natural Infrastructure and Integrated Dam Operation Management for the sustainable use of natural assets. All the above strategies are highly suitable for the basin management and they are interconnecting and overlapping accordingly.