Satellite data was obtained from online archives and pre-processed with atmospheric, geometric, and picture mosaic corrections. These images were resampled to a spatial resolution of 30 m using the average resampling method in ArcGIS (version 10.8). To study the condition of water bodies in the study area, Normalized Difference Water Index (NDWI), was used to generate streams from satellite data. Apart from NDWI, two other indices were also used, namely, Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI).

It is an index method used in remote sensing analysis of satellite imagery to identify open water features 30. The method involves two bands from the EMR spectrum, namely, near-infrared (NIR) and visible green (GREEN) spectral bands. The NDWI can be calculated using the Equation 1 as follows: 

NDVI is a very popular index method employed for evaluation of the vegetation conditions through remote sensing and satellite imagery 31. The spectral bands used in the computation of NDVI are NIR (near-infrared) and red bands of satellite imagery such as Landsat. It can be calculated using Equation 2 as mentioned below:

The normal range of NDVI values lies between -1.0 to +1.0. The NDVI classifications are shown in Table 3. It is commonly observed that the value range of NDVI for non-vegetated surfaces hovers between 0.1 (NIR < Red) to as high as 0.9, as determined by plants and other Earth’s surface components. Factors contributing to higher NDVI values include increased green biomass, advantageous seasonal shifts, and beneficial rainfall 32.

This index can be well employed to assess the density and health of vegetation, mainly in densely vegetated areas. It is more susceptible to biomass, atmospheric background, and soil quality than the Normalized Difference Vegetation Index (NDVI), yet it is still an improvement. EVI calculates a number using the blue, red, and NIR (near-infrared) bands; values closer to 1 denote healthy vegetation. The general range of EVI value is usually between -1 and +1. It can be calculated using Equation 3 as mentioned below:

where  G is a gain factor, L is a soil adjustment, C1 and C2 are coefficients for corrections for atmospheric conditions 32.

The temporal map of precipitation for 15 stations in the Mangura Nala watershed was created using a bar graph. The respective period of variation of precipitation was considered against each rain Gauge station. The temporal fluctuation of rainfall in and around the Mangura Nala watershed varies from station to station. We chose 15 rain gauge sites based on Earth data climatic location, as shown in Figure 2. The average rainfall from Amethi to Chapri stations is more than 900mm, with the pre-monsoon having lower rainfall than the post-monsoon. The monsoon season brings the most rainfall. In comparison to the rest of the station, Bhodh Chak and Burhgharea received the most rainfall. The temporal variance of rainfall in 2013 is greater compared to other periods in the monsoon, roughly 2750 mm in August at the Burhgharea station. The seasonality of rainfall in 2019 and 2020 is larger compared to previous periods in the monsoon, is about 2000 mm in Bhod Chack, Bhagawanpur stations, and Burhgharea stations from June to September. The seasonality of rainfall variation in the entire watershed has a direct impact on the vegetation and water bodies.

Using the NDVI, the relationship between spectral vegetative variability and variations in vegetation growth rate was thoroughly investigated. The results of this study demonstrated that NDVI values differed with time. Table 4 and Figure 3, Figure 4 indicate the spatio-temporal variation of NDVI and EVI from 2013 to 2023. Vegetation indices have shown a significant influence on rainfall averages. The high NDVI suggests a more sustainable climate for photosynthesis.

An overview of the overall rainfall distribution is provided by the average rainfall for the year from 2013 to 2023 (Figure 2). Any departure from this means or the appropriate season is crucial since it provides information on the overall biomass and the health of the plants. The surface vegetation region's pre-monsoon NDVI values ranged from 0.9 to -0.20. The 2013 monsoon season had the most surface vegetation cover (0.99), followed by the post-monsoon season (0.97) and the pre-monsoon season (0.91), showing that 2013 had the most surface vegetation cover during the monsoon season. From 2015 to 2021, the monsoon season had the greatest average vegetation surface value, ranging from 0.98 to 0.96. When compared to prior study years, the lowest NDVI value in 2021 was -0.19 during the monsoon season, while the highest value during the monsoon season was around 0.98. The Mangura Nala watershed in general shows an increase in vegetation growth during the monsoon season except in the north-western portion, which comprises of hilly region. The spatial distribution makes it evident that areas with high NDVI and EVI are primarily covered by dense vegetation (Figure 3, Figure 4). Due to the establishment of the seasonal crop primarily in plain areas, the monsoon NDVI has an excellent NDVI distribution in most regions (Figure 3). In 2013, the surface vegetation region acquired pre-monsoon EVI values ranging from 0.60 to 0.09. The 2013 monsoon season showed 0.95 to 0.01, while the post-monsoon showed 0.45 to 0.10, indicating that 2013 experienced higher surface vegetation cover during the monsoon season.

Figure 4 and Table 4 depict the spatio-temporal variability of seasonal changes. From 2013 to 2023, the monsoon season has shown the highest vegetation indices. When compared to other research years, the lowest EVI value in 2015 was -0.02 during the pre-monsoon season. Whereas the maximum value during the monsoon season was approximately 0.99 in 2017. Overall, EVI maximum values during the monsoon season were around 0.99, whereas post-monsoon seasons ranged from 0.35 to 0.66. Similarly, the EVI with positive deviation demonstrates an excellent crop pattern, which is also strongly supported by the positive rainfall deviation distribution.

Normalized Difference Water Index (NDWI) was used to examine the temporal and spatial variability of surface water. McFeeters (1996) states that values greater than 0 signify that water covers the area. The accuracy evaluation showed that the Mangura River sub-basins' surface water bodies were successfully captured by the NDWI. Table 4 displays the NDWI values, which were obtained from the analysis for the years 2013-2023. In general, the NDWI and ground-truthing data showed full agreement. These findings are in line with several studies that showed the good performance and dependability of NDWI in mapping surface water bodies. Significant temporal fluctuations in the watershed surface water area are revealed by the NDWI results in Figure 5. In 2013, the region covered by surface water experienced pre-monsoon NDWI ranging from -0.20 to 0.89. The 2013 monsoon season showed -0.20 to 0.99, while the post-monsoon showed -0.20 to 0.96, indicating that 2013 recorded the highest surface water cover during the monsoon season. Figure 5 and Table 4 indicate the respective years' data and changes. In 2013, the monsoon season has the maximum water surface value of around 0.99. In 2021 and 2023, this area shows maximum values in the monsoon and post-monsoon seasons of 0.97 to 0.98 and 0.96 to 0.94, respectively. In the Mangura watershed region, variations in surface water are generally closely related to patterns of rainfall. However, this link has become more intense in the 15 rain gauge stations of Mangura watershed due to changes in land use and rainfall variation in the area.

Notably, natural terrain is fast being converted into agricultural and urban areas, resulting in less water reaching the subsurface to replenish aquifers and maintain stream baseflow. As a result, surface water bodies are becoming increasingly dependent on rainfall-driven runoff. The impact of rainfall on water bodies was assessed using the rainfall deviation method. The image includes clouds, although their impact on water extraction is minimal. As shown in Figure 5, NDWI computation can identify areas with higher concentrations than water because of the cloud's reflectance. The experimental approach is like that of an urban river area, as the shadows cast by clouds align with the solar azimuth. Figure 3, Figure 4, Figure 5 depict the entire method, while Table 2, Table 3 show the experiment's accuracy. These results offer significant new information for the watershed's environmental preservation, agricultural planning, and water resource management. The regional variability of drought intensity identifies regions that may benefit from focused efforts to lessen the effects of the drought and allow for more intelligent resource allocation. In this regard, framing policies in liaison with the prevailing conditions can help in ensuring adequate supply of drinking water as well as to maintain agricultural productivity, ultimately aiming towards a sustainable and climate-resilient future in the study area.