Degradation status and water-quality assessment of the lakes in and around Hyderabad (India) using multi-temporal remote sensing data and GIS

The cratonic southern India is known for a large number of freshwater lakes/ponds. Devoid of any major river in these regions and characterised by scanty rainfall, these lakes have been the major source of water supply for human consumption. With increasing population, the lakes are getting degraded physically (e.g., by disconnectivity, encroachments, siltation, land-use change) as well as in terms of their water-quality (due to pollutant inflows). In this work, we have assessed the degradation of the lakes (n = 2815) in and around the Hyderabad metropolitan area in south India. We have used lake full tank level (FTL) boundaries as a baseline and Sentinel-2 time-series dataset to assess the physical degradation of the lakes. We have compared the annual and seasonal water area and vegetation variations with the FTLs and defined an index, DIP, to assess the physical degradation of each lake for the period 2013-2020. The assumption behind DIP is that the FTL is the ideal condition of the lakes. So, if the terrestrial vegetation area in the lake is increasing, or if the water area itself is decreasing with respect to FTL, the lake is getting degraded. We have also used Sentinel-2 based multispectral indices for vegetation (GRVI), turbidity (NDTI), CDOM, electrical conductivity (EC), and red tide (RT) to define a water-quality based degradation index, DIWQ. Using the DIWQ, we estimated the water-quality status of each lake at pixel-scale (10 m) resolution at monthly scale for years 2019-20. Our results of DIP show that in the year 2013-14, 1867 lakes were highly degraded, 448 were moderately degraded, and 500 were least degraded in terms of physical degradation. In comparison, 1480, 819, and 516 lakes were highly, moderately, and least degraded respectively in the year 2019-20. The water-quality of 667 lakes were of least degraded status, 1400 of moderately degraded status, and 748 of highly degraded status in the December of 2020. In comparison, these numbers were 10, 2716, and 89 for least, moderately, and highly degraded lakes in January 2020. Most of the lakes with least values of DIWQ are situated in the northern parts of the study area and the lakes with moderate and high values are evenly distributed throughout the study area. The results exhibit that time-series remote-sensing dataset in a GIS framework can be used for the assessment of the lake status. The method developed in the work can be applied to any other setting for the assessment of physical degradation and water quality which is the first step towards developing a restoration and management plan.