Google Earth Engine for Mapping Rooftop Rainwater Harvesting Potential Using Random Forest Classifier and Spectral Indices

As per the 2030 Agenda related to sustainable development, countries worldwide are necessitated to put substantial efforts in improvising the quality of life of their citizens. One such step is to abate the water stress through rainwater harvesting. The main objective of the study is to estimate the Rooftop rainwater harvesting potential in the study area of Hyderabad city by using Gould and Nissen Formula (1999) in integration with Google Earth Engine (GEE) datasets. The Climate-Hazards-Group-InfraRed-Precipitation-with-Station-data (CHIRPS) dataset has been used for annual rainfall and Sentinel-2A image collection for settlement mapping. Built-up area indices that are utilized in the present study include band ratio for built-up areas (BRBA), new built-up index (NBI), normalized difference building index (NDBI) and the urban index (UI). These spectral indices are involved in the extraction of settlement areas. These spectral indices were thoroughly incorporated through a random forest supervised procedure. From the study, it can be concluded that Random Forest classifier is worthy enough to extract spatial distribution of settlements in Hyderabad region (India) with the accuracy metric kappa coefficient, exceeding 90%.