Click a lake to see its catchment, drainage network, rainfall history, and rooftop harvest potential.
Click anywhere else to see where water from that point drains — which lake it reaches and which river it eventually flows to.
Catchments. A catchment (or watershed) is the area of land where all surface water drains to a single lake. Catchments are delineated from a digital elevation model (FABDEM, 30m resolution) using D8 flow direction analysis. The DEM is processed at Telangana state scale: lake beds are burned into the terrain to ensure flow convergence, pits are filled, flats are resolved, and flow direction and accumulation grids are computed. Each lake's catchment is then traced upstream from its outlet cell using breadth-first search on the flow direction array.
Rainfall. Daily rainfall data comes from Telangana's Automatic Weather Station (AWS) network (~1,090 stations). Each lake's rainfall is computed as an area-weighted average using Thiessen (Voronoi) polygons: the HMDA region is divided into cells, each assigned to its nearest weather station. For each catchment, the fraction of area falling within each station's cell determines the weight. The result is a spatially interpolated daily rainfall value specific to each catchment — not a simple city-wide average. The HMDA-level rainfall shown on the home page is the average across all lake catchments for each date.
Rooftop harvest. The rooftop harvest estimate represents the volume of rainwater that could theoretically be captured from building rooftops within a catchment. It is calculated as:
Harvest = rooftop area × rainfall depth × runoff coefficient
Rooftop area is the total building footprint area within the selected catchment, derived from Overture Maps building data (~2.5 million buildings in HMDA). The rainfall depth is the cumulative catchment-specific rainfall over the selected date range. The runoff coefficient (0.8) accounts for evaporation and other losses. The estimate changes with the selected date range — it is not annualised. Each lake panel shows harvest for that specific catchment; the home page shows the total across all HMDA catchments.
Drainage network. The lake-to-river drainage graph traces how water moves from each lake to the next, and ultimately to a named river. From each lake's outlet, the D8 flow direction is followed downstream cell-by-cell until another lake or a river is reached. This produces a directed graph of ~32,000 edges connecting ~3,500 HMDA lakes to their downstream lakes and terminal rivers. Stream paths follow actual terrain-derived channels, not straight lines between lakes.
Streams. The stream network is extracted from the flow accumulation grid using a threshold of 50 contributing cells. Stream order is computed using Strahler classification. Streams are tagged with the catchment they belong to, so only relevant streams are shown when a lake is selected.
Data sources. Elevation: FABDEM (Forest And Buildings removed Copernicus DEM), 30m. Buildings: Overture Maps Foundation. Rivers and lake polygons: Telangana state geospatial data (EPSG:7756, reprojected). Rainfall: Telangana AWS daily observations. Administrative boundaries: HMDA ward, circle, and corporation shapefiles.