Surface runoff water potential in the ungauged multi-basin Manchar Lake of Pakistan

This study estimated the runoff potential of Manchar Lake (Pakistan) as having enormous economic, environmental, and strategic value. In the recent floods, the lake overflowed, inundating surrounding villages. Had the timings and quantum of the expected flows known, the government could have minimized the damages by diverting the water to the sea. A hydrologic model can provide such information, but Manchar is ungauged, and model calibration/validation is challenging without observed data. An upstream Gaj basin had daily stream flow records only for four months. However, these data were insufficient in an ideal condition and were the only model calibration/validation option. For runoff simulation, Hydrologic Modeling System (HEC-HMS) was used. Land use/cover (Sentinel-2), soils (Food and Agriculture Organization), Digital Elevation Model (Advanced Land Observing Satellite), and hydro-meteorological information were model inputs. The Gaj basin model simulated well with coefficients of determination (R2 = 0.92/96), Nash–Sutcliffe efficiency (NSE = 0.82/0.85), root mean square error (RMSE = 159.4/98.9 ft3/s), RMSE-observation standard deviation ratio (RSR = 0.42/0.39), and percent bias (PBIAS = 8.01/31.75%) during calibration/validation periods. The calibrated parameters were then applied to the entire watershed using the regionalization concept of spatial proximity. The simulated runoff (mean annual volume) entering the lake from 2004 to 2020 was 1.662 ± 0.871 MAF per year and during the 2022 floods (June–September) was 4.28 MAF. This approach is helpful in water and land resource planning and managing ungauged catchments during dry periods and will significantly reduce flood damages through early warnings and timely flood management.