Impact of renewed solar dimming on streamflow generation in monsoon dominated tropical river basins

Abstract

From 1950s to 1980s, various observational studies around the globe found a significant decrease in surface solar radiation (SSR), which reversed in late 1980s for most of the countries including India. SSR observations at 12 stations located across India revealed that a much stronger dimming has reappeared during the last decade (2006–2015) after a brightening during 1996–2005. In the present study, effects of renewed solar dimming on actual evapotranspiration and runoff were analyzed using a semi-distributed hydrological model, Soil and Water Assessment Tool (SWAT) in 24 river basins (ranging from 1260 to 40000 km2) located in peninsular India. For these river basins, calibration (2003–2009) and validation (2010–2014) were performed using the observed daily discharge data, obtained from water resources information system (WRIS) of India, with a 3 year warm up period (2000–2002). The sequential uncertainty domain parameter fitting algorithm (SUFI-2) of SWAT-CUP (calibration and uncertainty program) was used with modified Nash–Sutcliffe efficiency (MNS) as the objective function to calibrate 13 model parameters, which can potentially affect streamflow. In nearly all the river basins, the p- and r-factor of 95 percentage prediction uncertainty (PPU) were more than 0.7 and less than 1, respectively. At daily timescale, MNS values were more than 0.5 in most of the river basins, reaching up to 0.66 and 0.71 during calibration and validation periods, respectively. Calibrated model was used to analyze the water balance of these river basins and different sets of experiments (with observed SSR trends) were performed to find SSR impacts on it. The model was simulated with and without the observed declines in SSR trends. The average change in SSR (in terms of evaporation equivalent) was −267.93 ± 100.92 mm/day/year (−5.62 ± 2.12%) with maximum reaching up to −417.12 mm/day/year (−8.99%). Due to this SSR change, actual evaporation was reduced resulting in 18.97 ± 9.78 mm/day/year (4.13 ± 2.50%) change in percolation. The percolation changes were higher for river basins having areas covered by forests and cropland/woodland, and having loam and sandy-clay soils. The increase in runoff generated was 6.90 ± 3.42 mm/day/year (2.14 ± 1.58%) with a maximum of 15.25 mm/day/year (7.56%) whereas corresponding increase in streamflow was found to be 9.93 ± 5.27 mm/day/year(4.21 ± 2.38%) with a maximum of 26.71 mm/day/year (11.86 %). The study reveals that the recent observed SSR changes are significant enough to have resulted in increased streamflow in the monsoon dominated tropical river basins of India.