Hydrologic impacts of climate change: Quantification of uncertainties

Abstract

The three visible signals of climate change, viz., increase in global average temperature, change in precipitation patterns and rise in sea levels, are known to cause a significant impact on regional hydrology, at different spatial and time scales. Processes such as stream flow and evapotranspiration, magnitudes and frequencies of hydrologic extremes of floods and droughts are likely to be affected by climate change. A commonly adopted procedure for assessing the climate change impact on hydrology is to use the projections provided by the global scale General Circulation Models (GCMs) and downscale the hydrometeorological variables to regional scales, and consequently execute the distributed hydrologic models to obtain hydrologic projections. Significant uncertainties are imparted due to the use of several models at different spatial and time scales. Quantification of such uncertainties at all levels - from the choice of climate models and emission scenarios to downscaling and hydrologic models - is a current area of research. This paper summarises the recent work carried out by the second author's team on quantification and reduction of uncertainties in assessing hydrologic impacts of climate change. Both aleatory (irreducible) uncertainties that arise from the inherent uncertainties due to randomness, and the epistemic (reducible) uncertainties that are characterised by lack of knowledge are addressed. Applications to impact on streamflow, evapotranspiration, pluvial and fluvial floods and droughts are discussed. A brief discussion on the changing frequencies of concurrent extremes of droughts and heatwaves in India is also provided.