Uncertainty in flow and sediment projections due to future climate scenarios for the 3S Rivers in the Mekong Basin

Reliable projections of discharge and sediment are essential for future water and sediment management plans under climate change, but these are subject to numerous uncertainties. This study assessed the uncertainty in flow and sediment projections using the Soil and Water Assessment Tool (SWAT) associated with three Global Climate Models (GCMs), three Representative Concentration Pathways (RCPs) and three model parameter (MP) sets for the 3S Rivers in the Mekong River Basin. The uncertainty was analyzed for the short term future (2021-2040 or 2030s) and long term future (2051-2070 or 2060s) time horizons. Results show that dominant sources of uncertainty in flow and sediment constituents vary spatially across the 3S basin. For peak flow, peak sediment, and wet seasonal flows projection, the greatest uncertainty sources also vary with time horizon. For 95% low flows and for seasonal and annual flow projections, GCM and MP were the major sources of uncertainty, whereas RCPs had less of an effect. The uncertainty due to RCPs is large for annual sediment load projections. While model parameterization is the major source of uncertainty in the short term (2030s), GCMs and RCPs are the major contributors to uncertainty in flow and sediment projections in the longer term (2060s). Overall, the uncertainty in sediment load projections is larger than the uncertainty in flow projections. In general, our results suggest the need to investigate the major contributing sources of uncertainty in large basins temporally and at different scales, as this can have major consequences for water and sediment management decisions. Further, since model parameterization uncertainty can play a significant role for flow and sediment projections, there is a need to incorporate hydrological model parameter uncertainty in climate change studies and efforts to reduce the parameter uncertainty as much as possible should be considered through a careful calibration and validation process. (C) 2016 Elsevier B.V. All rights reserved.