Projections of Selenga river runoff in the XXI century and uncertainty estimates

Abstract. The Selenga River, which originates in Mongolia, contributes nearly 50 % of the total inflow into Lake Baikal. Since 1996, the longest low-flow period has been observed. This paper focuses on the spatially distributed process-based modeling to assess possible runoff changes under climate projections in the XXI century using an ensemble of global climate models (GCMs) from ISI-MIP2 (Inter-Sectoral Impact Model Intercomparison Project, phase 2) and RCP-scenarios (Representative Concentration Pathway) as inputs. The ECOMAG hydrological model was applied to simulate possible runoff changes in the Selenga River basin. According to the simulations, the low-flow runoff into Lake Baikal will be continued throughout the XXI century under all the RCP-scenarios. Furthermore, as hydrological projections for the future are characterized by significant climate projections uncertainty, ANOVA (analyses of variance) test was used to quantify the sources of this uncertainty. The calculations showed that the contribution of the model uncertainty (differences between GCMs) is much higher than the contribution of the scenario uncertainty (variability of RCP-scenarios) although it increases at the end of the century.