Quantification of uncertainties in streamflow extremes in the Chaliyar river basin, India under climate change

Projections of future streamflow extremes are subjected to various sources of uncertainties. Uncertainties arising from different stages in the modelling process of extremes contribute to the overall uncertainty in the final estimates. In this study, some important sources of uncertainties that occur in the modelling chain for projection of future extremes are analysed. The contribution of climate models, hydrological models, representative concentration pathways (RCPs) and the parameters of the extreme value distribution to the overall uncertainty in the streamflow return levels of the Chaliyar river basin in Kerala, India, was quantified using analysis of variance. 5, 10, 25, and 50-year return levels for the near (2021-2050) and far (2070-2099) future periods were computed. Results of the study indicate that the median values of return levels for the near (0.6-13.5%) and far (21.5-37.3%) future periods are higher than the corresponding values for the observed data. Also the uncertainty ranges of the return levels in the near (2-2.4 times) and far (3-3.7 times) future for different return periods are higher compared to that of the observed. The contribution to uncertainties in the return levels from climate models, parameters of the extreme value distribution employed and hydrological models is higher than that from the RCPs. The contribution of parameter uncertainty increases with the return period and that of the climate models reduces with the return period for the two future periods considered in this study, indicating that parameter uncertainty is an important source of uncertainty and has to be considered while predicting the future extremes. The contribution to uncertainty by the hydrological models and by the RCPs does not exhibit much variation for different return periods in both the future time periods considered. Understanding the uncertainties in future extremes would help to devise proper adaptation strategies to mitigate the probable impacts of climate change.