Impacts Of Climate Change On Runoff And Their Uncertainty With Contributions From Internal Climate Variability

Global climate change has already aroused a series of impacts on ecological environment and social economics.Climate change gives rise to changes in hydrological impacts,thus influencing regional water resources distribution.Climate change projections are generated along with multiple sources of uncertainty,mainly consisting of scenario uncertainty(S),climate model response uncertainty(M)and internal climate variability(V).The existence of former two is due to a lot of assumptions about future greenhouse gases emissions scenarios,and a lot of climate models to simulate the climate system.While V is a property intrinsic to the climate system.Climate change projections with uncertainty are input into hydrological models.Due to differences among hydrological models,the uncertainty is further amplified as runoff impact uncertainty.However,in assessment studies on runoff impacts of climate change,uncertainties from S,M and hydrological model(H)are commonly analyzed,while V has been paid less attention.Therefore,this study is aimed at investigating runoff impact uncertainty under climate change in mainland China,taking V into account.Detailed study contens and conclusions are as follows:(1)This study proposes a method of using an initial condition ensemble from a single climate model to estimate V.Based on this method,this study has investigated climate change projection multi-source uncertainty taking V into account.And this study also investigates the significance of climate change taking V as a reference.Results show that internal precipitation variability changes significantly with time.This may imply an advantage of using initial condition ensembles to estimate V.Before 2050 s,V and M are main sources of climate change projection uncertainty in mainland China,particularly for precipitation change.S gradually becomes a main uncertainty source after 2050 s with its contribution increasing over time,especially for temperature change.In terms of spatial variations,temperature change projection uncertainties tend to have high values in cold regions(e.g.northern China and Qinghai-Tibetan Plateau).Precipitation change projection uncertainties tend to have high values in northwestern dry regions.Analysis of significance of climate changes show that: Average temperature changes are higher than their internal variabilities after 2020 s.Annual precipitation change in most northern and partial southern regions become higher than its internal variability since 2030 s to 2040 s.Winter and summer precipitation changes in parts of northeastern,central and southwestern China become significant since 2030 s.Annual and summer maximum daily precipitation changes in partial southwestern China become significant since 2030 s to 2050 s.Internal precipitation variability overwhelms precipitation changes in other regions throughout the 21 st century.(2)This study has established hydrological simulations at basin scale in mainland China to look into the propagation of V in runoff impacts.Meanwhile,it makes use of multiple monthly water balance models,to take hydrological model structure uncertainty into account.And it investigates the significance of changes in runoff impacts by using its uncertainty due to V as a reference.For each hydrological model,it first conducts calibrations and validations to get calibrated parameters in 63 gauged watersheds.Based on this,parameter values for 214 watersheds in mainland China are gained by using a parameter regionalization method in ungauged regions.This method is based on climate and geographical characteristics of watersheds.Results show that multiple sources of runoff impact uncertainty keep growing in most regions of mainland China in the 21 st century.In terms of spatial variations,runoff impact uncertainty due to V has the highest value in western China(Xinjiang,Tibetan,Qinghai inflow areas).Its relative contribution reaches 45%-90% in most regions except partial northern regions in the beginning of the 21 st century while decreases with time.The other three sources have the highest values in northern part(Xinjiang and Ganmeng inflow areas).Relative contributions of S and H are lower than 25%.The contribution of M exceeds 85% in northern China,and between 45%-75% in southern China since 2050 s.Annual and summer runoff changes in northeastern China are significant since 2070 s to 2080 s.Runoff impact uncertainty due to V in other regions overwhelms runoff changes throughout the 21 st century.Overall,internal climate variability is an essential source of climate change projection uncertainty.It even exceeds uncertainties from climate model response,greenhouse gases emission scenarios and hydrological model structure before 2050 s.Therefore,it is essential to fully consider contributions from internal climate variability in assessments of runoff impacts of climate change and constitution of adaptation strategies at basin scale.