Advances in hydrological frequency analysis of non-stationary time series

The conventional methods of flood frequency analysis are based on the assumption that the hydrological time series is statistically independent and identically distributed.The latter assumption implies that the series is subject to a same distribution in the past,present and the future,i.e.the series should be consistent.However,this as-sumption is usually not valid because of the impacts of climate change and human activities on hydro-meteorological conditions.Therefore,it is necessary to develop new approaches for frequency estimation of non-stationary time series of extreme values.The advances in the subject are presented in this study.The recent development has focused largely on the following two aspects,i.e.,the approach based on the backward restore or forward restore,and that directly based on the non-stationary hydro-meteorological extremes.Finally,perspectives on the hydrological frequency analysis of non-stationary time series are prospected.