Uncertainty assessment of non-stationary hydrological frequency analysis based on the reconstruction of extreme hydrological series

Abstract: In this paper, a simple statistical model is introduced to reconstruct the non-stationary extreme hydrological series (EHS) to make the reconstructed EHS meet the requirement of stationarity, and further ensure the conventional frequency analysis methods is valid. This method is based on the assumption that the variation EHS owns an expected vibration center (mean), which is a linear combination of the two mean values of the two subsample series obtained through separating the original EHS by an optimal splitting point. Besides, considering the impact of sample length on the accuracy of design value estimation, the Bootstrap is applied to analyze the impact of sampling uncertainty on the design value. A large number of bootstrap samples are constructed from the reconstructed EHS, the corresponding design value or quantile is estimated for each bootstrap sample, so that the sampling distribution of design value is constructed, furthermore, the uncertainty of quantile estimation can be quantified. Keywords: Hydrological frequency analysis, non-stationary, reconstructing extreme hydrological series, uncertainty assessment, Bootstrap.