Abstract
Due to climate change and human activities, the statistical characteristics of annual runoff series of many rivers around the world exhibit complex nonstationary changes, which seriously impact the frequency analysis of annual runoff and are thus becoming a hotspot of research. A variety of nonstationary frequency analysis methods has been proposed by many scholars, but their reliability and accuracy in practical application are still controversial. The recently proposed Mechanism-based Reconstruction (Me-RS) method is a method to deal with nonstationary changes in hydrological series, which solves the frequency analysis problem of the nonstationary hydrological series by transforming nonstationary series into stationary Me-RS series. Based on the Me-RS method, a calculation method of design annual runoff under the nonstationary conditions is proposed in this paper and applied to the Jialu River Basin (JRB) in northern Shaanxi, China. From the aspects of rationality and uncertainty, the calculated design value of annual runoff is analyzed and evaluated. Then, compared with the design values calculated by traditional frequency analysis method regardless of whether the sample series is stationary, the correctness of the Me-RS theory and its application reliability is demonstrated. The results show that calculation of design annual runoff based on the Me-RS method is not only scientific in theory, but also the obtained design values are relatively consistent with the characteristics of the river basin, and the uncertainty is obviously smaller. Therefore, the Me-RS provides an effective tool for annual runoff frequency analysis under nonstationary conditions.
Highlights
River runoff, as the most important form and component of water resources, has changed significantly in a number of rivers worldwide due to the impact of climate change and human activities
Considering that a large number of check dams has been constructed in Jialu River Basin (JRB), and the construction has been saturated in recent years, the reservoir index (RI) calculated based on the control area and the storage capacity of the check dams should be basically maintained at the level of 2010, so the RI data in 2010, as shown in Figure 5, were taken as the value at the design stage
It is not consistent with the statistical principle to analyze the nonstationary hydrological series by using the traditional method based on stationary samples, so the obtained design values cannot be guaranteed to conform to the reality
Summary
As the most important form and component of water resources, has changed significantly in a number of rivers worldwide due to the impact of climate change and human activities. The statistical characteristics of annual runoff series exhibit complex, nonstationary changes This change poses a serious threat to regional water resources security [1,2,3,4] and leads to the inability to analyze, predict, and manage water resources effectively, which is because the analysis method of the traditional design annual runoff based on the stationary assumption is no longer applicable. The design quantile obtained for a given reliability over the design lifetime varies with the choice of initial time and the curve type used for fitting the relationship between the statistical parameters and the covariates This means that the reliability of the future design values depends heavily on the time-varying characteristics of statistical parameters; the uncertainty about the prediction of statistical parameters is greatly increased due to the lack of ergodicity of the time series. We took the nonstationary annual runoff series in the Jialu River Basin (JRB) in northern Shaanxi as an example and analyzed the uncertainty of the deduced design annual runoff by Bootstrap method to verify the practicability and reliability of the nonstationary frequency analysis method proposed in this paper
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
