Font Size: a A A

Research On Hydrological Frequency Of Rivers In Arid Areas Based On Normal Transformation

Posted on:2022-05-07Degree:MasterType:Thesis
Country:ChinaCandidate:G T CaiFull Text:PDF
GTID:2480306548988369Subject:Hydraulic engineering
Abstract/Summary:PDF Full Text Request
The calculation of hydrological frequency is one of the most important research contents in the field of hydrology.It mainly explores the stochastic process of hydrological variables in the basin,analyzes the hidden statistical characteristics,and calculates the hydrological extreme events with the specified recurrence period based on the statistical results.Provide scientific basis for planning,design,operation and management of water conservancy projects.With the development of hydrology,new hydrological frequency calculation methods have gradually emerged.In order to enrich the calculation theory in time,the new method needs to be researched and summarized.Therefore,this paper uses the non-parametric normal transformation method to carry out the hydrological frequency study on the Manas River(hereinafter referred to as the Mahe),and compares the traditional method(P-? distribution)to study the pros and cons and adaptability of the hydrological model.Take Mahe Kenswat Reservoir as an example to carry out the application research of normal transformation.The results show that the normal transformation method used in this paper has good adaptability in the Mahe in the arid area of Xinjiang,and can be used in the planning,design and verification of water conservancy projects.The research conclusions are as follows:(1)The inconsistency test was carried out on the annual runoff sequence and flood sequence of the Mahe,the test results show that the annual runoff sequence mutated in 1996,and the flood sequence mutated in 1993.The main mutation forms of the two hydrological variable sequences are jump mutations.(2)Both normal transformation methods can significantly reduce the skewness of the Mahe sample sequence,and the Johnson transformation is always better than the Box-Cox transformation in improving the skewness of the samples.(3)The hydrological frequency of Mahe River was studied based on the normal transformation method and P-? distribution,The results show that the Box-Cox transform has more advantages in the high flows instability of the hydrological series because of its single parameter structure characteristics,and the Johnson transform has better fitting degree in the whole part of the hydrological series because of its multi-parameter structure characteristics.Moreover,the better the normality of the original skewed hydrological series after normal transformation,the higher the accuracy of hydrological frequency calculation.The Johnson transform is inferior to the Box-Cox transform when the normality of the transformed sequence is similar.(4)The 1999 flood in the Mahe was selected as a typical flood process,the flood process was designed based on the normal transformation method,and it was used as the reservoir flood for flood regulation calculation.Finally,the flood control risk rates of the reservoir under the two normal transformations are respectively 0.018% and 0.017%.That is,the reservoir is in a safe state.The results show that normal transformation can be used to design the flood and check the water level of engineering characteristic,and the accuracy of hydrological frequency calculation will directly affect the design flood process line and flood control calculation of reservoir.Therefore,it is necessary to improve the calculation accuracy of hydrological frequency by comprehensive use of various calculation methods to ensure the safety of engineering.The research results of this paper can enrich the theoretical system of hydrological frequency calculation in Xinjiang arid basin,and provide technical support for other basins to study the model.
Keywords/Search Tags:Hydrological frequency calculation, Normal transformation, Manas River, Non-parametric method, Flood control risk analysis
PDF Full Text Request
Related items