Abstract
The big data era in hydrology is on its way and will be of great significance to design flood prediction. How to discover and extract the most useful information from the abundant data with high correlation becomes an important problem. This paper aims to do a trial on data selection based on peak over threshold (POT) method. Attempts were made to evaluate the impacts of generalized Pareto (GP) distribution for design flood prediction. The daily discharge series of Danjiangkou Reservoir from 1954 to 2014 was taken as case study. The suitable threshold was determined as 1352.7m3/s by applying an automatic threshold technology. And a POT series with 369 independent flood peaks was obtained consequently, which was 6 times more than annual maximum (AM) series. The reasonable results indicated that GP model was a good fitness for POT series. The performances of the POT/GP and AM/P-IH models were compared on the basis of the T-year event estimates. It was concluded the design flood estimates of POT/GP model was more reliable than that of AM/P-HI model during the rational utilization period of water resources and hydropower works. The experience knowledge derived from small samples can be extended to big data by application of POT method.
Sign-in/Register to access full text options 
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 callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
