| Title |
A Study on the Difference of Rainfall Intensity According to the Omission of Short-Term (20, 30, 40, 50 Minutes) Rainfall Data in Inducing I-D-F Curves |
| Authors |
이희창(Lee, Hee Chang) ; 성기원(Seong, Kee Won) |
| DOI |
https://doi.org/10.12652/Ksce.2020.40.5.0465 |
| Keywords |
I-D-F곡선;Box-Cox 변환;빈도해석;상대오차 I-D-F curve;Box-Cox transformation;Frequency analysis;Relative error |
| Abstract |
I-D-F curves were induced by Box-Cox transformation using rainfall data from five major cities in Korea: Seoul, Busan, Daegu, Daejeon, and Gwangju, as well as from Sancheong (South Gyeongsang province) and Yeongcheon (North Gyeongsang province) stations. The practicality of the Box-Cox transformation is more scalable than the traditional method of frequency analysis in terms of applicability because it is available even if the analysis data are insufficient to perform general frequency analysis and do not produce an appropriate probability density function. For the case in which rainfall data for the entire period (10-1440 minutes) and short-term period (20, 30, 40, 50 minutes) at the foregoing 7 stations are omitted, there was a relative error of -23.0 % to 14.7 % at a duration of 10 to 60 minutes below the 100-year frequency. Accordingly, rainfall analysis requires inducing I-D-F curves, including for the short term (20, 30, 40, 50 minutes), and if rainfall data are omitted for the short term (20, 30, 40, 50 minutes), it is necessary to increase the existing margin rate depending on the point in order to ensure the safe design of small-scale hydraulic structures. |