Title Comparison of Extreme Wind Pressure Estimation Methods for Structures Using Long-Term Measurement Data: Observation vs. Translation Approaches
Authors 김용철(Kim, Yong-Chul) ; 천동진(Cheon, Dong-Jin)
DOI https://doi.org/10.5659/JAIK.2026.42.8.347
Page pp.347-355
ISSN 2733-6247
Keywords Extreme Wind Pressure; Non-Gaussian Process; Observed-Peak Method; Translation Method
Abstract Accurate estimation of extreme wind pressures is essential for the safe and economical design of building cladding systems, particularly in regions exhibiting strong non-Gaussian characteristics. This study comparatively evaluates four representative methods?Gumbel, XIMIS(eXtended Improved Method of Independent Storm), HPM(Hermite Polynomial Method), and JTM(Johnson Transformation Model)? using long-duration wind pressure data corresponding to approximately 400 hours at full scale. The extreme values estimated by each method are assessed in terms of accuracy and uncertainty under identical conditions. The results show that all methods provide comparable accuracy in terms of mean squared error (MSE), while significant differences are observed in uncertainty. The XIMIS method exhibits the smallest uncertainty range, followed by HPM, indicating more stable performance. In contrast, the Gumbel method shows increasing uncertainty with larger extreme values, whereas the JTM method presents the largest variability overall. These findings suggest that XIMIS and HPM provide more reliable performance in practical applications.