The hypothesis that better estimates of wind force extremes are obtained by analyzing time series of squares of wind speeds, rather than time series of wind speeds, has been considered in the engineering literature. We present results of Monte Carlo simulations, from which we conclude that the hypothesis does not appear to have merit. The results also suggest the possible use of comparisons between percentage point estimates based on wind speeds and on squares of wind speeds as a means for deciding on an appropriate probabilistic description of extreme wind speeds.
Proceedings Title: 13th ASCE Engineering Mechanics Division Conference
Conference Dates: June 13-16, 1999
Conference Title: ASCE Engineering Mechanics Conference
Pub Type: Conferences
building technology, extreme value statistics, structural engineering, wind engineering, wind forces