Closure to "Wind Speeds in ASCE 7 Standard Peak-Gust Map: Assessment" By Emil Simiu, Roseanne Wilcox, Fahim Sadek, and James J. Filliben
暂无分享,去创建一个
The ASCE 7 peak-gust map divides the U.S. into 2 main adjacent wind speed zones that do not reflect correctly the country's differentiated extreme wind climate. This closure article is written by the authors of a paper (Simiu et al, April 2003) that showed that the methodology used in the map's development averages out real climatological differences and causes severe bias errors for the following reasons: 1) estimation of the speeds was based on superstations, of which 80% included stations that were also contained in 1 or more other superstations; 2) stations with significantly different physical geography and meteorology were in many cases included in the same superstation; 3) legitimate wind speed data was omitted from data records in cases in which analyses resulted in speeds different from those postulated in the map; and 4) off-the-shelf smoothing software was used that does not account for physical geography and meteorological differences. In the commentary article, Peterka and Esterday (June 2005) contended that Simiu et al failed to demonstrate an improved analysis methodology and did not understand the improvement of the gust map in reducing sampling error over the earlier fastest-mile map. The commentary authors conclude that while the current ASCE 7 wind map is based on reasonable analysis, additional analysis incorporating recent wind data would be beneficial. Simiu et al respond in this closure article by emphasizing that reduction of sampling errors is only part of a larger story. The authors conclude by reiterating their position that the current ASCE 7 map does not reflect correctly the country's differentiated extreme wind climate. In addition, given the large role played by extreme wind speeds in the design of many types of engineering structures, unnecessarily misestimating extreme wind speeds is a disservice to the structural engineering community and the public at large.
[1] C.M.L. Dorman,et al. Extreme values — an improved method of fitting , 1982 .