The analysis of extreme values is important in many areas of application. One concern of the Structures Group at NIST is designing structures, especially tall ones, to withstand extreme wind events. These engineers are typically interested in estimating return values or the distribution of the maximum or minimum value from univariate and possibly multivariate time histories, collected outdoors and from wind tunnel tests. Novel statistical methods for these kinds of data could help in designing more cost-efficient structures that remain safe.

 

Reference

Pintar AL, Simiu E, Lombardo FT, Levitan M: "Maps of Non-tornadic Wind Speeds With Specified Mean Recurrence Intervals for the Contiguous United States Using a Two-dimensional Poisson Process Extreme Value Model and Local Regression". NIST Special Publication 500-301 http://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.500-301.pdf

 

key words

Extreme values; Peaks-over-threshold; Generalized extreme value distribution; Generalized pareto distribution; Poisson process;

Citizenship:  Open to U.S. citizens

Level:  Open to Postdoctoral applicants