ESDU 88037:2012

ESDU 88037 provides two examples of the use of the methods of
ESDU 87034 for the analysis of wind speed maxima: using data for a
single dominant storm mechanism and using data where there are two
dominant mechanisms. In the first example, the quantity of data to
be considered is reduced by taking only those wind speeds above a
certain threshold, and the procedure for calculating confidence
limits on the variation of estimated extreme wind speed with mean
recurrence interval is illustrated. For the other example, a Monte
Carlo method is used to determine confidence limits on the estimate
of extreme wind speed by generating a sequence of simulated annual
extreme storms having the required characteristics. The errors in
estimated extreme wind speed if the wind speed maxima from the two
mechanisms are not analysed separately are also shown. A third
example is included, showing the inverse use of the method of ESDU
88038 by which the parameters of the Fisher-Tippett Type 1 extreme
value distribution are deduced from the Weibull parameters of the
parent cumulative distribution (in this case a two-year continuous
wind speed record). This approach is particularly useful where
complex topographic effects on wind speed at a site cannot be
determined so that extreme wind speed cannot be predicted using a
design wind speed map.