How to estimate the period associated with the 100-year wave? 1

[buoy]

I know how to estimate the 100-year wave height from a time series of wave measurements, but not how to estimate the wave period associated with the 100-year height. Let's say based on a 20-year time series of sea surface elevation measurements, the 100-year height is estimated to be Hs=15m, but meanwhile the highest observed wave in this period is Hs=13m. How is the period associated with Hs=15m estimated?

Presently I am calculating it as follows: (a) For 0.5m-bins of wave height extending from Hs>2m to Hs=13m, I calculate the most probable wave period for that bin (in other words, at Hs=2.5m it's Ta=9 seconds, at Hs=3m it's 10 seconds, etc). (b) I fit a line to those most probable periods. (c) I extend the line to Hs=15m.

Is there a better or more standard way to estimate the period associated with a 100-year Hs?

 

[buoy]

I found the article 'Joint modelling of wave spectral parameters for extreme sea states' by P. Jonathan, J. Flynn, and K. Ewans, which discusses how to estimate extreme Hs and associated Tp. My method above should not be used. I will be trying to implement this using the tools in the WAFO toolbox. If you know of this being done somewhere already please let me know, since extreme value analysis is new to me.