Dates: 04/03/09 (Part I) and  04/17/09 (Part II)

Speaker: W. Xiao

Title: Predicting Rogue Waves from Directional Spectra

Abstract:
Rogue waves are defined as waves whose wave height is multiple times  larger than the
significant wave height. Those waves are suspected to  be the reason for many platform
damage and ship loss. There are two  essential research topics in this area. The first
one is to understand  the generation mechanisms of these giants and the second is to
develop  a reliable statistical model to predict the rogue wave occurrence. In  this
talk, we describe an investigation of the occurrence, statistics,  and generation
mechanisms of rogue waves in the open sea using large- scale three-dimensional
phase-resolved wavefield simulations. A large  number of phase-resolved nonlinear
wave-fields, initially specified by  directional wave spectra, is obtained by applying
direct simulations.  The typical domain size of such numerical nonlinear wave-fields is 
O(10^3 km^2) over evolution time of O(hr). From the simulated  nonlinear wave-fields,
rogue wave events are identified and their  occurrence statistics are studied. It is
shown that the classic linear  and second-order narrow-banded theories significantly
underestimate  the rogue wave occurrence in wave-fields with moderately large wave 
steepness, relatively narrow spectra bandwidth and narrow directional  spreading.  The
influence of key wave spectrum parameters (such as  typical steepness, spectra bandwidth
and directional spreading) on the  rogue wave occurrence is analyzed. It shows, unlike
the case found in  long-crested wavefield, the occurrence of rogue wave does not only 
depend on the so-called Benjamin-Feir index but also the crest length  and group length
of the 3D wavefield. The first part of the talk also  serves as a short up-to-date review
on statistical models developed  for large waves.