[Sea-mat] IMF's and Hilbert spectrum

[Derek Goring]

Hi there

Has anyone out there tried using IMF's (intrinsic mode functions) 
and the Hilbert spectrum to analyse seiche data as per Huang et al 
in JPO, 30: 2001-2012?

IMF's seem to work very well once you figure out how to deal with 
end effects of the cubic spline (I simply use a best fit linear curve 
from the penultimate knot to the ends).  My problem is how to deal 
with negative frequencies arising from:

h=hilbert(imf);
phi=unwrap(angle(h));
w=diff(phi)/dt;

Has anyone figured this one out?

 
Derek Goring
Principal Scientist
Coastal and Freshwater Hydrodynamics
NIWA
P O Box 8602
Christchurch
NEW ZEALAND
Ph 64 3 343 7868
Fax 64 3 348 5548
http://www.niwa.cri.nz