Geoscience Reference
In-Depth Information
We compute the wavelet coei cients using
cwt
.
coefs = cwt(series3L,scales,mother);
We convert the scales
scales
to pseudo-frequencies using the mother wavelet
mother
and the sampling period
dt
.
f = scal2frq(scales,mother,dt);
We use a i lled contour plot to portray the power spectrum (Fig. 5.17 a).
contour(t,f,abs(coefs),...
'LineStyle','none',...
'LineColor',[0 0 0],...
'Fill','on')
xlabel('Time')
ylabel('Frequency')
title('Wavelet Power Spectrum')
set(gcf,'Colormap',jet)
set(gca,'YLim',[0 0.04],...
'XGrid','On',...
'YGrid','On')
h e graph shows horizontal clusters of peaks at around 0.01 and 0.025 kyr
-1
,
corresponding to 100 and 40 kyr cycles. h e 40 kyr cycle (a frequency of
0.025 kyr
-1
) only appears at ca. 450 kyrs before present. Using
cwtft
instead
of
cwt
again creates a much smoother result (Fig. 5.17 b).
sc.s0 = s0;
sc.ds = ds;
sc.nb = nb;
sig = struct('val',series3L,...
'period',dt,...
'wavelet',mother,...
'scales',sc);
cwtstruct = cwtft(sig);
scales = cwtstruct.scales
f = 1./(4*pi*cwtstruct.scales/(w0+sqrt(2+w0^2)));
contour(t,f,abs(cwtstruct.cfs),...
'LineStyle','none',...
'LineColor',[0 0 0],...
'Fill','on')
xlabel('Time')
ylabel('Frequency')
title('Wavelet Power Spectrum Using FFT Algorithm')
set(gcf,'Colormap',jet)
set(gca,'YLim',[0 0.04],...
'XGrid','On',...
'YGrid','On')
Compared to the windowed power spectrum method, the wavelet power
