STC = SPATIOTEMPCORRRF(F) computes the spatio-temporal correlation
function of the scalar fields F, defined as
C(X,T) = < w(x,y,t) w(x+X,y,t+T) > / < w(x,y,t)^2 >,
with <..> the average over x, y, and t. The normalization is such that
C(0,0)=1. This is useful to determine the convection velocity of a
pattern (e.g. a wave), which is given by the slope X/T maximizing the
correlation C.
The spatial correlation is computed along the X direction only (rotate
your field if you want a correlation along Y). Since C(-X,-T) = C(X,T)
(for an homogeneous and stationnary process), only two out of four
quadrants are computed: (X,T) = (>0, >0) and (<0, >0).
The output structure STC contains:
STC.cor Matrix containing the correlation C(X,T)
STC.X Vector (of size 1+round(Nx/2)) containing the separations
STC.T Vector (of size 1+round(Nt/2)) containing the time lags
where Nt = LENGTH(F) and (Nx, Ny) is the size of the fields.
STC = STCORRF(...,'full') computes the correlation for all possible X
and T (remark: large X and T are noisy because of poor average).
STC = STCORRF(...,'verbose') displays the work in progress.
Example
stc = stcorrf(h,'verb');
contourf(stc.X, stc.T/200, stc.cor',16);
xlabel('X (mm)'); ylabel('T (s)');
See Also
corrf, tempcorrf, tempspecf, specf, tempfilterf.
Published output in the Help browser
showdemo spatiotempcorrf