EOF output normalisation
In an Empirical Orthogobal Function (EOF) decomposition the
normalisation of the EOF maps and PC time series in arbitarry.
Consider the definition
E(x,t) = ∑i=1N
One can multiply the EOF Ei(x) by an
arbitrary factor A as long as you divide the time series
pi(t) by the same factor A.
You can choose between four normalisation schemes.
- Spatial maximum one. This implies that all EOF plots will run
from -1 to +1, which makes them easy to plot. The PC time series
will have the units of the original field and smaller amplitudes for
higher values of EOF number i.
- Spatial variance one. resembles the previous option, but makes
the EOFs and PCs more comparable if there are few isolated areas
with large amplitude in the EOF.
- Time series maximum one. Not very useful, included for
Time series variance one. This option implies that the EOF maps
have the units of teh original field. They will have smaller
amplitude for larger EOF number i.