This is indeed an interesting topic. Robert's method may work if you average
the months over serveral years, e.g. an average january over 5-10 years may
not be influenced too much by seasonality...
But if you find a method which works better i think some would be interested
if you post your results
> From what I understood, most landscape classification methods available
> from GRASS (through R) do not account for the temporaI aspect of
> spectrum. So for example, if you have a raster where each band
> represents NDVI values at a given date, from March to September for
> example, the clas
On 06/13/2011 03:37 PM, Christopher Fusting wrote:
> Second, have a look at the CRAN 'spacetime' package for R. Although this
> will not solve your query, it may offer a different perspective.
Looking at packages is always good advice. There was also some discussion at
https://stat.ethz.ch/pip
Hey,
From what I understood, most landscape classification methods available
from GRASS (through R) do not account for the temporaI aspect of
spectrum. So for example, if you have a raster where each band
represents NDVI values at a given date, from March to September for
example, the classif