Hi Luis,
To calculate the 'average magnitude of changes in precipitation' is straightforward - I calculated the delta (change) between average precip by 2070-2100 w.r.t 1961-90 at each grid box, but then before spatially averaging I calculated the root-square of the change at each grid box. This makes the 'magnitude if change' positive at every grid box so that when spatially averaged, the regions with negative changes don't cancel out the positive regions.
This is described in more detail in McSweeney et al 2012 - i'll try to upload the paper (the file is too big - will try to compress and add later).
Hope this helps,
Carol
|