Hello Babu,
WorldClim is already downscaled, and downscaling a downscaled
dataset is generally not recommended. If you have homogenized
weather station data, then it is much more suitable input for your
downscaling process than the 1-km-resolution WorldClim data. Or a
homogenized and gridded but not downscaled dataset, like E-OBS.
I suggest the following steps:
- calculate the mean and standard deviation of the raw monthy
climate variables (Pres01, Prec02, ..., Tmax01) of the
coarse-resolution grid (E-OBS, WorldClim, whatever)
- standardize the variables: (x-mean)/sd
- extract the altitude from a DEM for each point of the
coarse-resolution grid
- create the fine-resolution (e.g. 100×100 m) grid
- extract the altitude from a DEM for each point of the
fine-resolution grid
- using for(), lapply() or future_lapply(), iterate through the 48
monthly climate variables, and do regression kriging with altitude,
latitude and longitude as auxiliary variables, using package "gstat"
- destudentize the kriged, fine-resolution monthly variables using
the previously stored means and standard deviations: x*sd + mean
- finally, calculate the bioclimatic variables from the monthly
climate variables. (I do not suggest downscaling the bioclimatic
variables, instead, downscale the raw monthly variables)
Hope this helps,
Ákos
_____________
Ákos Bede-Fazekas
Centre for Ecological Research, Hungary