Hello,
I'm using Kernel Interpolation with Barriers (for contamination in a bay area). I log-transformed my data to run the analysis. Finally, I exported the GA layers to rasters (prediction and standard error).
Now, I want back transform my data. Any recommendation on how to do this? Can I just transform my prediction surface, e.g., calculate exp10? I read that for kriging one should consider variance. What about the standard error surface?
Thank you for any pointers,
IP
