Confusion about the prediction of β parameter in universal kriging

[Bamboobill]

I have a question about how to obtain a map of regression coefficients that varies with space when I do Multiple Linear Regression in universal kriging. For example, a linear regression of Y = k1X1+k2X2. Could I predict a spatial pattern of parameter k1 in gstat package?

I noticed that Liang et al (2016) used gstat package to estimate a global map of effects of species diversity on productivity, which means that the regression parameter (e.g., k1 and k2) varies with spatial location. However, I do not know how to implement this， though I have tried many times. Is there any function to get regression coefficients that varies with space in gstat package? Many thanks!

Reference: Jingjing Liang. et al., 2016, Science 'Positive biodiversity-productivity relationship predominant in global forests'

[edzer]

Yes, but you'll have to fake it by providing a newdata object that has X1 values equal to 1 and X2 values equal to 0.