[R] interpolation of climate data

[Mintewab Bezabih]

Dear R-users,  
 
I am working on interpolating the station level temperature data to farm level data. I have z vector consisting  of station level temperature observations and my x and y are latitude and longitude corresponding to a farm. My understanding is I can use raster combined with tps. While I am clear with the tps bit, I am not sure how I can construct the raster with teh data I have. Here is the reproducable example I made

Many thanks in advance
regards, 
Mintewab 

 

 library(fields)
x <-1:20
y<- runif(20)
z<- c(11, 15, 17, 2, 18, 6, 7, NA, 12, 10,21, 25, 27, 12, 28, 16,
17, NA, 12, 10)
 
mydataset<-data.frame(z, y, z)
mydataset[complete.cases(mydataset),]
tpsfit <- Tps(cbind(x, y), z, scale.type="unscaled")
 
library(raster) 
r <- raster(system.file("external/test.grd", package="raster"))
p <- raster(r)
p <- interpolate(p, tpsfit)
p <- mask(p, r)
plot(p)
se <- interpolate(p, tpsfit, fun=predict.se)
se <- mask(se, r)
plot(se)