How do I ignore/delete points in the ocean when extracting values from a map using raster::extract?

Question

I have a list of coordinates from recorded sightings for a plant species.

I am extracting mean annual temperature from a map at each coordinate (using raster::extract).

But I now realise that a few coordinates are in the ocean due to some sampling error (e.g. they are just off the coast), causing the extracted values from maps to be inaccurate, because the map only gives information about terrestrial sites.

Is there a quick way of removing ocean coordinates form the list of coordinates, or ignoring ocean coordinates when extracting temperature values from the map? (I have to do this for many species and environmental variables)

Thank you

Answer 1

You can use na.omit or similar after you extract the values and detect the NAs

Example data:

library(raster)
set.seed(1)
r <- raster(ncol=36, nrow=18, vals=1:(18*36))
r[sample(ncell(r), 300)] = NA
s <- stack(r, r)
names(s) <- c('temperature', 'rain')
xy <- cbind(lon=-50, lat=seq(-80, 80, by=20))
sp <- data.frame(sp=rep(c('A', 'B', 'C'), 3), xy)

Use extract

e <- extract(s, xy)
x <- data.frame(sp, e)
x
#  sp lon lat temperature rain
#1  A -50 -80          NA   NA
#2  B -50 -60          NA   NA
#3  C -50 -40         482  482
#4  A -50 -20          NA   NA
#5  B -50   0         338  338
#6  C -50  20         266  266
#7  A -50  40          NA   NA
#8  B -50  60         122  122
#9  C -50  80          NA   NA

y <- na.omit(x)
y
#  sp lon lat temperature rain
#3  C -50 -40         482  482
#5  B -50   0         338  338
#6  C -50  20         266  266
#8  B -50  60         122  122