Hfrhyu

Working in R. I have rasterstacks of temperature / precipitation data for particular countries and agricultural seasons. For instance, a rasterstack of daily noon temperature for the growing season only, for 2011, covering Ethiopia only. That particular rasterstack would be 275 layers, since the growing season is 275 days.

I'd like to examine histograms of the cell-specific values across layers. That is, I want to see the rain / temp distribution that occurs within a season, over days, for any given location. I'm not wondering about a particular cell; I'd like to be able to quickly visualize these histograms for a bunch of different cells.

How would I do this?

Use the extract() function to find all values associated with a particular location, and visualize with hist(). Your raster's projection and extent will determine what points you want to use for PointOfInterest. A reproducible example:

library(raster)
r1 = raster(matrix(c(1:25),
 nrow=5,ncol=5))
r2 = raster(matrix(rep(5, 25),
 nrow=5,ncol=5))
r3 = raster(matrix(c(35:11),
 nrow=5,ncol=5))
r_stack = stack(r1,r2,r3)
extent(r_stack) # Bounded by (0,1) in both dimensions

PointOfInterest = matrix(data=c(0.1,0.1),
 ncol=2)
myValues = extract(x = r_stack,
 y = PointOfInterest)

hist(myValues)

If you want to do this with more than one point, you can add rows to the matrix with coordinates you're interested in, or use raster::sampleRandom() to randomly choose points. Then you just need to format the graphical parameters accordingly:

randomvals = sampleRandom(r_stack,
 size = 4)
randomvals # This did the extracting for us

# Set up paneling so you can look at a bunch of histograms at once
n.points = nrow(randomvals)
par(mfrow=c(ceiling(n.points/2),2))
# Create a histogram for each point
for(i in 1:n.points)
 hist(randomvals[i,])