Data Science with R for the Life Sciences

Spatial Data Raster Math

Notes

Show > * Canopy Height Model picture

  • We’ve been working with the DTM data, which is the Digital Terrain Model, or the elevation of the ground
  • LIDAR also collects data on the highest point in each location
  • This is used to create a Digital Surface Model or DSM - the elevation of top physical point

Panel 1: Drawing of trees on undulating terrain. A green line along the top of the trees indicates the Digital Surface Model. Panel 2: Drawing of trees on undulating terrain. A brown line along the top of the terrain indicates the Digital Terrain Model. Panel 3: Drawing of trees on a flat terrain. A green line along the top of the trees indicates the Canopy Height Model. Panel 4: Equation: DSM (Digital Surface Model) - DTM (Digital Terrain Model) = CHM (Canopy Height Model).

  • In forested areas we can combine these to create a Canopy Height Model (CHM)
  • Do this by subtracting the DTM from the DSM
library(stars)

dtm_harv <- read_stars("data/HARV/HARV_dtmCrop.tif")
dsm_harv <- read_stars("data/HARV/HARV_dsmCrop.tif")
chm_harv <- dsm_harv - dtm_harv
  • Math happens on a cell by cell (elementwise) basis
  • Can then graph this new raster
ggplot() +
  geom_stars(data = chm_harv)
  • This lets us see where there are the tallest trees on the landscape and where there are none

Do Task 2 of Canopy Height from Space.