How can I render a three-compositional raster map using a ternary colormap?

Say there are three raster maps, representing percentages of sand, silt and clay respectively. The three values for each cell are hence summed up to 1. I want to use one raster map to characterize the relative contribution of the three materials. I found an R package named 'tricolore' for this purpose, but it is for a vector map, like this:

but my compositional data are stored in raster format.

Answer 1

[Alan]

Hi Jian,

If you have sand, silt, and clay data for each point, you could try out the geoscatter function: https://in.mathworks.com/help/matlab/ref/geoscatter.html#d124e556156

You could scale the material content to range between 0 and 1 and provide them as R, G, and B values (since you have exactly 3 quantities). The following syntax can be used:

geoscatter(lat, lon, A, C, "filled")

Where,

1. lat: Array containing latitude values.
2. lon: Array containing longitude values.
3. A: Array containing size of markers (They can be increased or reduced based on the density of the data you have)
4. C: Nx3 array in which each row represents the RGB values of each point, where N is the number of points in the raster data.
5. “filled”: Creates filled markers.

There are functions specifically meant to plot raster maps, but they might only support a single dimensional colormap (i.e., one cannot supply 3 channels). But if you have a unified quantity like the sum of all 3 quantities like you have mentioned you could use the functions mentioned in the “Raster Data” section of this documentation page: https://in.mathworks.com/help/map/vector-and-raster-map-display.html