Creating a map to show the dominant size class

I am currently trying to create a map to show the dominant size class of diatoms in a region.

The data sets for these size classes are in a 36x36 masked array (with each grid containing the concentration of the diatom)

I am looking to make a new data frame which compares these three size classes and plots a set value depending on which is larger in each of the grid spaces.

for example

if 1umdiatoms > 10umdiatoms and 100umdiatoms then the plotted value for that grid space = 0

if 10umdiatoms > 1umdiatoms and 100umdiatoms then the plotted value = 1

if 100umdiatoms > 1umdiatoms and 10umdiatoms then the plotted value = 2

This way I can assign each value a colour and show which size class is dominant there (only needs to be basic)

I am sure there is a better way to show this but I am not well versed in python so any ideas would be greatly appreciated.

Cheers.