I am a rookie in ArcGIS and I am learning raster.

My problem is that I have a area of different kinds of landcover(actually it's rock type). Now the road is set to NoData, therefore the whole area is divided into several regions/blocks.

Is there a way for me to get know the area percentage of the most frequent rock type within each block?

This is a theoretical problem. My thought is to get the most frequent rock type, I need the help of attribute table-the one with the most count shall be the winner. For percentage I will just need to use the counts of cells of that rock type divided by the number of all the cells of that region.

Is there any way for me to get the total count of the block? There must be a better way of doing this using certain tool than stupidly count cells and do calculation.

the rock type map

1 Answer 1


There is a tool to complete your task without all the manual work you've described. It's called Zonal Statistics and is located in the Spatial Analyst Toolbox. In order for it to work, you need to prepare a feature class (or raster) that contains information about your zones (i.e. your road blocks).

You could:

  1. Create a new polygon feature class and draw the polygons of your road blocks (quick & dirty). (Depending on the number of blocks you have, this method might not be suitable).
  2. Create the polygons for your blocks by multiplying your raster with 0 (raster calculator tool from the spatial analyst toolbox), then running the int tool (spatial analyst) in order to create an integer raster and finally use the raster to polygon tool (conversion toolbox) in order to create polygons from your blocks. As your roads are no data values, the polygons will be created for each block.

Then use the feature class from 1) or 2) to run the zonal statistics tool. Use the command "MAJORITY" for the most frequently occurring rock type, or "VARIETY" for a summary for all your rock types. Good Luck.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.