Map of colour-coded grid cells according to a classification scheme

581
3
01-29-2021 09:02 AM
Labels (1)
ConnieHarpur
New Contributor

Hi all, 

I am trying to create a map consisting of a gridded area, whereby each grid cell is colour coded according to it's classification (1-4). The classification system is based on:1. The dominant feature class present, and 2. Average slope angle. There are 4 possible classes. 

I have a raster of slope gradient and a number of different features mapped out as shape files. 

What would be the easiest way to do this? The study area is large (>250 km^2) and I'd like the grid to have a resolution of 1 cell= 250m^2... I'm therefore reluctant to add classes for each cell manually, and it would be great if I could automate this. 

I tried to work out how to do this using Zonal Statistics (which I think is the way forwards?), but couldn't quite figure it out... 

I'd be really grateful for any help! 

Thanks,

Connie 

0 Kudos
3 Replies
DavidPike
MVP Frequent Contributor

For a start:

Create a fishnet - Create Fishnet (Data Management)—ArcGIS Pro | Documentation

Run Zonal statistics 'AVERAGE' against the fishnet and the slope raster.  Each cell in the fishnet is a polygon feature, against which will need a unique ID, maybe you can use ObjectID, not sure - not hard to create one either way.

Tabulate Intersection (Analysis)—ArcGIS Pro | Documentation

Should get you dominant features, if area is dominance.  

 

There's going to be a bit of messing about with the tabulate intersection table, but it shouldnt be too tricky.

Unsure how you want to bring it all together to classify though.

 

ConnieHarpur
New Contributor

Thanks David!

That's really helpful. I had got as far as making a fishnet but stalled at that point. 

Maybe easiest to run the two methods you suggested, rasterise the outputs and use zonal stats or map algebra to somehow automate the final classification.  

0 Kudos
DavidPike
MVP Frequent Contributor

Yes I think so, makes more sense for checking your outputs too.

0 Kudos