This is what we are aiming for. It is fairly complex so it would best to create a new project to keep it separate from the opportunity mapping project. We can save the output from the fragmentation model to use later in the opportunity mapping.

First we will identify woodland edges. Create a new project and add these components to the model canvas.

Connect them up as shown, save and run the model. The Numeric constant needs to be set to 25 because the cell dimensions at z20 are just over 24m and we want to capture patches of woodland two cells wide and edge patches a single cell wide.

Rename the components to indicate the purpose of each. This part of the model finds woodland edges by identifying areas of woodland adjacent to non-woodland.

Next we need to identify interior (or non-edge) woodland areas and generate a distance map. Add these additional components.

And connect them up as shown.

You will need to zoom out a bit to add the final two connections from “Woodland” and the “Numeric constant.”

Rename the “Set difference” component to “Interior woodland.” Save and run the model to preview the results so far.

Add an “Intersection” component and connect it to “Woodland” and the output from the “Greater” on the end of the “Interior woodland” row.

Rename the final “Intersection” component to “Isolated woodland”, save and run the model. The preview displays all woodland areas which are adjacent to non-woodland and non-adjacent to interior woodland.

Finally we need to create new “Continuous woodland” class. Add a “Set difference” component.

Rename it “Continuous woodland” and connect the input node A to “Woodland” as shown.

Drag the “Continuous woodland” component to the other end of the model near to “Isolated woodland.”

And connect “Isolated woodland” to input B.

Add a “Boolean to categorical dataset” converter and connect the outputs from “Isolated woodland” and “Continuous woodland.” Save and run the model.

Finally add the “Map layer” component to export the output to the Map view.

You can save the model and then toggle Map view and run the model from Map view.

When the model has finished running the new layer will be displayed showing the two new categories of woodland.

You might want to change the colours for each category to make them more visible.

If you zoom in on the map you will be able to see clearly the distribution of isolated and continuous woodland identified by the model.

The last thing to do before closing this project is to add the component to save the dataset. Name it something meaningful (e.g. wld = wealden district and z20 refers to the zoom level).