One more Tree Data figure. I had been messing around with the tree data clustering NTA neighborhoods by tree species and count profiles. But we can also cluster tree species by similarity of neighborhood profiles—that is, by trying to put the species most likely to be found together in neighborhoods close together in clusters. Here’s what that looks like, as a circular dendrogram. It is of course pleasing to make a dendrogram that clusters things that have real leaves and branches. The figure is also available as a PDF file.

New York City's street tree species clustered by similarity of neighborhood profiles

New York City’s street tree species clustered by similarity of neighborhood profiles

Update: While I’m at it, here’s the other side of the matrix, with the neighborhoods clustered by tree profile similarity.

New York City's NTA neighborhoods clustered by similarity of tree-species/prevalence profiles

New York City’s NTA neighborhoods clustered by similarity of tree-species/prevalence profiles

The figure is also available as a PDF file.