Spatial correlation depending on attributes

336
2
11-10-2021 10:42 AM
TobiasMachnitzke
New Contributor

Hello everybody,

 

I am currently writing my Master's Thesis in forest sciences and one topic is spatial correlation between small and large tree individuals. As I am completely new to working with ArcGIS i have absolutley no idea how to prove said correlation. I already looked through some papers and found people performing Ripley's K to do just what I want to do. But as far as I get it, Ripley's K searches for clustering over all individuals.

 

So my question is, is there any tool on ArcGIS, which allows me to analyse spatial correlation depending on attributes?

 

Regards

Tobias

0 Kudos
2 Replies
DanPatterson
MVP Esteemed Contributor
0 Kudos
LaurenGriffin
Esri Contributor

You have a couple choices depending on the question you want to answer. This is how I would approach it:

1) To answer the question do tall trees cluster together spatially, and do small trees cluster together spatially (or , alternatively, do tall trees repel small trees and vice verse)? Run the Spatial Autocorrelation tool using tree height as your analysis field. This document tells you how to interpret the results: https://pro.arcgis.com/en/pro-app/latest/tool-reference/spatial-statistics/h-how-spatial-autocorrela...

2) If I want to know WHERE tall trees cluster spatially and where small trees cluster spatially, I would run Hot Spot Analysis or Optimized Hot Spot Analysis... again, tree height is your analysis field. The hot spots will be locations where tall trees cluster spatially. The cold spots will be where the small trees cluster spatially.

2) If I want to understand if there are locations where a tall tree is surrounded by small trees or where a small tree is surrounded by tall trees, I would run Local Moran's I (Cluster Outlier Analysis), with tree height for the analysis field. This document describes how to interpret the results: https://pro.arcgis.com/en/pro-app/latest/tool-reference/spatial-statistics/h-how-cluster-and-outlier...

I hope this is helpful! 

Best wishes,

Lauren

0 Kudos