2

I am using ArcGIS Desktop.

I have a point shapefile for 5 different plant species and rasters (elevation, slope, aspect, canopy height). I am trying to find clusters for each species in terms of raster values (elevation, slope, aspect, canopy height). I need to figure out where cluster are spatially and significantly. So far, I ran extract multi values to points, and I have raster values for each point in attribute table. Then, I got stuck!

Do you have any ideas what to do next to have a visual layer ( raster or vector) where it shows the significant clusters?

enter image description here

enter image description here

  • 1
    Have you looked into Ripley's K? – Aaron Nov 19 '17 at 5:49
  • Thank you @Aaron . Yes I did. It produced a table if there is any significant clusters. I am looking for something more visual for a power presentation along with statistical values. – Amadeus Nov 19 '17 at 5:56
  • 1
    It looks like the data was colected in transects, as such point pattern and density analysis are not valid here. All I can think of is an actual cluster analysis such as k-means but, it is not avalible in arcgis. – Jeffrey Evans Nov 19 '17 at 22:10
  • @JeffreyEvans Yes I thought the same. Now I am making Similarity Search rather than spatial analysis. Thanks for your comment! – Amadeus Nov 20 '17 at 0:59
0

One possible solution to your problem could be the following:

1) Go to Layer > Add Layer > Add/Edit Virtual Layer...

enter image description here

2) In the Create a virtual Layer window you could rename the default layer name, then press the Import button. After that, you will select the layer to embed and then you'll press ok.

enter image description here

3) Now, you have a virtual layer which will serve for our analysis. Open the virtual layer Properties, go to General tab and press the Query Builder button.

enter image description here

In the Provider specific filter expression please enter the following statement: "PlantType" = 'Persimmon'

enter image description here

5) If you wish, you could specify a label for the virtual layer features...

enter image description here

and finally you set a Heatmap style.

enter image description here

6) Now, you can easily identify where the Persimmon cluster is spatially significantly. Of course, you can set more complex queries, as desired.

enter image description here

The entire project can be downloaded from here.

  • Thanks for your thorough explanation and answer but I am working with Arcgis Desktop version. I need to start learning qgis soon. – Amadeus Nov 19 '17 at 17:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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