I have been asked to do a weighted suitability analysis using all vector based polygon feature layers (8 different layers). Each layer has been ranked from 1 to 8 in order of importance.

My problem is that I need to convert each layer into rasters in order to do a weighted overlay, but many of the layers only contain boundary polygons that don't contain specific values in which to create a raster with the ability to be analyzed in a weighted overlay.

I get once I have rasters, I can reclassify them to match to do the overlay, but what happens when one (or two or three) of the those layers contain only polygons that represent boundaries and not a range in values (ie. one layer containing 5 equally important zip-code boundaries, one containing 3 equally important zone features, one containing 44 equally important census tracts, etc.)?

It was also requested that I make it look cool like a heat map.

Can this be done with a weighted Kerenel Density (then how do you weight kernel density...create a "population" field with ranked values 1 thru 8?)?

I am using ArcGIS Desktop 10.5 Advanced with most all extensions.

closed as too broad by PolyGeo Jul 7 at 1:17

Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer. Avoid asking multiple distinct questions at once. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • 1
    I would split this into two questions. In the second question you can outline the details of the "heat map" effect you have been asked to create. If you could expand on what your data looks like - are you using points / polygons, are you looking for concentrations of values or the outpu of the weighted overlay to look like a heat map. – Keagan Allan Jul 2 at 0:01

This is how I would approach the problem:

If you are set on using a raster method you will need to ensure that the spatial extent of all datasets are the same. You will need to create a "base" polygon, this polygon will contain a value for what you would like to represent a "no data" value, possibly a 0 or a 1, depending on what the values in your other polygons look like.

For each of the input layers (this would be the 8 layers you want to convert to raster for the weighted overlay), you need to run an "Erase" between the input layer and the base layer. This erases the common area that lies underneath the input layer or removes areas which contain information according to the input layer.

You will be left with a base layer full of "holes", you can then run a merge between this layer and the input layer you used to do the erase. This will give you a final polygon layer containing the values from the input layers and a 0 value for areas where that input layer does not have any data. Create a new field in this layer and copy the data values across. You can use this field to generate the raster.

Repeat this process 8 times. The result will be 8 rasters containing data from each of the input layers and "0" values where data does not exist. These can then be used in the weighted overlay function.

  • Interesting idea, and I will use this method and let you know how it goes. I think the idea of standardizing the spatial extents is great, something I hadn't even considered. The method you describe is for binary data (either yes=1 or no=0). What happens when 2 of the 8 layers are not binary, but rather contains data for value ranges which I would like to incorporate. Let's say my ranges for those two layers include 5 classes each...can I reclassify the binary data to match 5 value classes? I appreciate your continued guidance. – Justin Cahoon Jul 2 at 14:53
  • I think I may have muddied the water with the 0 and 1 statement. I meant for the 0 and 1 to represent the no data in the base layer. The range of values in the overlay polygon can be whatever you liked. I mentioned a 0 or a 1 because I wasn’t sure how you were planning on doing the overlay. Where a 0 could through an exception. Any value can go in the polygons. As long as your raster format is compatible. – Keagan Allan Jul 2 at 20:54

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