Select to view content in your preferred language

Buffer Density Analysis

5912
13
Jump to solution
08-08-2012 08:21 AM
MikeWissner
Emerging Contributor
Hello All,

I am running an analysis to prioritize locations that have the highest buffer densities, ie places where many buffers overlap.  Any idea how to determine this?  I don't think it's feasible to separate each buffer into its own layer.

Any help would be much appreciated,

thanks,
mike
0 Kudos
1 Solution

Accepted Solutions
by Anonymous User
Not applicable
One thing you can do (this is more ideal for points) depending on the scale of your buffers is to convert those to points then to rasters (you will have to play around with the cell size depending on your scale) and run focal statistics using the sum. 

Before converting the buffer points to raster, you need to add a field and give all the points a value of 1 and use this as the value field of the raster.  Once you convert this to a raster you can run focal statistics using the sum to show areas with high densities of buffers.  In order to get the effects of your buffers you will need to set the neighborhood to something that would be close to the size of your buffers to find out how many overlap.  I have never tried to do this with polygon features and you will probably have to play around with cell sizes  and neighborhoods to get good results.  Hopefully this will be a good place to start.

Caleb

View solution in original post

0 Kudos
13 Replies
by Anonymous User
Not applicable
One thing you can do (this is more ideal for points) depending on the scale of your buffers is to convert those to points then to rasters (you will have to play around with the cell size depending on your scale) and run focal statistics using the sum. 

Before converting the buffer points to raster, you need to add a field and give all the points a value of 1 and use this as the value field of the raster.  Once you convert this to a raster you can run focal statistics using the sum to show areas with high densities of buffers.  In order to get the effects of your buffers you will need to set the neighborhood to something that would be close to the size of your buffers to find out how many overlap.  I have never tried to do this with polygon features and you will probably have to play around with cell sizes  and neighborhoods to get good results.  Hopefully this will be a good place to start.

Caleb
0 Kudos
by Anonymous User
Not applicable
Here is a quick analysis I just did.  I have used focal statistics to identify the number of structures within 150 ft of each other.  This method should give you a pretty good idea of the spots that should be high priority. 

[ATTACH=CONFIG]16781[/ATTACH]

Caleb
0 Kudos
MikeWissner
Emerging Contributor
Thanks Caleb,

that makes perfect sense, but when I tried to convert the buffers to points, I got the following:

WARNING 000117: Warning Empty Output Generated

Should I not be using the Feature to Point tool?


-m
0 Kudos
by Anonymous User
Not applicable
Mike,

Hmm that is strange. I would have gone with the feature to point function as well, but I don't know why you would be getting that error message.  Were your buffers based off of a point feature class?  If so, I would just use the original point FC to convert the points to raster.  Also, what is the size of your buffers?

Caleb
0 Kudos
MikeWissner
Emerging Contributor
No, buffers were originally based on polygon features.

Buffers are 1/4 mile.
0 Kudos
by Anonymous User
Not applicable
Did you try using the original polygons as the input for the Feature to Point tool?  I don't understand why the buffers did not work for you though.

If that does not work either for some reason I have one other idea.  You could also try converting the original polygon features to raster if your polygons are fairly close in size.  You would want to set a cell size slightly larger than that of the largest polygon to ensure that you only get one cell per polygon.

If neither of the above works for you, would you mind attaching a small sample of the data?

Caleb
0 Kudos
MikeWissner
Emerging Contributor
I converted the original polygons to points and got a blank raster when I tried the feature to raster tool.

A sample of the data is attached (it's in xls format since the forum doesn't allow csv uploads).

Thanks again for your help with this.
0 Kudos
by Anonymous User
Not applicable
No problem.  When you converted the points to raster what cell size did you use?  The output raster had no cells at all?  Also, instead of using the feature to raster tool it may be better to use the point to raster tool.  One more thing, before converting the points to raster did you create the value field (all = 1) and use that as the value field for the raster?

And sorry I wasn't very clear about the data sample.  Could you attach a sample of the shapefile with just a few of the polygons? The .xls did not have any coordinate info. I just want to take a look at the data to see if I can figure out why these things are not working for you.  Thanks

Caleb
0 Kudos
MikeWissner
Emerging Contributor
I tried the default cell size of 200 and 60 and neither produced a raster with any cells at all.  I tried the same with the Point to Raster tool and got the same results.  Yes, I included a field called 'weight' with values of 1.

There is a sample of the polygons and their buffers attached.

thank you
0 Kudos