How to determine the changes between two polygon shapefiles of similar data

[WhitneyNewcomb]

The city I work for gets its parcel data from the county once a month so we have several years of almost the same data. I want to be able to determine where changes occur between every new month. So what are the changes between September to October, October to November, etc. And by changes I mean any change in geometry (like parcel splits or merges) or changes in attributes (especially changes in ownership).

I've contacted ESRI about this and they weren't able to do a whole lot for me. 

Here's what I've tried so far:

I am able to find changes in geometry by using the Select by Location: Identical features and inverting the selection. 

This works fine but does not get me changes in attributes.

I would use the Feature Compare tool but this does not work as it requires the same number of rows in both tables - which I don't have. 

I also found an old arcgisscripting tool and converted it to arcpy and it worked with point datasets (but at a certain point in time the County changed how they did their parcel data and went to polygons only which the script does not work with).

So I'm at a loss for what to do now.

Does anyone have any advice for me?

Please let me know if there's any info I can give to help you help me.

-Whitney Newcomb

[MichaelVolz]

Does the data contain any date/time fields that you can use to differentiate the data?  Do you know if the county enforces attribute changes to be tracked by populating a date/time field (Editor Tracking would be useful in this scenario, but you would need to contact the county to see if they are implementing this procedure)?

[WhitneyNewcomb]

Unfortunately, no it doesn't. They don't even fill out their metadata. It's rather upsetting. Hah!

[DanPatterson_Retired]

for your last option... doesn't work with polygons....produce a centroid file of your polygons, then you can use that at least for the attributes

[WhitneyNewcomb]

Dan,

That actually might work for me to some extent, but I'm worried that I'll have a hard time converting it back to a polygon file.

The script I found grabs the different fields of the two parent files and dumps them into one new file so that for every point in it there is (usually) actuallly two points stacked on top of each other. One from the earlier month and one from the later month.

[DanPatterson_Retired]

you wouldn't need to convert the centroid file back to a polygon file... it would just be used to identify which record attributes have changed... Your polygon file would remain untouched 

[WhitneyNewcomb]

You're totally right.

Thank you Dan. You have been most helpful!

[NobbirAhmed]

Could you try first converting the polygons to polylines and then using the polylines to compare? If you want you can use Polygon to Line under Data Management Toolbox, in Features toolset.

http://support.esri.com/en/technical-article/000007707

[WhitneyNewcomb]

Hi Nobbir,

This doesn't work for because attributes do not come along with the polylines. I really need to see the differences in attributes. 

[DanLee]

Hi Whitney,

We are developing tools to help solve conflation problems. There is a GP tool Detect Feature Changes which currently only works on line features, such as roads, parcel lines, etc. This tool detects spatial (geometry) and attribute differences between two sets of line features.

We are investigating polygon use cases. What you described seems a good case for us. Would you be able to share some sample data, two or three monthly data of a small area? If so, you can attach your data in email to dlee@esri.com.

In the meantime, I think it's a good idea to first exclude the spatially identical ones, as you did. For the remaining parcels, perhaps try Spatial Join Oct vs. Nov, and then Spatial Join Nov. vs Oct. That could give you 1:n relationships both ways to catch what's been split or combined.

Thanks,

Dan (Esri Conflation Team)