Municipal GIS: ownership of data / details

[mapmeld]

Municipal GIS should include sections on ownership of data:

• Many cities do not have their own GIS data, getting it from the county or another regional/state authority (Macon being an example, oh man...)
• Some GIS datasets (water supply, for example) will raise public safety concerns.
• Regulations vary from state to state about whether a local government can charge for GIS data.
• If a local government can charge for data, they may share data for a project.
• If a local government can charge for data, they may see the value in having a data portal to streamline/automate answering data requests. I think Cook County, IL is an example but I am not 100% certain.

Technical implementation (not sure if this is too low-level)

• Many cities use ESRI ArcGIS software. This is a toolbar for the software to export open formats https://github.com/project-open-data/esri2open
• If data is public domain, you can share on OpenStreetMap. For data imports there is an imports e-mail list and wiki page http://wiki.openstreetmap.org/wiki/Import to guide you on how to avoid using data without permission or overwriting existing data.

Bonus points:

• Public safety GIS data (including crime) is often handled by a separate team and under a separate policy from how the rest of the city releases its GIS data.
• If a local government charges for data, they may give you data but ask you not to share it.

[daguar]

+1 to @mapmeld's comments, and add:

Assessment data is some of the highest-value GIS data out there from the standpoint of direct economic value creation. Unfortunately, for that very reason, lots of counties try to charge large fees for this info. Some kind of organized strategy to push on this front would be incredibly useful (e.g., mobilizing the actors who benefit, producing prepackaged talking points for local pushing, etc.)

[waldoj]

This is enormously helpful, @mapmeld and @daguar—thank you.

My hope is that some localities will be interested in sharing some or all of their data, because doing so will enable them to provide a modern, functional, free web-based GIS system on their own website, for the benefit of their citizens. But even localities that don't share assessment data are still sharing a great deal of useful data. Something is better than nothing. :) But, yes, @daguar, I'm not prepared to give up that easily.

[jpmckinney]

Not sure if it's relevant, but we've been successful in getting Canadian governments to send us electoral boundaries and getting permission to distribute those boundaries, which we’ve then been able to aggregate on GitHub and make available via API. That’s boundaries for 353 governments (though we only had to ask 90, because some are aggregators). By population represented, we’re 80-90% of the way to having all electoral boundaries at all levels of government.

[waldoj]

Wow, that's really great—y'all are doing great work! I'll have to read through that to learn more about it. Thanks for that!

[jpmckinney]

We haven't published much about our data collection process, but I'm happy to answer any questions here or by email.

[daguar]

I've been really intrigued by the OpenAddresses project as a potential model for something like "OpenParcels" — they store links to source data, and then (appear to) write one-off crosswalks for the source data fields to the standardized data schema.

For example, here's the source data record for Alameda County, CA:

{
    "coverage": {
        "country": "us",
        "state": "ca",
        "county": "Alameda"
    },
    "data": "https://data.acgov.org/api/geospatial/8e4s-7f4v?method=export&format=Original",
    "license": "http://www.acgov.org/acdata/terms.htm",
    "year": "",
    "type": "http",
    "compression": "zip",
    "conform": {
        "merge": [
            "feanme",
            "featyp"
        ],
        "lon": "x",
        "lat": "y",
        "number": "st_num",
        "street": "auto_street",
        "type": "shapefile"
    },
    "fingerprint": "177cd91707ab2c022304130849849255",
    "version": "20140330",
    "cache": "http://s3.amazonaws.com/openaddresses/us-ca-alameda.zip",
    "processed": "http://s3.amazonaws.com/openaddresses/us-ca-alameda.csv"
}

What's more, in general I think this is an interesting approach for doing mass open data collection in a structured way, and worth considering across other project domains.

[waldoj]

We're thinking along the same lines, @daguar. I filed my first pull request with OpenAddresses a few days ago, and while I think the address data is valuable, I know that's raw material for an open, parcel-based map of the entire country. I'm trying to figure out how we can support OpenAddresses, toward the goal that @mapmeld created with this issue.