[talk-au] Tracks flagged as missing from government data

Little Maps mapslittle at gmail.com
Mon Aug 23 22:59:45 UTC 2021

Hi again, I’ve dominated this discussion to a painful extent so this will be my last message on the topic unless explicit questions are raised. (Thank god, I hear everyone exclaim!). We’ve moved away from the original topic of this thread so i suggest we start a new thread if there’s interest in discussing track access imports more broadly.

AH: I guess it comes down to the data, if it's reliable enough. Maybe I could try to compared existing tracks in Victoria which have an access value with Vicmap to see if it's usually in agreement or not. How do you rate the accuracy of the Vicmap data when it comes to track access?

Vic Gov agencies advertise track closures (permanent, seasonal and short term; we don’t care about the latter) on web and in apps, and these closures are legally enforced, so I guess the question is: how close is the Vic Transport dataset to what is advertised on those web pages?

My *guess* is they’re extremely close and if anything the VicTransport dataset might just be a bit behind the times, but I emphasise that’s a guess. We could easily test that though (where ‘easily’ means conceptually but someone still has to do it).

These types of tests would cover tracks on public land, but not tracks marked private on gov maps. (I’m unclear from your earlier messages if this tag is in the Vic Transport dataset or held elsewhere but it must exist somewhere). 

For private tracks (most are on farmland), the only accuracy assessment I can think of would be to take a large, random subset of all private tracks and visually assess whether we are confident they are on private land, using air imagery, Mapillary, land tenure maps and perhaps cadastral maps too. Such a test would be more robust if more than one person did the visual assessments, but that may not be an option. I imagine that most would be perfectly obvious. The question is, how many aren’t? How (and whether) we proceed on tagging private tracks would be best left until everyone saw the results of an agreed accuracy test I think.

If the broader community accepted this approach, then we could step by step get to a stage where OSM provides an up-to-date map of track restrictions on major public lands and accurate depiction of tracks on private land on Vic.

Effectively, this is a strategy of reducing uncertain in access=unknown by focusing on restrictions first, so we have greater confidence that remaining tracks without an access tag in osm are more likely to be, by default, access=yes.

If we ever got to that stage,  it would be informative to analyse all the remaining tracks on public land that don’t have an access tag. These could be separated into 2 categories: those that are also on the Vic gov maps, and those that aren’t. The former would most likely be access=yes, based on all available data, and the latter access=uncertain. How we deal with that is another question. But that question is a long way down the track.

AH: it's just about how to bridge that gap in the meantime before all access tags can be set.

Yes, I guess these suggestions highlight that, if we want to accurately add access tags at large scales, and the community is ok with using rigorous imports of gov data to work towards this, then it will be easier and more accurate to keep chipping away at tagging known access restrictions first. My feeling is that the Vic Gov mapping is among the best in Aus (maybe ACT is better?) so Vic would be an ideal state to start discussing and developing these procedures. 

Thanks for promoting the discussion. Cheers Ian

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-au/attachments/20210824/bd99338c/attachment-0001.htm>

More information about the Talk-au mailing list