Possible error in Linz tree import

Communities Österreich (Austria)
, ,
1

Hi, sorry for using English. I was browsing trees with weird height/circumference ratios and stumbled upon the 2015 Linz tree import. It seems like a lot of trees with a circumference of 0.01 were added. I think they are likely to be errors because:

For example there are 7m tall Carpinus betulus in Linz with 0.01 circumference, in the rest of the world the average circumference for a 7m Carpinus betulus is around 0.48.
I also checked some in streetview imagery and they look wrong, you can see this one on Mapillary for example: link.
Or check the tree row in Harrachstraße for example, the trees appear to be all similar (likely planted at the same time as most tree rows) but the first four trees on the left are 0.01 while the rest of them are >1m, despite having the same height.
94% of circumference=0.01 values in the whole OSM database are all in Linz.

I don’t know deutsch to browse into the import documentation to understand what happened. Maybe 0.01 was a wildcard value meant to identify trees with unsurveyed/unknown circumference? Maybe an error in the original data or in the import process? Is there a way to understand and fix this error?

7 Likes
2

In the newest tree database the values seem to be correct Baumkataster (Linz) - Datensatz - data.gv.at. You could import the trees in the newest database and replace trees with circumference=0.01 in OSM based on tree:ref=* and tree:ref_area=*.

However if a new import is too much effort, I would just delete all values with circumference=0.01 in Linz. Better an empty value than a false value. And I don’t think the circumference is super important for OSM.

2 Likes
3

I don’t know how to import in OSM so I will go for the deletion since they are wrong values. Based on what you said if the refs are still there, someone interested in importing the new data could re-add them without problems.

EDIT:
I have used the latest data to fix 344 circumferences and deleted 274 since I couldn’t find the refs in the updated file: changeset. It is possible that the trees have been cut down, the reference numbers have changed, or another issue I don’t know of.

4

Looks good👍