Hey everyone - I’m working on mapping my university more and wanted to import an existing tree “database”, for lack of a better term. The website is https://otm.wcu.edu and there is a data download button. They have coordinates and I was able to put them in Tableau - is there a way to similarly put objects into OSM? (Area in question: https://www.openstreetmap.org/#map=17/35.30899/-83.18444&layers=N))
Hi and welcome to OSM and the forum
Before considering an automated import please read https://wiki.openstreetmap.org/wiki/Import . Here we describe how an automated import can be performed and what conditions have to be fulfilled (licence etc).
Regarding mapping trees see https://wiki.openstreetmap.org/wiki/Tag:natural%3Dtree . In my opinion trees should only be mapped when they are significant, like a landmark, historic or POI.
For CSV import try JOSM with the “Open Data” plugin.
This is very much a personal opinion, and “significant” is probably a difficult thing to treat objectively.
There are numerous examples of imports of tree registers typically at a city level, and quite a few of us map trees in urban environments for a range of reasons.
It would be an import so please do abide by the import guidelines. These are there for good reasons, for instance the Belfast tree register turned out to have trees over 2000 feet tall when the data were analysed.
What I would recommend if you have time is taking a, preferably random, sample of trees from the data set & visiting them to check:
- Is it still there?
- Is it the same type of tree the dataset purports it to be?
- Is it in the same place? Insofar as one is able to tell.
I’ve done this on a few tree datasets and roughly:
- 1% of trees are no longer there
- upto 10% of other data is incorrect (usually species/genus).
- there may be a small number of extreme outliers, peculiar values (e.g. tree pit, planned & so forth)
You may check on other attributes too (e.g…, height, circumference, diameter etc) to see how accurate the dataset might be. A good sample is perhaps 100 trees or 1% of the data. It takes about 1 minute per tree or a little less if there are identical trees in rows or avenues.
Note if the dataset is just of tree locations then I would suggest not importing it into OSM as it can be used as is & then one doesnt have a problem of updating it. If the dataset has additional information or you want to enhance sparse information, particularly species or genus then it is likely to be more valuable, particularly if you plan to enhance this with more information.
Certain items of data are rarely worth importing (height), whereas others (diameter, girth, date of planting) are both more valuable & less subject to short term change. Knowing the kind of tree is much the most useful thing: this can then be used for things like urban foraging, teaching tree identification, shade calculation, and so forth.
I also miss a suitable licence for the data on the website which is absolutely needed for OSM imports.
I would recommend visiting around 50 trees to validate the data.