Recently I’m into tree mapping. I created a MapRoulette challenge about trees validation and a diary entry. I thought to check for the tallest trees in the OSM database and I’d like to share them with you as a fun bit of trivia (queried with QLever from 29.04.2024 OSM dataset):
Not surprisingly, the tallest trees are sequoias from the USA. I haven’t checked them all yet, but I suspect the entries from other countries might just be errors. Additionally, I’ve added some of the tallest trees (according to Wikipedia and monumentaltrees.com) that have not been mapped yet or lack the height tag. It would be cool to have them.
Perhaps next week I will update this thread with a list of the thickest trees.
node/9877651243 seems to be an obvious typo. The other trees next to it have a height of 10 or 15 meters. Also that tree has a circumference of just 1 meter, which further indicates that this is a simple typo.
You might find other suspicious trees by looking at the circumference Tag, and maybe even a circumference to height ratio.
Also looking at imagery for the listings for France: OpenStreetMap , & Sudan: OpenStreetMap , neither of them appear to be anything like ~200m tall!
Wiki page for the French school: Lycée Bartholdi — Wikipédia also includes a photo of the courtyard trees, which show them to be about the height of a 3 story building + gable roof - 19m perhaps?
Yep, this is also why I run these kinds of queries, for QA reasons. QA tools for trees are very scarce at the moment. The only one I know is Osmose, which flags trees overlapping buildings and trees with type= values. If someone knows more, please let me know.
EDIT: Oh, the like from Mateusz reminded me there’s also his tool that have a check for species added as wikidata/wikipedia instead that as secondary tags such as species:wikidata/species:wikipedia: https://matkoniecz.github.io/OSM-wikipedia-tag-validator-reports/
And iD has a check for leaf_type tagged as wood (es. wood=deciduous).
Of course! This thread is not about tall trees only, that was only the first post, as announced I would like to post (weekly? monthly?) more data, maybe dedicated to specific genus or species. I wanted to make a new post about the thickest trees but the data seems to be a mess (lot of incorrect values), there are still ~300 values to be fixed, so I have to postpone it or it would be a list of errors. So maybe a list of oldest trees could be the next one, thanks for the idea! I also created a task to fix trees’ start dates: here
In Italy the government has 10 different parameters for the notability of monumental trees: age, circumference, height, crown diameter, shape, ecological value, botanical rarity, plant architecture, landscape value and historical value.
Thanks for the idea. I found a possible problem with a tree import looking at circumference/height ratio (by species): here
If you search all the trees that have been mapped as areas and look for the biggest ones (in terms of geometry), all you find is a lot of objects where people presumably meant to type natural=wood. All from JOSM it seems - not sure if there’s a specific reason.
The tree that is actually the biggest in the world, by area, has been double mapped as a natural=tree node and a natural=wood area, so it won’t be found with this query…
flag trees that have leaf_type!=broadleaved|needleleaved|palm|mixed|leafless
species check: check species values with only one word
genus check: flag genus values with more than one word
but imho what trees mapping needs is something like NSI for trees, I made a table that can be easily converted to .csv with the most popular species values in the database: link
This list could also be used to flag incorrect values. For example, it would flag an Acer platanoides with leaf_cycle=evergreen, suggesting that either the species or the leaf cycle is incorrect. Or could suggest to add leaf_cycle and/or leaf_type if missing (in Poland something similar is made based on species:pl=* values: here).
I’m not sure why OSM should be recording data such as leaf type, whether leaves are evergreen or deciduous, et.c. This information should be derived from the species name, not added separately, and especially not on a mapping website where there is the potential for error. For most types of trees (species) that are found in public areas, there ought to be a Wikipedia page.
This would follow the database principle of not duplicating data.
Isn’t it true that the same species can be evergreen in one climate and deciduous in another?
Some of this information is a generalization if the mapper doesn’t know the species. For example, a lot of people don’t know the difference between pines, spruce, fir, etc, but they know they are needle-leaved - and this bit of information may be helpful to a data consumer if the species is not recorded.
Evergreen and deciduous - yes, but teak is the only one that I’m sure of at the moment - a tropical dry-weather deciduous tree, but a rain forest evergreen. But I believe I’ve seen the California live-oak Q. agrifolia in full leaf in a native Californian winter, decades ago, but when I saw it another time in London in late winter, it had lost most of its leaves.
For mappers that don’t know the species, surely they should be encouraged to leave the genus name only, or just a description, to stop misleading others?
There is already an informational warning (others) Presets do not contain property value - Value '*' for key 'leaf_type' not in presets. though the preset currently only includes broadleaved, leafless and needleleaved
Nice, these two would be useful for a lot features besides natural=tree. Note to myself: Beware of semi-colon separated multiple values for certain main tags.
All other suggestions need some external list which are out of the scope of mapcss but could be written in java.
(“Berg-Ahorn” should be in