Update Level 2 Standard Place Boundaries for Australia

I’m not sure how iNaturalist identifies locations, in particular for records in Australia.
Scientific records in Australia tend to include locality names and there are clearly identified boundaries for these localities.
I have records in, say, the locality Vinegar Hill. iNaturalist shows Gatton (the name of a nearby small town, but also a locality in its own right).
This is confusing.
It is easily possible to download a kmz or kml file with localities - at least for Queensland.
Is it possible that iNaturalist could use the coordinated of records to put them into the correct localities in Australia - OR provide a field where observers can indicate the locality?

In other areas with this sort of geographic data the inat admins have added places of this sort so you might ask them at help@inaturalist.org

Can you share an example or screenshot of what you’re referring to? It sounds like it may be the automatically generated names from Google Maps, which I’ve learned to just ignore. You can already manually edit the locality name from the website and Android app.

1 Like

it was hoping it would be something like the Vermont towns we added - discrete jurisdictions that already exist - where if there is an existing KML of the boundaries the admins may be able to add them as part of the official place infrastructure. Bouteloua is correct, if you’re just using Google locations… they don’t work very well with the site.

@gordon_c can you please change the subject to a request? And as @bouteloua said, please provide some examples with URLs.

I wonder if what is meant is the typical “collecting areas” such as what we have for NZ from Walker/Crosby 1988:

The Localities I am referring to are “address” localities. So if I am asked for my address, or if it comes up in, say, the post office database it will be 99 Xyz Street, Vinegar Hill, 4343 where Vinegar Hill is the locality and 4343 is the postcode. That’s why I think iNat could relatively easily pick up on this dataset. And yes, it turns out that it is a Google calculation that is built into the iNat system that is providing the current locations.

@charlie - exactly. There are kml files with the boundaries and names of the official Localities in Australia. I’m not sure what you mean by ‘using Google locations’. I work with decimal degrees of Lat and Long because they are precise and widely used, and because databases can generally translate this into e.g. an address that includes the locality if this is needed. I get the Lat/Long from either my GPS, a GPS app (MotionX) or, if I didn’t collect it in the field, from Google Earth. I’ve just had all my observations moved across from Bowerbird so it would be quite a task to manually change all that show the wrong locality.

I could be wrong, but as far as I am aware this would be a locality naming error to fix on Google’s end, rather than iNat’s.

I’m new to iNat, having just migrated across from the Australian Bowerbird website, so I’m still discovering elements of iNat that others are probably very familiar with. I’ve just discovered the basis for the Locations shown against my observations - the locations shown are the old Queensland Shire names - see attached screen shot. These were changed as a result of amalgamations to create larger local government areas years ago. From my point of view, if iNat used the current local government area (LGA) names this would be useful, because a lot of what I upload is from a community group which does wildlife surveys within that LGA. However the LGA is still at a rather coarse scale for recording wildlife observations, and what is used now is the official Locality name (boundaries and names available in a kml file). So from my point of view the Locality name would be the preferred Location for an observation. However the ability to search/sort on the basis of LGAs would be very useful, because the LGA is the lowest level at which government has policies and regulations related to biodiversity conservation. I had a project on Bowerbird in the LGA name (Lockyer Valley Region biodiversity) where users could list their observations.

1 Like

Ah, okay, I think I see what you mean now. Translated to iNat-speak, it sounds like your request would be something like “Update Level 2 Standard Places for Australia”. You can read more about “Standard Places” here: https://www.inaturalist.org/pages/inat+standard+places (the content there was written in 2016)

Just in case this is complicated by the BowerBird data migration which included some observations from gordon_c…

During the BowerBird data migration, we just grabbed latitude and longitude which means that the ‘locality notes’ vacuum was filled by Google’s geocoding service (which we don’t control) which takes a lat/lon and returns an address string that we stick in ‘locality notes’. There’s been quite a bit of questions/confusion about this from participants in the BowerBird data migration. And we’ve since noticed that BowerBird has an ‘address’ field which we could use to populate ‘locality notes’ so we’ve update the migration script to use this (its been in effect for data being migrated starting around 6pm PST Tuesday) and plan to go back and try to update the ‘locality notes’ for the remaining data transferred from BowerBird with this ‘address’ data.

But as others have said in this thread, we can’t control what the Google geocoder returns, but the default behavior of ‘locality notes’ being populated by the Google geocoder can always be overridden by typing something manual into that field.

Hope this helps


This can branch off as a new request if necessary, but would it be possible to parse out only information on country, state, county - which are useful - and not street addresses, which are so often wrong as to be worse than useless? I would love an option to never have street addresses filled in for my observations. And certainly not the house number, which is the part that is most often erroneous, and is irrelevant when an observation is made on a public street, in a park, or on a beach.


Android is sort of doing this.

I’ll close this thread.