A note about "Geoprivacy" and the TEA's Ontario Butterfly Atlas

The TEA maintains an online Atlas of Butterfly observations for the province of Ontario. The Atlas aggregates observations from a number of sources, including iNaturalist. In recent years, iNaturalist users have made increasing use of the "Geoprivacy" setting to obscure the locations of their observations. In the past, the Ontario Butterfly Atlas was able to include obscured observations if observers granted project curators permission to "see" their hidden locations. In 2021, iNaturalist changed how they handle obscured observations, and now iNaturalist obscures both the location and the date of these observations. Because of this change, and the increasing concern around privacy settings in general, the Ontario Butterfly Atlas will no longer include obscured observations in our database, even when the observer has granted us permission to "see" the precise location/date of these observations.

With thousands of observations posted to iNaturalist every year, we don't have the resources to keep track of who is "OK" with the level of information that the Atlas reveals and who is not. The TEA simply can't take the risk of revealing information about observations that observers want or need to keep hidden. Therefore, the safest course of action is for the Atlas is to simply exclude all iNaturalist observations where the Geoprivacy is set to obscured (or private). Long time contributors to the Ontario Butterfly Atlas who want to see their observations included in the Atlas should consider whether obscuring their observations is really necessary. There are certainly a number of circumstances where it's the appropriate thing to do, but only in very rare circumstances is it because the butterfly in question needs "protection".

In summary, if you obscure your observations using the Geoprivacy setting, the Ontario Butterfly Atlas will not include those observations. Furthermore, I will no longer verify the identifications of these observations (because TANSTAAFL).

Please note that this policy will (eventually) be applied retroactively to all observations in the Atlas database, so long term contributors may see some of their older observations "disappear" from the Atlas if those observations have the Geoprivacy setting set to "obscured" or "private".

Also note that I'm not telling anybody that they shouldn't use the Geoprivacy setting. What I'm saying is, if you ask that your observations be obscured, the Atlas project will respect your request. Just consider whether it's what you really want.

For those who are OK with the level of detail revealed by the Butterfly Atlas, but don't want to set the Geoprivacy on their iNat observations to "open", there is a solution. Observers can submit their observations to the Atlas Project using a spreadsheet. Contact me (rcavasin) for details.

Posted on December 17, 2021 02:38 PM by rcavasin rcavasin


Just to be clear - can you please confirm the steps to reverse previously obscured or private butterfly observations.
If one changes the privacy setting of individual posts to "open" does that allow observations to be properly seen or does it require additional adjustments

Thank you.

Posted by chdonati over 2 years ago

For the observer, all you need to do is to change your Geoprivacy setting for the observations in question. On the Atlas side, it gets complicated as to when/how that change will take effect, since the Atlas doesn't get updated "live".

For old observations (those added to iNat before 2021):

if the observation is in the Atlas, and you set Geoprivacy to "open", there will be no change. The observation remains in the Atlas
if the observation was previously excluded from the Atlas, and you set Geoprivacy to "open", we will attempt to include the observation at some future date, probably at the same time that we are removing all the old observations that have Geoprivacy set to "obscured"

For newer observations, I'm currently processing observations that were added between Jan-Oct 2021. If you set Geoprivacy to "open" at this point, the observations will probably get added at some future date when I perform at "audit". If I get a bunch of people chiming in with "WAIT! I'm changing my Geoprivacy setting to OPEN!!", I'll consider re-doing the download of the 2021 data so that those observations can be included.

Posted by rcavasin over 2 years ago

Thanks, Rick. I will review my posts from this year. I am pretty sure I switched after an earlier conversation with you but will check to make sure. Butterfly sightings are special and I have noticed a change in what I see just in the few years I have been on INat. It is good to have any data possible recorded. Thank you for your considerable investment in the Butterfly Atlas.
Best wishes

Posted by chdonati over 2 years ago

I am increasingly disappointed with iNaturalist for it's poor computer recognition of any species, and it's "democratic" opinion of ID when as many as 4 different experts "opine" differently, some with ridiculous "suggestions", while other recognised experts are beyond dispute. I only wish there was an Ebird for butterflies! Is there not some way I can record butterflies directly to the Ontario atlas? Or am I doing that already? Keith Wickens, Ottawa.

Posted by kawickens over 2 years ago

Hi Keith,
This is a bit off topic, so I will take the discussion to a personal message. I will only comment that IMHO, both the computer vision and the community ID works quite well most of the time (at least for butterflies here in Ontario). Nothing is 100%. Other models implemented by other citizen science websites tend to work very poorly by comparison.
Cheers, Rick

Posted by rcavasin over 2 years ago

Add a Comment

Sign In or Sign Up to add comments