Obscured observations only showing month observed for other users

An earlier thread from @bouteloua (I forget the title) but - a single woman out hiking - don’t want to announce - I am now here. And also we don’t want to announce - I am not at home, now is a good time to break in. iNat is also social media with its risks and benefits.

2 Likes

Pinned location is ONE circle. Works for me.

This one tells you it is in the hospital parking lot etc.
https://www.inaturalist.org/observations/107245253

4 Likes

I do understand that. iNat just seems like a very unwieldy way to gather personal information. Clearly it does happen - it just surprises me.
And saddens me.

1 Like

I saw a situation on iNat a while back where a (male) user was repeatedly demanding that a particular (female) user un-obscure her moth observations that were clearly taken at her home porch light. He was claiming they were ‘useless’ unless the exact location was visible. But he didn’t seem to be making any such demands to any other users who posted similar things… fortunately he got moderated pretty quickly, but it was still very uncomfortable.

In conversations with male friends and colleagues, they’re usually completely shocked about how much harassment women face on a regular basis, and how much time we have to spend thinking about our personal safety.

6 Likes

Actually not, it’s done so that they create a big rectangle, it was discussed in another topic.

1 Like

I was talking to my son about this because he shares the account with me and I wanted to see what he thought about that idea, and he brought up something that I had totally forgotten about. Wouldn’t the metadata still be accessible this way? You can see the location in the metadata of a photo when uploaded on iNaturalist but when it is obscured, you can’t view that.

I think only some observations include the exact coordinates in the metadata and some don’t, but I’m not really sure how to know what would contain that in the metadata and what wouldn’t.

3 Likes

I think this will depend on the device that you’re using to take the photos and its settings for adding metadata to files. It is possible to strip metadata from files before uploading, so you could explore that for added security if needed.

4 Likes

I just checked my observations. For home based observations, I use the large circle method with ‘open’ privacy settings. At first I thought the exact metadata was available but then I remembered to check using an incognito window (to simulate being signed out of my account) and the metadata was not visible at all.

edit to add: I stopped the experiment too soon. Seeing metadata is a function of being signed into an iNat account or not. Not a function of my observation vs someone else’s observation.

3 Likes

yes, if you are recording coordinates on your photos. you would need to strip the metadata before uploading, as cthawley notes. however, the better thing is to turn off location services on your device when taking photos at locations you don’t want to reveal.

2 Likes

hmm, interesting. I viewed my own observations while signed out because I know that I can see some data on my own observations that isn’t visible to anyone but me. And I couldn’t see the metadata. So I assumed it wasn’t visible to anyone but me. But that’s not the case. It’s not visible to someone not signed into any iNat account. So, signed in… I could see your metadata. Signed out, I couldn’t.

1 Like

yup. that’s why you need to make sure you avoid uploading the metadata one way or another if you want to go with a “manual obscure” approach, as opposed to using the system’s built-in obscuring functionality. (you could do a combination of the approaches to hide the photo metadata from other users, but that wouldn’t help avoid the situation described in the original post.)

2 Likes

You’re correct about iNat’s obscuration, but that’s not what I was talking about (but I did mention “obscuration” in my post, so I apologise for the confusion).

I was referring to the practice of setting the location of an observation to somewhere in the vicinity of the true location, then stretching the accuracy circle far enough to cover the true location. If you had many of these sorts of observations, it’s pretty easy to narrow down a window that the true location must be inside of, by identifying the space where all the accuracy circles overlap.

Well, then @dianastuder answered to it better than I did, you pin one location and only use it for your house!

2 Likes

That would not be an issue with a pinned location and a constant radius of uncertainty. All observations would sit on the same spot in the map

Correct.

I have a ton of obscured observations at my house. I am considering changing them to a circle with a pinned location nearby, as discussed above. But I have a question about how this metadata stripping works. I didn’t strip the metadata when I originally uploaded the photos, because the observations were set to obscured. If I go back and change those observations to Open location now, with a random dot in a circle, will the true metadata show up now? Is there a way to strip the metadata now after the observation has already been uploaded? Or was the metadata permanently stripped when I uploaded the original observation as obscured?

yes. photo metadata on open observations can be viewed by other users.

technically, uploading the photo will strip the metadata from the image files stored by the system, but it records the metadata in the database.

once the metadata has been recorded, there’s not a way to delete the bits of metadata without deleting the photo record. (it would be a nice new feature to be able to delete bits of metadata that you don’t want to show up with your photos though.)

on existing observations, to get rid of the photo metadata, you’d have to delete the iNat photo record entirely, strip out the metadata from your original source image file (or make a copy that lacks the metadata), and then load the image file (without metadata) as a new photo in your observation. (then at that point, you could change the location of the observation and un-obscure it.)

don’t just delete your observation and reload as a new observation because that would be annoying. don’t just remove your photo from the observation (as opposed to actually deleting the photo record) because that would potentially leave you with an orphaned photo record whose metadata would be viewable by other viewers.

1 Like

How would you find your orphaned photo records and how do you delete them if they are no longer associated with an observation?

there’s not a good way for regular users to find orphaned records, but others would still be able to access them. for example, if someone saved a link to a particular photo record, then they would be able to still view it that way, even if it was no longer associated with an observation. orphaned photo records are supposed to eventually get deleted eventually, but sometimes you never know. it’s best to explicitly delete them rather than letting them get orphaned.

My question would be, is the original location of the photograph included in the metadata which is saved in the iNat database? My brief investigation of a relocated observation shows many photo metadata fields, but only the new coordinates (the false location I assigned when uploading) and not the original coordinates. But I did not look at very many examples.