On species pages you can look at what time of the year people have seen species along with a few other nice data sets. I think similar to the time of the year graphic we should have a time of day graphic on species pages. This would allow scientists to easily look at when species are typically active as well as tell users the best time of day to find a taxa they’re looking for. While many may appear more common during the day or night, it could also highlight things such as the hours of the day a particular butterfly species is active. (Some butterflies are really only active at certain times of day).
Personally I was thinking it could be separated either in hours or in half hours with the observation being rounded to the nearest time. As much as I would love it in minutes I don’t know the feasibility of showing that nicely while taking lower processing power.
This would be a feature only on animal pages as with other taxa like plants or fungi though they may have different times of day in which they are active this does not change their observability.
It’s a nice idea, but a few problems. All it takes is your camera/phone to be set to the wrong time and you’re immediately generating lots of erroneous/misleading data. I know one user (can’t remember his name off the top of my head) who says on his profile that all his times are off by 12 hours. So all of his night photos would seem like they were taken during the day and vice versa.
Also, if someone picks up a shell and photographs it at home, the recorded time will be wrong unless they remember to change it to when they actually found it.
Time observed also does not necessarily correlate with activity. If a bird is photographed heaps at ~1pm it doesn’t mean that’s when it’s active; 1pm is around lunchtime and people might be taking photos during their break, thus skewing the data and giving a false impression of activity.
I mean it will always be a bit messy for example some species may be common throughout the year but appear uncommon in the winter due to people going outside less. However that’s the nice thing about larger data collection is that noise becomes less apparent in a way.
I think the wrong time on camera/phone is the biggest problem. Many people don’t have the time set correctly on their devices. This wouldn’t be a problem if no time was set, but the set time is incorrect = incorrect data.
Similarly, many people traveling will upload observations after they come back from traveling or at the end of the day.
If this is done from a phone often the time will be the time it was uploaded (image imported to iNat app rather than taken via the iNat app), and when it’s taken with a proper camera the time will be wherever the camera is set to and I don’t know anyone who sets their camera time when they travel… myself included even though I set the time on all my other devices).
Many people don’t even set the camera date/time initially when they buy the camera, so those can be way off.
To the phone point I’ve personally never had an issue with my photos being set to the incorrect time and I don’t take photos from in the app. I can’t say anything about cameras because well I don’t use one. Though I would imagine that most observations are set to the correct time frame and incorrect ones could sum up the typical “noise” of data.
I think it is possible that some meaningful signal could show up through the noise. Only way to say for sure would be to try a couple of test cases where we have a decent number of observations and know what the expected pattern should be. Anyone have the chops to take a crack at that with some downloaded data?
Don’t know that I would want to limit the functionality to just animals either. Plant species that flower morning and/or evening, but are very inconspicuous mid-day, are not uncommon and could also have detectable observation patterns.
Perhaps you could mitigate some errors in the data by adding a control. For example add a click box on the side of the observation (where the "Annotations’, ‘Projects’, ‘Tags’ panes are currently.
The box might say something like: “I wish to submit ‘time of day’ data for this species; I confirm that this time is accurate to within 30 minutes of actual sighting”.
You would probably lose a lot of data from people not bothering to confirm their times, and it’s not the most practical solution. Just a thought to get others thinking.
:D Cool idea though, it would be awesome to see this. I also agree that many people are not totally accurate with their ‘time’ of observation sightings.
I often use the app to add photos I’ve taken on my phone camera. It has not been a problem as photos taken on my phone (and I assume most phones) are time-stamped, which means the original picture time is imported into the app with the photos. So this shouldn’t be too much of an issue, unless (as @thebeachcomber mentioned) phones are set to the wrong date/time to begin with.
I think that’s me you’re talking about!
I’ve noticed that not only people’s cameras but also the software they use to process an image can change times and possibly the uploading to iNats can get confused and set to some other time zone. When setting up a project for a day field blitz I find I have to add twelve hours to either end of the day to capture all the photos that have odd times for whatever reason.
However, globally I suspect that the total picture would end up being fairly accurate. Just because a few people claim to have seen the particular moth in the middle of the day doesn’t mean it’s true.
This would be extremely useful, +1! I am studying a few jumping spider species where it would be incredibly useful to know what time of day they are being found on. Making my own graph would take much too long.