Seeing a Silk Floss Tree (Ceiba speciosa) blooming in Portugal in November was super odd for me. As a Brazilian, I am used to see these flowers much earlier in the year.
So I wondered when people saw them bloom around the world.
I co-vibe-coded a little open-source vanilla javascript app to see that:
I have also added a little button to try and help to get more phenology annotations; most observations don’t really have that degree of information now:
AI made it easy to have an interface in multiple languages. It should be straightforward to add more (though with the usual caveats of machine translation)
i think it’s an interesting idea, but the issue with this sort of thing is that it requires getting observation-level data. for taxa with large numbers of observations, you either have to download a ton of data, or you download only partial data. (i think your page does the latter.)
you also end up with potential situations where markers overlap other markers, and you have to decide which one to display. for example, if one is yellow for August and another is red for May, and they overlap, which one do you display?
for what it’s worth, iNaturalist has a page that allows you to see a time series animation of observations. it address both of the problems above by showing different intervals at different times. for example, this cycles through 12 sets of data (one per month) in series: https://www.inaturalist.org/observations/torquemap?interval=monthly&taxon_id=126634.
i’ve also created custom maps using iNat’s UTF grids. one technique allows you to display 4 sets of data in four corners, and it could display, say, 4 quarters of data. the advantage over a time series animation is that you can see all the data in one glance. here’s an example of a custom map that shows Monarch butterfly observations over 4 months in 2019: https://jumear.github.io/stirfry/iNat_UTFgrid_based_custom_density_map?example=5.
1 - Yeah, the dashboard does a random sample of the observations there. I found it okay to see trends in general
2 – I am displaying both markers with some transparency; the sampling in fact helps with that, there is not that much room for overlap. Some overlap is okay for an exploratory tool, but I agree, “publication-quality” plots would need some alternative there
Also, markers are clickable and show a popover, I found it an interesting to navigate to particular obs.
3 – Oh, I did not know about : https://www.inaturalist.org/observations/torquemap?interval=monthly&taxon_id=126634, that is interesting, thank you for sharing. Though, there is not a . Though I really wanted to plot this third “monthly” dimension in some static way. Different colours in a loop seemed to be the simplest way, but I am open
this has 2 layers that you can activate to click and see the latest observation that falls in a particular marker. (one allows you to see the latest observation in the subset, and the other allows you to see the latest observation in the whole set.)
this torque map is really cool. I noticed that it slows down between intervals if there are a bunch of observations during that particular interval… i’m guessing it’s adhering to the rate limited fetching requirements in the background as it’s retrieving data for each individual interval. does anyone know if there’s a way to tell it to retrieve all the data for all intervals up front?
could there be a button to record an expensive animation such that it could be shared broadly without the iNat servers having to re-execute the queries?
Thanks, that’s very useful! It doesn’t give a step-through of months on a map, but I can pick them one at a time, and it lets me overlay all observations to see what fraction of plants have ripe fruit in different months.
I am trying to find nymphs clues of two species of bugs (we know adults and I IDed all of them) - now by adding three fields of adult sp1, adult sp2, nymphs in that link - i can readily see on map if there are exclusive pockets of one adult only and find better mapping to nymphs
