Indicate overlapping observations for map pins

I think that it would make sense to include some map symbology indicating that there are overlapping observations. It makes sense when recording interspecies interactions like the hemlock and hemlock woolly adelgid.

Nice idea, could be along the lines of how Google Earth does it; when you hover over the pin it’ll diverge radially with each overlapping pin having leader line pointing to the location.

Just as a note, for some reason iNat doesn’t automatically register a vote when you make a feature request, so even though you are the originator of the request you still have to vote on it.

1 Like

To me the devil’s in the details. Many many observations like this won’t have exactly the same coordinates. Earlier today I took photographs seconds apart with my phone: one of a monarch, one of the milkweed it was on, but both photos don’t have the exact same coordinates or the exact same horizontal accuracy. So what would the thresholds be here?

And if iNat does this for observations that don’t have the exact same coordinates, that would likely necessitate extra calculations when loading the map.

1 Like

Cragislist does something sort of like this, splitting out into smaller groups as you zoom in.

image

1 Like

consolidation of markers already occurs when observations are close enough to each other. so there is already some sort of threshold set for this in the current mapping interface.

probably the easiest way to accomplish this without changing the mapping interface too much is to simply allow users to select whether they want to see the grid markers, the pin + obscured circle markers, or the heatmap. right now, the grid markers show up in the Explore page map at zoom levels 2 to 10, and the pin markers take over from zoom level 11 and on. the grid markers and heatmap do provide an idea of observation density via opacity and color, respectively so if you could use those rather than the pin markers in some of the more granular zoom levels, then you could get an indication that a particular spot has a higher density of observations.

(if you want to see what the different observation map tile sets might look like at different zoom levels, you can look at https://jumear.github.io/stirfry/iNat_UTFgrid_based_density_map_for_Leaflet.html. this page shows some of my own custom density markers by default, but you can use the layers selector in the top-right corner of the map to see iNat’s standard grid markers, heatmap, etc…)

another thing that could probably be done relatively easily is when you click on a particular pin, you could have the pop-up show not just the latest observation’s details, but also additional text at the bottom that could be something like “+10 other observations”.

i don’t really advocate new “symbology” to indicate multiple points because at lower (less granular) zoom levels, sometimes most of the markers on the map might represent multiple observations. so the only time such symbology might be useful is at really granular zoom levels, i suspect.

there’s really not an elegant way to do this technically in the current mapping interface without totally changing how it works, i think. but even just conceptually, exploding points is fine when you have a limited number of points, but when you have the potential for hundreds or thousands or even more observations in a given spot, then exploding points might not be the best way to present such data, i think.

2 Likes

After discussion, it’s not something we’re opposed to but it would require technical exploration that we probably don’t have the time for in the near term.

just for grins, i mocked up what this could look like in my own mapping page (https://jumear.github.io/stirfry/iNat_map.html). here’s an example screenshot from that page with the popup displaying some “additional observation” text:

1 Like

I agree that a radial split is far from the most elegant way of dealing with this. It was meant more along the lines of exploring options for different ways of graphically exploring information. The nested system that Craig’s List uses, as @jwidness suggested, is a good idea.

@tiwane, the threshold issue is a good one to pay attention to. I’d suggest that there are two broad ways to deal with this, one being assigning whatever the average location accuracy range for a smartphone (which is what I suspect most users are using for location data, or more accurately, what the accuracy range for a single antenna AGPS system averaged over the world is), and the other assigning a static radius, something like 20m or 100m (I’d vote for the larger number) as belonging to the ‘same’ point. Whatever system might be employed would have to scale with the map view.

Your display reminds me of the grid mapping in observation.org:

1 Like

it looks like ArcGIS Living Atlas has ingested a portion of the monthly iNat AWS Open Data Set into one of their layers. at lower zoom levels, the layer does some hex binning to present data in a way sort of similar to the Craigslist example here. you could also probably query observation data from the layer, although since it’s based on the AWS Open Data Set, it has all the limitations of that dataset.

i don’t think this directly addresses the feature request, and i personally wouldn’t advocate for this kind of visualization over iNat’s current map visualizations, but i thought it was interesting, if only as a curisoity.