Visualizing 100k Years of Earth in WebGL | Hacker News Reader

Visualizing 100k Years of Earth in WebGL

technistuff.com

83 pointsagnosis20 comments

Nice work! This is like a much better version of Ancient Earth[0], which I made ~10 years ago using GPlates[1]. I like your approach of rendering the map itself from data, which makes it continuous, rather than just wrapping map textures around a globe.

[0] https://dinosaurpictures.org/ancient-earth#240

[1] https://www.gplates.org/

Cool viz! The demo shows the channel forming gradually but iirc there's actually evidence it happened super fast - like a giant lake in Doggerland had a dam that broke and "fast flushed" to carve the channel in one catastrophic event

Northern winter and summer look very different[1] (and those don't even capture sea ice).

I've puzzled over how to represent such variation. Especially with deeper time paleogeography, where those 100 kyr of ice ages and sea level changes can be the variation which needs to be aggregated.

One approach is sampling biased by similarity. So you snag points in time, from similar times of year and climate states. If the interactive allows twiddling those, it might not be too misleading.

One approach is open-shutter motion blur. The sometimes-there sometimes-not semi-transparent ice sheets.

One approach is, maybe call it flickered multiples. If one was showing a year, the visual could rapidly cycle through the months.

Any others?

Clouds raise similar issues. It's interesting how time-blurred cloud cover changes with seasons and decades and climate.

[1] Jan 2004: https://eoimages.gsfc.nasa.gov/images/imagerecords/74000/742... (2 MB) June: https://eoimages.gsfc.nasa.gov/images/imagerecords/74000/743... from Blue Marble Next Generation https://visibleearth.nasa.gov/collection/1484/blue-marble Much higher resolutions are available there.

Thats an interesting point. I think that if the climate difference is important I would allow the user to toggle between summer and winter. Or choose based on the context if you are showing specific events like wars that were impacted by winter weather. From my research (not professional or scientific) ice sheets didn’t move much between seasons so I wouldn’t include them. When you have very large intervals of 100k years when you go further back there could be several ice ages in between so I don’t think it makes much sense there. In what context do you think that this would be important to consider?

My context was mostly "zooming" time. Like power-of-ten zooms of size, but for time. Or timelapses. So a single rendered moment, represents an aggregate of some extent of time. Say you're linearly scrubbing over a 10 kyr span, made of 100 steps. Then each step summarizes a 100 yr extent, with all its heterogeneity. Similarly for a span of 100 yr of 1 yr steps, or 100 Myr span of 1 Myr steps. How might one do that well, gracefully handling the heterogeneity?

Now educational graphics are notorious for negative training. Some aspect is done carefully perhaps, but others less so, and a menagerie misconceptions are reinforced. Some solar system introductions for example, start with such a wretched graphic, which so reinforces many common misconceptions, that even before you hit the text, you've dug a net-negative learning hole that you're never going to climb out of. Many readers would understand the topic better if they'd never seen the page.

Raising a general question: in what ways might we render an Earth globe, that gracefully aggregates/summaries some extent of time? How do you handle things that varied during that time? Diverse clouds, day/night, diversity of seasons, diversity of years, climate changes and sea levels and ice ages, moving continents and paleoclimates. An extent of some mere 10s of Myr encompasses west antarctic as both temperate rainforest and arctic tundra - so what might one paint, to represent west antarctica over that extent as a whole?

So I was taking advantage of your post, to raise a general question which has long puzzled me.

Thanks for sharing your nice work.

Err, out of respect for Murphy, I'll emphasize that none of that was intended as critique. It's just a challenge I've repeatedly faced, mentioned in hope that readers attracted by the post might have thoughts on it.

Interesting, but you're missing geologically important proglacial lakes, like Lake Missoula and Lake Agassiz.

Also the African humid period isn't visually apparent (https://en.wikipedia.org/wiki/African_humid_period).

But very cool!

if we're nitpicking, glaciers push down the crust they are on (https://en.wikipedia.org/wiki/Isostatic_depression) so when glaciers melt the land underneath is at first underwater before emerging

Hardly nitpicking. The size of the two glacial lakes was phenomenal, bigger than current great lakes. Glacial lake Missoula emptying into the Pacific dug out the Channeled Scablands in Washington State.

Very cool, the interactivity of this makes this a much better learning tool than a set of static images (for me, at least).

One minor suggestion: on mobile put the date scrubber on the bottom, otherwise my thumb gets in the way of the UI while sliding back and forth through time :)

Also, I’m not sure if a log scale for time makes sense in this case. It confused me for a second, at least.

Great job, thanks for sharing!

Log scales can be educationally confusing. One alternative is a stack of scrubbers of different zooms.

Thanks for the feedback! I used log scale because it will be easier to show historical events on the timeline once I implement that. Since much more happened closer to present day

Awesome. I'd suggest implementing lakes that were created by ice sheets blocking rivers in the northern hemisphere (in Canada & Siberia).

Great point, I thought about it but decided it was too hard to do. But I will take another look and see if I can find some good data sources for it.

Any geography buffs able to explain why the greenspace in Northern Africa seems to ebb and flow across this time period? That looks quite interesting but I can't seem to find anything about it.

This is great. Always wanted something exactly like this for teaching or learning History and Geology. For some reason I had a real struggle with history in books format.

doesn't seem to work on macOS with either Safari or Chrome. am i missing something?

Yeah - not rendering on MacOS Chrome or Firefox.

Current FF on Linux renders OK (on Intel HD Graphics 630 @ 1.10 GHz/Kaby Lake).

Reply on news.ycombinator.com