24 comments

[ 2.5 ms ] story [ 55.0 ms ] thread
A similar one, but for earth :-) http://earth.nullschool.net/
Our coordination was so precise it's surprising we didn't cause some kind of race condition in HN. (edit: No wait, beat me by at least 20 seconds)
And your answer shows earlier cause your karma :-(
The stuff done by this duo is amazing (they're currently leading Google's data visualization research group, see the About page).

This particular map was expanded by another dev into an epic global map of different climatic variables:

http://earth.nullschool.net/

Check this guys: http://windyty.com

Wind, temperature and waves.

This one is like a culmination of the other two projects. More features and a lot smoother than earth.nullschool.net, I can really see this being useful for navigation.

It's also a perfect example of why open source kicks ass, with the chain of inspiration from one project to the next.

No globe though but the mercator projection (awful for a global map like this).
Well, this is one of the rare cases where using Mercator as the default projection is justifiable, since this map is most useful for navigation.

An option to switch projections would be very nice, but I can imagine very difficult to implement. earth.nullschool.net has this option, but I'll wager it was only doable because of the the ultra-minimalist map used.

We were looking at this in the office on Friday, watching Hurricane Patricia (it's not much use for the high-speed predictions we usually need, but it's a really beautiful way to visualise/understand the data after the fact). The Mexican government prepares for these storms, issuing catastrophe bonds that pay a high rate of return but lose all their value if high-impact storms hit populated areas so the government can fund their aid/reconstruction efforts. As someone who sometimes finds it difficult to follow all the links between what we do and real-world value, it's great seeing financial products being used so directly for good.

[1] https://www.windyty.com/?2015-10-22-12,17.958,-101.558,5 (hit play and watch the next 48 hours, 200 mile an hour winds disappear as soon as they hit land, it's amazing)

>catastrophe bonds

Would you care to expand on these and how they work? They sound interesting!

Cat bonds were developed by insurance companies to offset the potentially large losses associated with natural disasters.

Investors purchase cat bonds, which pay interest as long as no predefined trigger conditions are met. If the trigger condition, such as a damaging hurricane, occurs, investors forfeit their principal, which is used to pay claims arising from the catastrophe.

Investors like them because they offer enticing, uncorrelated returns, and insurers like them as an alternative to reinsurance.

You can learn more about them here: http://www.artemis.bm/library/what-is-a-catastrophe-bond.htm...

And actually for the whole world, not just one country...
Fantastic visualization! Any idea about the tools used to achieve that result?

Inspecting the code a bit I'd say regarding the viz it's based on leaflet.js+canvas(custom overlays)? No trace of d3js, threejs,etc.

I'm happy to see riotjs so nicely used as well.

I saw this posted probably here a while ago. Why is Alaska left out still?
It would be enlightening to see a time-lapse of this graphic that instead charts the course of a year ... or decade ... or century.

This site is great work - I enjoy rediscovering it on what I suspect is an annual basis.

Does anyone have recommendations on good toolsets for visualization like this? I'm diving into canvas/d3/raphael but would love to look into other options if they're out there.
The fantastic thing about this project is that they've open sourced the code [1]. With a little help from NOAA and some work with the GRIB data format [2], I was quickly able to build this into an customized interactive museum exhibit that's in two science centers now.

We could never have afforded to build this from scratch. Open Source to the rescue, once again.

[1] - https://github.com/cambecc/earth

[2] - https://github.com/scimusmn/earth-latest-data

The map's creators, Viegas and Wattenberg, gave a talk on how they made the visualization at the Eyeo Festival 2012 -- https://vimeo.com/48625144#t=16m55s . They show the experiments that took them from static vector fields to the fading brushstrokes analogy: Static vector field: https://vimeo.com/48625144#t=20m34s | Particles: https://vimeo.com/48625144#t=21m16s | Particles with geographic outlines: https://vimeo.com/48625144#t=21m47s | Warping blocks: https://vimeo.com/48625144#t=22m21s | Static brush strokes: https://vimeo.com/48625144#t=23m16s | Fading brushstrokes: https://vimeo.com/48625144#t=24m57s