Amazing. I've been tracking a delivery (Seiki 4k monitor incidentally) which I believe is currently on this flight: http://www.flightradar24.com/ABX2040
It always looks as though the course they take is unnecessarily far north. I have to consciously remind myself that the map projection and the prevailing winds play a factor. This really does beautifully illustrate why they take the route they do.
The map projection on google maps is only chosen to produce good maps at every point by themselves (square city blocks will still be square, wherever you zoom in). I assume they would have chosen a different design (a globe) if they designed it to be used as a world map service.
No, it's chosen because it preserves direction. They use a single Mercator projection centered at the equator.
It's a pragmatic choice, but it's certainly not chosen to produce good local maps. It produces bad local maps everywhere except the equator.
If they wanted to produce "good maps at every point by themselves", they'd use a different projection every time you recenter the map.
It's a terrible projection at every point not on the equator. However, it does preserve direction, so straight N-S or E-W streets will appear perfectly straight.
However, square city blocks _do not_ appear square. You just don't notice the distortion unless you're quite far north or south.
Sort of, journeys in the other direction are pushed even further north (or, actually north of the straight line) though which I believe is because of the winds. - i.e. it's a factor, isn't it?
[edit] I think it's fair to assume the winds are relatively stable and more intense further north so eastbound (US > Europe) flights might deviate even further to the north.
Beautiful. Reminds me of how I used to run xplanet as live desktop with bump and cloud maps (updated every few hours), accurate stellar background, etc. I wonder if there's something similar, more modern, with more possible data feeds.
Hi. I'm the creator of the site. Unfortunately, each time the globe orientation changes, the distortion caused by the projection needs to be recalculated. Tried to think of ways to make this faster or save computations, but haven't found any good solutions yet.
From a usability aspect at least, you could keep the same distortion until the interaction ends, and then tween the distortion values to the new orientation over a second or two. Technically it'd be 'wrong' for a short time, but it'd be nicer to use.
Not sure if it is feasible but a slider control for height would show the transition from high altitude (circulation currents) through to surface conditions really nicely. Hard to cache the animation I imagine.
It's modelling all the water flow levels in the world using very detailed GPS data. I mean, you can model a sea raise or a river level raise.
I didn't even get 1 point. OK, design isn't as cool as this wind one, but I think it deserved a bit more love since they process a crap ton of terrain data for the water flow modelling...
35 comments
[ 3.3 ms ] story [ 111 ms ] threadIt always looks as though the course they take is unnecessarily far north. I have to consciously remind myself that the map projection and the prevailing winds play a factor. This really does beautifully illustrate why they take the route they do.
It's a pragmatic choice, but it's certainly not chosen to produce good local maps. It produces bad local maps everywhere except the equator.
If they wanted to produce "good maps at every point by themselves", they'd use a different projection every time you recenter the map.
It's a terrible projection at every point not on the equator. However, it does preserve direction, so straight N-S or E-W streets will appear perfectly straight.
However, square city blocks _do not_ appear square. You just don't notice the distortion unless you're quite far north or south.
Great circle distance (Curves in Mercator, straight lines on globe) vs. Rhumb lines (Straight in Mercator, constant course angle against meridians)
http://demonstrations.wolfram.com/ComparingLoxodromesAndGrea...
[edit] added picture for explanation
http://i.imgur.com/wJS0ry1.png
Lines of same color have identical course
The appearance as a northbound curve is the effect of the straight Orthodrome/Great Circle line projected into Mercator.
http://en.wikipedia.org/wiki/Transatlantic_flight#Transatlan...
[edit] I think it's fair to assume the winds are relatively stable and more intense further north so eastbound (US > Europe) flights might deviate even further to the north.
http://en.wikipedia.org/wiki/North_Atlantic_Tracks#Route_Pla...
http://earth.nullschool.net/#current/wind/isobaric/10hPa/ove...
Edit: Wasn't working before, Chrome restart fixed it. My bad.
https://news.ycombinator.com/item?id=6924854
Relative Humidity [%], Air Density [kg/m3], and Wind Power Density [kW/m2]
https://twitter.com/cambecc/status/438674275285757952
here is the obligatory link to the old discussion: https://news.ycombinator.com/item?id=6924854
Awesome work regardless though.
Not sure if it is feasible but a slider control for height would show the transition from high altitude (circulation currents) through to surface conditions really nicely. Hard to cache the animation I imagine.
Example of a high altitude with temperature:
http://earth.nullschool.net/#current/wind/isobaric/10hPa/ove...
And many different projections, Stereographic is awesome:
http://earth.nullschool.net/#current/wind/surface/level/over...
http://earth.nullschool.net/#current/wind/isobaric/1000hPa/o...
Please someone make this as an app for a desktop background!
-----
I found this option for total cloud water[0] and the checked the weather in France[1]. Pretty useful.
[0]:http://earth.nullschool.net/#current/wind/isobaric/1000hPa/o...
[1]:http://www.weatheronline.co.uk/France.htm
It's modelling all the water flow levels in the world using very detailed GPS data. I mean, you can model a sea raise or a river level raise.
I didn't even get 1 point. OK, design isn't as cool as this wind one, but I think it deserved a bit more love since they process a crap ton of terrain data for the water flow modelling...
http://www.youtube.com/user/Suspicious0bservers
Also, how live is this data? If the data is live enough, this could be immensely useful for small-craft off-shore fishing.
All around, superb work!