top | item 3215610

How Dark Sky Works (Weather prediction using computer vision)

13 points| thegrossman | 14 years ago |blog.jackadam.net | reply

7 comments

order
[+] itsnotvalid|14 years ago|reply
I like the idea except it is U.S. only.

Okay back away from trolling, I thought these kind of modeling was something to be done on server side, but you app seems to be doing it on the client side. What's more, there are many more interested in these (including weather stations) things but I am guessing there is no way you guys are giving away the algorithm to make the prediction.

[+] thegrossman|14 years ago|reply
We definitely want to expand it beyond the US! But dealing with each country's different network of weather radar is a daunting task.

As far as giving away the algorithm: Unfortunately, it isn't a single algorithm that can be packages nice and neatly. So I'm not sure how that would work.

But, we do plan on developing an API that 3rd-parties could use.

[+] thegrossman|14 years ago|reply
As promised in an earlier thread (http://news.ycombinator.com/item?id=3186978), I've written a blog post describing how Dark Sky operates under the hood.
[+] lsllc|14 years ago|reply
Fantastic, can't wait to get the app! I've always wondered about taking weather radar over time and creating a "heat map" of rainfall to prove that during the summer the t-storms seem to split in half, pass us by and rejoin, leaving me to go out and have to water the lawn while ALL the adjoining towns get inches of rain!
[+] maxerickson|14 years ago|reply
Have you compared the results of your computer vision approach to the velocity data provided by the NWS?