Using your Android phone to better predict weather

[Victorious Secret]

 

So your Android phone probably has a barometer, did you know that? Well, now you do and now you can join a crowd sourced way of creating far more accurate weather reports. 

 

In other words, we are finally going to kill off the weathermen. God damn weathermen. Always leading us astray in our times of need.

 

 

 

Atmospheric scientists are working with an app developer to take air pressure information that is already being collected from thousands of Android phones and feed it into sophisticated new climate models. If they get enough buy-in from Android owners, you may be able to receive warning hours in advance about thunderstorms and tornadoes coming to your precise location with far more certainty than you can today.

 

http://www.wired.com/2013/01/android-weather-prediction/?tn

[Askew]

Sounds like a cool idea so long as there's a high standard of accuracy in the barometers inside devices, I don't really know how to judge that.

[kuzimins]

Don't weathermen also use barometers?

 

Sounds like a cool idea so long as there's a hgh standard of accuracy in the barometers inside devices.

Not those in our smartphones.

[Victorious Secret]

Don't weathermen also use barometers?

 

Since you're gonna be that guy, weathermen don't have access to hundreds of not thousands of them in a few city blocks (hell, upwards of hundreds of thousands in larger cities) to better map out the weather patterns in real time and make far more reliable models.

[Askew]

Since you're gonna be that guy, weathermen don't have access to hundreds of not thousands of them in a few city blocks (hell, upwards of hundreds of thousands in larger cities) to better map out the weather patterns in real time and make far more reliable models.

 

Yeah I know I just wrote that because I knew you'd jump on it because it was me who wrote it lololol

 

I changed it already.

[kuzimins]

Also the post is one year old.

[Askew]

Also the post is one year old.

 

Oh shit if I'd seen that I'd definitely have been that guy.

[Builder]

So your Android phone probably has a barometer, did you know that? Well, now you do and now you can join a crowd sourced way of creating far more accurate weather reports. 

 

In other words, we are finally going to kill off the weathermen. God damn weathermen. Always leading us astray in our times of need.

 

 

http://www.wired.com/2013/01/android-weather-prediction/?tn

Ah yes, I remember. Back when they thought a barometer would make people buy it over an iPhone...

 

We have this app, it's called Dark Sky, and it works perfectly. We don't need barometers because we have developers that know what they're doing.

 

And if you're worried about consistency...

 

http://gameovenstudios.com/bounden-on-android-delayed/

 

Also the post is one year old.

Yeah, this is like, really old news, brah.

[kuzimins]

 

Yeah, this is like, really old news, brah.

I mean the source.

[Builder]

I mean the source.

So do I.

[flibberdipper]

It seems like a neat idea, but even if it was posted a year ago, how come I haven't heard of it until now?

[kuzimins]

So do I.

Damn, thats a fail for me.

[Victorious Secret]

It seems like a neat idea, but even if it was posted a year ago, how come I haven't heard of it until now?

 

You forget, that if something wasn't published 5 seconds ago it can't be valid news. Even if it was a year old and most people (including me, and I love reading tech news) somehow missed it. 

 

Still seems like a valid thing to post, but I'm not in it for page views or to look good to Linus so he mentions my name during the WAN Show. I thought it was neat, it wasn't posted here before, I posted it. 

[Builder]

You forget, that if something wasn't published 5 seconds ago it can't be valid news. Even if it was a year old and most people (including me, and I love reading tech news) somehow missed it. 

 

Still seems like a valid thing to post, but I'm not in it for page views or to look good to Linus so he mentions my name during the WAN Show. I thought it was neat, it wasn't posted here before, I posted it. 

I'm not sure anyone posts here to attract the attention of Linus...

[Victorious Secret]

I'm not sure anyone posts here to attract the attention of Linus...

 

You'd be surprised...

[flibberdipper]

You forget, that if something wasn't published 5 seconds ago it can't be valid news. Even if it was a year old and most people (including me, and I love reading tech news) somehow missed it. 

 

Still seems like a valid thing to post, but I'm not in it for page views or to look good to Linus so he mentions my name during the WAN Show. I thought it was neat, it wasn't posted here before, I posted it. 

It's neat news. I want it to take off so bad. It would be so cool to never have to really watch the news just to get more accurate weather or what have you.

 

I'm not sure anyone posts here to attract the attention of Linus...

Yes, because the people that come here and post a thread with the name "LINUS!!! PLZ HALP!!!" aren't trying to attract his attention. I've actually seen a topic with that name.

[Builder]

It's neat news. I want it to take off so bad. It would be so cool to never have to really watch the news just to get more accurate weather or what have you.

 

Yes, because the people that come here and post a thread with the name "LINUS!!! PLZ HALP!!!" aren't trying to attract his attention. I've actually seen a topic with that name.

Haven't seen that. Anyways, that wouldn't be the way to get his attention. If you find a good news story, you would post it here and if it was good it would get activity and the most popular posts would probably get spoken about. 

[Victorious Secret]

It's neat news. I want it to take off so bad. It would be so cool to never have to really watch the news just to get more accurate weather or what have you.

 

 

 

Its certainly worth caring about. I'm all for better algorithms and better radar and better modelling in that sense, but sometimes you can never have enough data and if you can leverage the few kb of data from hundreds of thousands if not millions of devices, thats just more for your models to play with and thus be way more accurate than if you were only using 5-6 data gathering stations. 

 

Our local news has access to some 6 weather stations across the city and outlying areas and sometimes they get it wrong hilariously. Having people all over the city and surrounding areas giving you info though? Hello better forecasting. That stuff actually matters for many people. 

[flibberdipper]

Its certainly worth caring about. I'm all for better algorithms and better radar and better modelling in that sense, but sometimes you can never have enough data and if you can leverage the few kb of data from hundreds of thousands if not millions of devices, thats just more for your models to play with and thus be way more accurate than if you were only using 5-6 data gathering stations. 

 

Our local news has access to some 6 weather stations across the city and outlying areas and sometimes they get it wrong hilariously. Having people all over the city and surrounding areas giving you info though? Hello better forecasting. That stuff actually matters for many people. 

I always laugh when they're hilariously

 

Haven't seen that. Anyways, that wouldn't be the way to get his attention. If you find a good news story, you would post it here and if it was good it would get activity and the most popular posts would probably get spoken about. 

Yup. But you don't just do it to try and get the comments and the attention, you should do it since you think that it's interesting, and that's all that matters.

wrong. So funny.

[LukeTim]

I'm no meterologist, but is a barometer really all you need to measure and predict the weather?

How can a pressure measurement, for example, predict thunder and lightning? Distinguish light and heavy rainfall? Snow? Hail?

 

I can't even really see the point of having a barometer in a phone...

[Builder]

I'm no meterologist, but is a barometer really all you need to measure and predict the weather?

How can a pressure measurement, for example, predict thunder and lightning? Distinguish light and heavy rainfall? Snow? Hail?

 

I can't even really see the point of having a barometer in a phone...

Low pressure areas indicate bad weather. This is basic meteorology.

[LukeTim]

Low pressure areas indicate bad weather. This is basic meteorology.

 

Okay. But how can you get any more specific information than "bad weather" from a barometer reading?

I am genuinely interested.

[StingBull]

There are many applications that do this.

[Builder]

Okay. But how can you get any more specific information than "bad weather" from a barometer reading?

I am genuinely interested.

So air moves from high to low pressure areas, it's called a pressure gradient. You know about diffusion and such? Wind is air trying to reach equilibrium between the high and low pressure areas.

 

When this wind comes through, depending on the direction, it can create what is called a front. A front happens when cold, dry air moves out of a high pressure area and into a low pressure area with warm humid air. This warm, humid air collides, and the front is formed. The front sweeps across the pressure gradient, pushing cold air and warm air through the areas. This differential in conjunction with high humidity is what causes storms. Evaporation of water into the cold dry air of the front causes saturation and clouds form. When the clouds become saturated, rain occurs.

 

By measuring differences in pressure, you can discern when a front is going to form and when winds occur. With enough, you could form a location-accurate pressure map and use that data on storm simulation supercomputers to determine when storms are going to form.

 

Disclaimer: I'm not a meteorologist, nor am I a meteorology student. I'm a well rounded computer scientist. If you want to really know this stuff, look up articles.

[LukeTim]

So air moves from high to low pressure areas, it's called a pressure gradient. You know about diffusion and such? Wind is air trying to reach equilibrium between the high and low pressure areas.

 

When this wind comes through, depending on the direction, it can create what is called a front. A front happens when cold, dry air moves out of a high pressure area and into a low pressure area with warm humid air. This warm, humid air collides, and the front is formed. The front sweeps across the pressure gradient, pushing cold air and warm air through the areas. This differential in conjunction with high humidity is what causes storms. Evaporation of water into the cold dry air of the front causes saturation and clouds form. When the clouds become saturated, rain occurs.

 

By measuring differences in pressure, you can discern when a front is going to form and when winds occur. With enough, you could form a location-accurate pressure map and use that data on storm simulation supercomputers to determine when storms are going to form.

 

Disclaimer: I'm not a meteorologist, nor am I a meteorology student. I'm a well rounded computer scientist. If you want to really know this stuff, look up articles.

Thanks! There's clearly a lot I didn't know about the weather.