More data for better forecasting?

Question

Here's an interesting proposition. Apparently, some smart phone manufacturers have installed pressure sensors to help determine the phone's elevation. Cliff Mass, an atmospheric scientist, has envisioned a different application. If a large database of these phone users could report instantaneous pressure data and geographical location, this data could be used to improve weather forecasting.
http://cliffmass.blogspot.ca/2013/02/smartphone-pressure-observations-take.html

This sounds like a good idea to me but one issue I see is a privacy issue of cell phone location. What are your thoughts?

Answer 1

Nice idea, but I don't think it works quite how folk have interpreted it here.



The GPS in the phone already returns a z-axis position (that's the altitude in the wgs84 coordinate system).



There's no problem at all pushing the data from the phone, it's absolutely commonplace these days.



A similar technique is already in regular use to do with travel times. The time taken to get from one place to another (in built up areas) changes as the amount of traffic changes. Several journey time app's use the users' GPS in the background, checking their speed, then using this to adjust the network model to current conditions.



It would be very easy and cheap to use this technology to build a very detailed and real-time pressure map.



Someone will figure out how to use this to work out the likely paths of tornado's in real time. That seems like an immediate win.



Edit. Re: using cell towers for location.



This is only a poor approximation of location... It can be done by triangulation between nearby cell towers IF the phone has contact with more than one at the time. But that requires the cooperation of the telco to achieve. Usually this technique only returns the location of the cell tower being used. Also, there's no z-axis value returned from this technique except that of the cell tower, which means my phone would report pressure where I am, but report it for the cell tower's location.



For those concerned about privacy, the app doesn't need to forward any user details to make this work, all that's needed is, lat, lon, alt, pressure and time.



The processing load isn't big in the scheme of things, but for performance reasons you'd probably seperate the data acquisition and management from the presentation and analysis.

Answer 2

I think it's a great idea, and if you let people opt in or out then there are no privacy issues. I think you need both the GPS and pressure sensors going--otherwise you can't tell the difference between pressure changes and elevation changes.



Most of what Sagebrush worries about is nonsense--there are no pressure differences between the inside and outside of a purse (where does he get this stuff?) and the pressure differences between the inside and outside of a building are usually quite small, certainly less than 1 millibar. Also (and what everyone in here seems to not be aware of) raw data always needs to be checked to ensure accuracy and reasonableness, if that weren't already be being done our weather models would be garbage.



Cell phone towers are under consideration as the Doppler radars of the future, since there are so many of them around.



The accelerometers inside cell phones are already being used in a similar fashion to observe earthquake movements, and I would guess that the noise issue associated with that application is much higher.

Answer 3

Phone location is already provided to the companies providing the service and thus law enforcement agencies when requested. It is probably no different to using facebook etc on a phone which allows location to be detected. It is just as simple as giving people the option wheterh or not they want to participate so I don;t foresee any privacy issues.



Wow good point saggy, since scientists just think up ideas then say them out loud without every considering the minutiae, there is no chance that Mass has considered things like calibration before putting this forward. Good thing there are true Poes I mean scientists like you to state the obvious isn't it?

Answer 4

as long as there is consent and data is limited to location and weather and does not connect the actual individual who owns the phone what's the problem? It's unlikely that this would be a commercial app with hidden data collection to target shoppers.



this app would be similar to using the inertial sensors for detecting earthquakes.



this is exciting, like millions of PC's analyzing SETI.

Answer 5

It seems the user of the phone has to install an application and allow it to upload that information. I don't see how this is a privacy issue as the user has to give explicit consent.

Answer 6

Ha! Ha! Somebody needs an experience in instrumentation. Great if all were calibrated and operated properly, but with all the jiggling and hard usage you would get noise up the wazoo. If you were in a car with the fan blowing that would be another factor. Buildings often pressurize the inside for efficiency. A woman's purse could have a different pressure than ambient pressure. That is just to get us started.



And then I don't like the idea of Big Brother.



James Hansen would have a field day with that set up. He has corrupted data that is highly controlled. What do you think he would do with all this uncontrolled data?

Answer 7

You would have to understand the difference between discrete and analogical signaling. Then ascertain nominal values for for each. Pretty much a dead end street, as Sage suggested.

Answer 8

It couldn't hurt, but it would probably require more powerful computers to process all the data.