Wait, did I say control, I meant query… Sorry to disappoint!
I am a home automation nerd. Recently I’ve noticed that as it gets closer to warmer or cooler extremes that it takes longer for my hvac system to bring my house to the temperature I want. I’ve also noticed that NEST claims to automatically learn these factors. Not to be outdone by the Griswolds, I decided to look at building this into my system.
curl http://weather.noaa.gov/pub/data/observations/metar/decoded/KMSP.TXT
The output is basically as follows:
MINNEAPOLIS-ST PAUL INTERNATIONAL , MN, United States (KMSP) 44-52N 93-13W 265M
Oct 01, 2013 - 10:53 AM EDT / 2013.10.01 1453 UTC
Wind: from the WNW (290 degrees) at 13 MPH (11 KT) gusting to 24 MPH (21 KT):0
Visibility: 10 mile(s):0
Sky conditions: mostly clear
Temperature: 68.0 F (20.0 C)
Dew Point: 48.9 F (9.4 C)
Relative Humidity: 50%
Pressure (altimeter): 29.82 in. Hg (1009 hPa)
Pressure tendency: 0.14 inches (4.6 hPa) higher than three hours ago
ob: KMSP 011453Z 29011G21KT 10SM FEW150 20/09 A2982 RMK AO2 SLP094 T02000094 51046
cycle: 15
I subtracted or added the difference in temperature to my desired temperature and am experimenting with how much more quickly I need to fire things up based on that (for my hvac system seems to be about a minute per 10 degrees of delta), but there are definitely plenty of ways to go about such number nerdery. Either way, I can now control the temperature based on the weather using curl, which is basically controlling the weather in my house, so not as untrue a title as with most front-page newspaper articles…
Finally, there’s also a REST API, available from NOAA at http://graphical.weather.gov/xml/rest.php.