Personal Weather Stations Helping With Weather Forecasting

Weather Storm writes "Weather information from thousands of personal weather stations are being used for weather forecasting by several private and government agencies, including the National Oceanic and Atmospheric Administration (NOAA) and the Department of Homeland Security (DHS). The Citizens Weather Observation Program (CWOP) was created by a few amateur radio operators experimenting with transmitting weather data with packet radios, but it has expanded to include Internet-only weather stations as well. As of September 2007, nearly 5,000 stations worldwide reported weather data regularly to CWOP's FindU database. The weather data is forwarded every 15 minutes to NOAA's Meteorological Assimilation Data Ingest System (MADIS), checked for temporary and spatial consistency, than utilized by computer forecast models and internal forecast verification programs. In a Febuary 2007 report (PDF) DHS listed CWOP as a national assets to the 'BioWatch' Network, stating that data from personal weather stations could be useful in weather forecasts for hazardous releases. In 2007, the FindU server received 422,262,687 weather reports which is a 29.5% increase over 2006." The personal weather stations certainly come in stylish shapes.

Comments

Installation Guidelines?

[im_thatoneguy]

Are you required to consistently install your weather station? It would seem like you could get some inconsistent data if some people mount them in the sun while others in the shade etc. Or are they just depending on the volume of reports to create a median result?

Mixed opinions

[ShaunC]

NOAA is the single remaining government agency that I still trust implicitly. DHS is nearly the polar opposite. I'm not sure what to make of a collaboration between them, but I'm encouraged by the fact that they're taking input from regular citizens. No doubt NOAA is behind that initiative.

There has been tremendous pressure from commercial weather interests to devalue the NOAA/NWS, to decrease its funding, to cripple its information distribution, in favor of making people pay to get the sort of weather data that NOAA has been distributing for free ever since they were chartered. I'm happy to see that new programs with public participation and cooperation are still being created. WX data is crucial for everyone, let's keep it open and free and public.

Re:Mixed opinions

[gotzero]

I do agree this data should stay open. They should make a larger appeal for weather data points. I would happily give them the data from a station at my home as long as I knew it would be a public information good...

Re:Mixed opinions

[GPSguy]

Actually, the data are now likely to remain free for some period. NWS got a boost when Rick Santorum (R-PA) failed of reelection last time.

I'm not an NWS employee but I do work with them on several projects. I've been present when some of the private sector have called for NWS to stop giving away data, forecasats, etc., to the public... just as long as they kept giving it away for free to the private companies. This isn't right. It's just plain wrong.

I do participate in CWOP, and several of my projects benefit from MADIS. Both of these are really cool projects, and they really do help improve weather forecasting both by allowing more data for data assimilation, whereby we use real world observations to enhance the data for the models, AND by allowing for more data for forecast verifications. NWS would love to have the funding to place more weather stations under their direct control out there for these purposes. Congress has seen fit to not make that possible. MADIS is an example of of a stellar program to gather in data from a variety of sources, and then to process it to afford some indication of data quality based on neighbor-comparison, reasonableness and internal consistency. CWOP is a sign of citizens volunteering their data to allow a really good Federal Agency to do their jobs better.

Oh, and while I'm at it, NWS is one of the few agencies who compute and use real metrics for performance, including how well their forecasts perform. So many other agencies seem to create metrics designed to prove they simply know how to play the game. I've gotten to see the process surrounding how they look at storm-based warning evaluations, and the effort going into making sure the evaluations mean something, and that they represent what's really happening were eye-opening for me.

Yeah, they are doing it right.

Re:Mixed opinions

[ScrewMaster]

The GP is right. The outfits that are currently taking data from the NOAA and disseminating it (for a fee) are dead-set against the NOAA releasing weather data in an open format. They apparently have a sense of entitlement here, feeling that it's their data, even though it was acquired by a taxpayer-funded agency.

Grants

[ascii genitals]

It would be great if I could apply to a program/agency and get a standardized weather reporting kit in order to participate for free. Basically just have them distribute grant money in order to set up even more stations.

Re:Grants

[jamesh]

That would be pretty cool. The weather station nearest us (~10km away) recorded no rain about a week ago, but we got heaps. The sort of rain that has been coming through here has been really really patchy the last few years. So if we had a weather station it would have recorded completely different rain, humidity and wind speed for that period. I'm sure that sort of activity happens all over the world, so more data collection points would give a much better picture of the 'now' weather. Not sure if that translates to better prediction though.

Re:Grants

[Anonymous Coward]

I believe I've seen something like that before. Either the NWS or CWOP do indeed have a program like this. They basically loan out additional hardware for this purpose. It's usually older decommissioned hardware (not 100% automated) that has been replaced with a newer, more up to date AWOS/ASOS.
Check with your local NWS forecast office, they might have some idea about this. We always need more surface observations.
The cost of purchasing your own hardware, depending on what capabilities you wish to have, generally starts at about 2 grand.

As for the NWS' dedication toward freely/openly available information is amazing. Although, even given the advances made in forecast models they would have nowhere near the accuracy they do now without input from additional sources like your average wx geek. So, they almost have to retain an openly available design. Additionally, I believe they have made a significant effort to remain as open as possible with both their information and software.

Re:Grants

[TapeCutter]

"Not sure if that translates to better prediction though."

Computer climate/weather models use finite element analysis. If they get enough data to improve the resolution of the models, and enough computing power to make use of the better resolution, then yes the predictions should improve.

Disclaimer:IANAMeterologist

Applications of distributed technology

[Dutch Gun]

This is an interesting application of distributed and networked technology (I'm not sure what else to call it - would this be classified as distributed computing?) - using technology to extract information from an extremely large number of processing nodes in a way that is impractical on a smaller scale. It reminds me in a way of projects like Seti@Home/Folding@Home. That is, using the combined technological resources from a large number of people to do some pretty cool things that just weren't possible a few years ago.

It will be interesting to see what other sorts of projects in this vein take shape in the near future.