Listening In On Weather Balloons With RTL SDR

sonde

Every day, twice a day, over 800 weather balloons are launched around the world at exactly the same time. The data transmitted from these radiosondes is received by government agencies and shared with climatologists and meteorologist to develop climate models and predect the weather. Near [Carl]’s native Auckland, a weather balloon is launched twice a day, and since they transmit at 403 MHz, he decided to use a USB TV tuner to receive data directly from an atmospheric probe.

The hardware portion of this project consisted of building a high gain antenna designed for 162 MHz. Even though the radiosonde transmits at 403 MHz, [Carl] was easily able to receive on his out-of-band antenna.

For the software, [Carl] used SDRSharp and SondeMonitor, allowing him to convert the coded transmissions from a weather balloon into pressure, temperature, humidity, and GPS data.

34 thoughts on “Listening In On Weather Balloons With RTL SDR

  1. So those strange transmissions above 400MHz are weather baloons, thank you for revealing the mystery :) Or are they?
    I received something similar at 428MHz with funcube dongle pro and single wire “antenna”.

  2. Hey,
    I find myself in need of some support regarding SDRSharp and was wondering if any of you would know the way to go.

    I have a EZcap tv tuner on hand (EzTV645) and I was wondering why is it that ADSBSharp detect it but pop a warning message saying : Device has been detected but is busy. From that point software is up and running but I can’t select my dongle in the drop down list and start,thus SDRSharp got nothing to work with.

    Has anyone here got this problem? If you did,how did you guys solved it?

    Thanks

  3. An interesting project something I’ll try to keep in mind. In where that antenna gain is achieved by directing radiation from one direction to more desired direction. In understanding that it’s easy to see most omnidirectional gain vertical that favor terrestrial operation may not be the best satellite use. I ask a local RF engineer if the is a better vertical antenna to use for satellite use that would give the longest in sight time, and hopefully gain as well. information on the web supports his suggestion I’d imagine all verticle antennas would have null above, but don’t have clue if it would be a big enough hole to be a problem most of the time

    1. I was concerned about the collinear antenna not working well as it’s gain is directed at the horizon, but it worked much better than expected. I can usually track the entire flight. I also tried a simple wire antenna which was useless.

      I think the collinear antenna worked well as the sonde never passes directly above my antenna, and was always somewhere on the horizon. Perhaps if you’re tracking from the launch site a more omnidirectional antenna will be needed.

  4. I got SondeMonitor working, and I think I have a very faint signal with my RTLSDR, but I can’t tell if it is just noise, I guess I’ll fire up SDR-Sharp again later and see if it is still there. I really need to pickup a real antenna.

  5. here’s a hack for you: ‘frustrated long time hackaday reader resorts to url crafting trickery exploiting the tags system to find old posts as hackaday no longer seems to have a search function!’ i wish there was a writeup for this one and not just a youtube video as i can’t get youtube; i havee a very legitimate interest in anything that accumulates weather data over radio and not the internet, though i did enjoy the seemingly totally unrelated antenna build link as i am extremely interested in ais for the same reason i need the weather data.

    1. Hi, i’m the uploader of the video. Curious as to why you can’t get youtube? I’m sure my video isn’t region locked as there’s no music or copyright issues. Anyway the video isn’t a tutorial, it’s just showing the data working.

      The antenna link is there because I used that 4-element collinear AIS antenna to receive the sonde signals. Even though it’s out of band it works very well. There are calculations on the antenna page where you can calculate the required lengths for a 400Mhz antenna which should work even better.

      Basically, the steps to get this going are very simple.

      1) Get a RTL SDR/Funcube dongle/any SDR or some sort of scanner that can tune to 400Mhz working and inputting sound into the PC. There’s lots of tutorials on this stuff already.

      2) Build/buy and set up an antenna that works well on 400Mhz.

      3) Find out where and when sondes are launched in your area.

      4) Tune to 400Mhz and find the sonde signal. Pass the audio to Sondemonitor through windows stereo mix and you’re away! Also, if you want to track the sonde location you’ll need to set up the GPS tracking in Sondemonitor, which is a matter of downloading an up to date Alamanac file (http://www.navcen.uscg.gov/?pageName=gpsAlmanacs) and knowing the Sonde launch coordinates at least roughly.

      To get AIS working in this way its the same steps, except build an antenna for 162Mhz, and pass the audio to a program called ShipPlotter made by the same company that makes SondeMonitor.

      1. I appreciate your reply. It’s my cellphone provider they’ve decided i stream to much video and need to be punished. I was going to ask if there was a linux compatable software for this purpose, but judging from the other comments there’s no way to do this with noaa baloons in the us so im probably going to try to pull data from the noaa satilites as per that post from a little while ago. Would any of you have any suggestions or pointers for constructing a more compact antenna for ais? Preferably encased in pvc pipe for weatherproofing? Though looking at the antenna build it seems pvc affects the signal so maybe i should wrap it in fiberglass instead? In my situation this antenna would be mounted on a 40 foot pole with an unobstructed horizon in every directionbut it can’t hang down and will be subjected to extreme weather conditions. I would ideally like to get something like a 30 mile range if i could. My setup is a modified police scanner jacked into a ubuntu laptop and ive got all the software figured out for ais, but i can only receive one frequency at a time like this and ais seems to use 2; is this a big deal? Is one more common then the other? Should i just get a second scanner?

        1. Yeah that sucks about the US using a different protocol. I didn’t even consider that. You could always get into launching your own sondes :D

          On the AIS antenna link he states that encasing his antenna in PVC pipe only affected the resonant frequency by 2%, so I wouldn’t worry too much about the effects of PVC. I’m not totally sure about this, but I think there are also 1/4 wave collinear antenna designs which are more compact. Anyway, I think any 162Mhz antenna you mount on a 40ft pole with unobstructed horizon will give you excellent AIS results.

          AIS uses two channels but I believe its the same data on both. I think commercial AIS monitors both frequencies in case of interference on one channel.

  6. Does this SondeMonitor software work in the US/North America? Given that it’s so often the case that we North Americans insist on using our own proprietary protocols on the radio waves I wouldn’t be surprised if SondeMonitor doesn’t work here. And the God-awful 1995-era web site for the software doesn’t seem to give any indication.

    1. Sondemonitor doesn’t currently work in the US. Radiosonde vendors have never made any attempt to use common signaling protocols. Vaisala has 70% of the world market, but L-M Sippican won the re compete for NOAA’s sonde supply contract. Per the current chit-chat on the Sondemonitor Yahoo discussion group, the amateurs haven’t yet cracked the code on the Sippican MkII.

    1. If you look up the SondeMonitor Yahoo group and join it, there are a bunch of example sound files of different protocols in the files section.

      Also the posted video does have sound of RS92SGP telemetry which should decode if you pipe the youtube audio to SondeMonitor.

      Just note the GPS data won’t work as the GPS almanac data will have probably changed by now.

    1. Most vaisala have a “dump” receiver GPS that needs post-processing (see sondemonitor), other radiosondes like MODEM M2K2 have a full featured TTL serial NMEA GPS. Think about it , daily 800 gps receivers are throw to land and sea.

      1. Oh. And I thought the 3 gigantic state-of-Texas sized sargasso seas partially did well because water blocked degarding uv rays. A bounty should be mandatory on these. If we paid more for weather broadcasts maybe they’d get better.

  7. Is the high-gain antenna really necessary? I was a meteorology student back in the early 1990s, and we used to launch and/or track radiosondes all the time. I just don’t remember using anything more complicated than a j-pole.

    But, what I do remember is having to run a little strip (sort of a punch card) that came with each sonde through a reader hooked up to the receiver to calibrate the instruments. Do you still have to do that?

  8. I find this very interesting. I live in canada and the northern part of the province of Alberta . I have looked for info on Canada weather ballon launchs and fregs used but have come up empty, can anyone help please.
    thanks
    don henry
    ve6vac

  9. If one bought a vaisala sonde off the internet brand new and had sondemonitor, do you have to calibrate the sonde before launch, or can you launch it and have sondemonitor pick it up and send the data back without having to use vaisala equipment to calibrate it?

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.