A Cloud Of Lightning Detectors


Here’s an interesting project to plot every lightning strike on Earth. Blitzortung is a project that uses many extremely low-cost sensor boards packed with an amplifier, microcontroller, and an Ethernet socket to detect lightning strikes. When multiple stations send all that data up to a server, the location of lightning strikes can be calculated, even if they’re hundreds of miles away from any station.

Each station works by detecting a change in the local EM field caused by a lightning strike with either a large loop antenna or a smaller ferrite core antenna. These signals can be amplified and turned into usable data, time stamped, and sent out on the Internet. From there, it’s a simple time of flight calculation to precisely locate where lightning strikes.

The hardware is actually pretty simple, with based on an STM32F4 Discovery board. A controller includes an Ethernet port, GPS unit, LCD, and all the hardware associated with detecting lightning strikes.

If you’d like to see what’s possible with a huge network of lightning detectors connected to the Internet you can check out LightningMaps for a look at what’s possible.

Thanks [Sean] for sending this in.


  1. Torben says:

    Wonder how this can be news in HackADay ? I have been using blitzortung and lighingmaps for quite a while. What’s next ? Presenting live flight trackers ? Or a “project to index web page content, where content can be found from a simple search field” ?

  2. sean says:

    They’re currently out of stock on boards, looks like Canada could use more coverage

  3. SwiftyTheFox says:

    Just a little hint for other that want to include lightning detection into their projects without too much tinkering – I guess that you can even make an easy arduino shield out of it:

    I used that for my clock since I live in an area with the highest lightning density in europe…

  4. Martin says:

    Just a little hint if you want to implement that feature into one of your projects (home automatisation would be an idea):

    I’ve put that into my clock project – works as it should.

    • kd0pgm says:

      I built a lightning detector around that chip, using the Embedded Adventures board. I wasn’t impressed. It detects lightning, but only about 20% of it. And sometimes the distance estimation appears to be accurate; other times just not even close.

      I feel like if I see a big bolt of lightning, it should see a bolt of lightning too. More than 20% of the time anyway.

      • Martin says:

        I had the same problem at first when I’ve put the reception coil too close to the rest of the electronics – an arduino within 5cm (shield) is enough to limit the reception quality. I was lucky enough to get my hands on the demo kit, so I had a “lightning simulator” as well for debugging.

        @ Admins: can someone kill my double post? I had troubles with the login at the first time…

  5. Joshua says:

    HackADay, thanks for posting this. I am very interested in geolocating lightning strikes and this project seems to be within my budget. I understand that others have seen this system, and are currently using this system, but others have not. I look forward to joining the network and expanding the coverage.

  6. Ast says:

    I was not aware of this project and I think it is really cool. Thank you for sharing this.

  7. qwerty says:

    I also am among the ones who didn’t know about that project. That makes this article useful to me.

  8. Lee says:

    Why no orders from Germany, Switzerland and Austria?

    • James says:

      From watching the “live maps” (http://www.blitzortung.org/Webpages/index.php?lang=en) and enabling the “detectors” checkbox (lower left corner), you will see there are already many detectors in that area (Germany, etc.) (you can only see this if you’re lucky enough and there’s some active lightning going on).

      Maybe this is why they don’t want / need any more orders from that region (if your goal is to provide accurate lightning maps, rather than to entertain “lightning electronics enthousiasts ;) )

      But, I’m just speculating.. there might be another reason.

      • Duwogg says:

        I think they are trying to limit their data pool to protect it from noise. Since lightning detection is basically searching for noise, the more systems you have gathering data, the more corrupt data you’re going to have from anything that creates noise and false positives (EVERYTHING ELECTRICAL), the more corrupt data, the more filtering and server load and such. They want data, just very controlled data. They don’t want you building your own detector from scratch either. I think they explain in the PDF whitepaper you can download.

        • robot says:

          Isn’t noise rejection a benefit of more stations? It allows the system operator to reject any event that isn’t corroborated by more than N stations, maybe with some constraint on distance as well.

          I would think that an approach like that would reduce false positives caused by smaller local events like the AC turning on or something.

  9. I had never heard of this before! Thanks for posting this :)

  10. josh says:

    This is an interesting project, I remember seeing something similar with air quality measurements on kickstarter a while ago. But never heard if it actually became a reality. It makes me wonder what other kind of community data mining projects are out there and if there is a way to bring some of these projects together to share resources and techniques that work.

    I was thinking about some modular system, with a cheap base unit and additional sensor modulars. Then you could pick which projects you would like to help and connect those sensors.

    Just a thought.

  11. sean says:

    I didn’t know about this either. I had previously been looking for a Maxim 1-wire lightning protector (appears to no longer be in production) for a weather station I was working on, based on the work here. http://forum.doozan.com/read.php?4,11039,11039

  12. I figured out how to use the same json long pole requests the sites use to plot the data. Might be cool for a project.
    It shows the next 100 strikes and gives links for ones that strike in Illinois.
    If you want the code let me know.

    • Whatnot says:

      i was impressed that the blitzortung site noticed and notified I had websockets disabled. Normally in the rare cases that a site uses them and things fail you really are left to your own devices as to why, if you normally keep websockets disabled.

  13. Hirudinea says:

    This should be combined with a UFO detector (seriously they exist, here’s a link to a kit… http://www.amazon.com/UFO-02-magnetometer-interfaced-controller-anomalies/dp/B000FVUKKO ) and maybe we could be finding aliems instead of smoldering trees.

  14. Wish they would implement geiger tubes on the same devices. Imagine the radiation maps! That would be awesome.

    • MRE says:

      Check out Safecast.org for radiation maps. Massive data for Northern Japan, but now expanding out all over the world. Need MANY MORE safecasters in the rest of the world.

      With data that is more stable over a length of time (oddly enough, radiation is appropriate here, most of the time), mobile logging is more critical than stationary logging. You want to get as many geolocations as possible. Weather, lighting strikes, and other data that changes constantly, you want stationary sensors.

      Stationary geiger counters tell you only one thing: That the environment on your roof has not changed all that much… but it still tells you nothing about the downspout of your gutters, your back yard, the concrete slap that swing set is on, or the bush your dog is sniffing around. Stationary geiger counters are about as useful as a bus stop in the desert. Certainly wonderful if you happened to be stranded in the desert, at the precise time and place, but otherwise a useless waste of time and effort for 99.9 percent of their readings.

      So long as your ‘mobile’ data logger back tracks through the same space every week or so, you will have sufficient data that a stationary sensor there is not necessary.

      • Sean says:

        That safecast.org looks really neat, thanks for sharing :)

        Interesting that at Fukushima the worst that I can find is 10uSv/hr, which is same dose as the average person receives per day.

  15. Ren says:

    20 years ago, (they may still be around) there was a company that provided near real-time lightning strike data in the U.S. The had a map display showing the strikes which would slowly age off. This data was useful for construction/mining companies that use dynamite, and airline ground crews. A web based display like this would certainly be helpful to many people such as Little League coaches.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s