Several times a day, NOAA POES satellites pass overhead and transmit a live camera view of the ground below them. Using a cheap RTL-SDR dongle and a homemade antenna, I've been experimenting with receiving and decoding these signals to get pictures...from SPACE!
My experiments with Software Defined Radio (SDR) and RTL-SDR started a couple of years ago after I purchased dongle off Amazon for about $10. Initially, I only dabbled in receiving terrestrial radio, particularly weather radio broadcasts and public police broadcasts. Soon though, I learned that it is possible to receive satellite transmissions with a specialized antenna.
In particular, NOAA POES satellites broacast analog low-resolution scans of the ground below them using the Automatic Picture Transmission (APT) protocol. These signals are broadcast at around 137MHz (depending on the particular satellite) using a right-hand circularly polarized FM carrier signal.
The first antenna I made for receiving NOAA APT signals was a simple 137MHz dipole antenna I built following this guide. While it was easy to build and was able to sucessfully receive APT signals, the gain was not very good and there were a number of blind spots where the radio signal faded out completely.
In pursuit of better reception, I then built a (rather makeshift) right-hand circularly polarized 137MHz Quadrifilar Helical (QFH) antenna following this guide. Although this antenna was harder to construct and required more particular wiring, it is specifically attuned to the APT signals and resulted in much clearer reception.
Close-ups of the two antennas can be seen below:
With the hardware out of the way, it was all a matter of finding the right software to decode the APT signals into images. Since I'm on Linux, I chose to use GQRX to drive the RTL-SDR dongle and read the radio signals into sound. I also installed GPredict to track the satellites and account for frequency shift caused by the Doppler Effect. Finally, I used WXtoImg to translate the sound from my sound card into NOAA satellite images.
Luckily, since RTL-SDR is a Software Defined Radio, it's laughingly easy to automate timing, tracking, and capture of the satellites and their data. Using a couple of homebrew python scripts (some more slapdash than others), I set up a system to track the nearest NOAA satellite, timestamp the images generated by WXtoImg, and archive them for future analysis. I've picked some of the clearest and most interesting downlinks to share below.
Click to see the full-resolution image.