HF Spectrum Analysis - Weather Fax

The first thing we’ll look at in the recordings are the weather fax broadcasts. There are several transmitters around the world broadcasting pictures of the weather to all manner of users, but mostly mariners. A variety of images (called “products” in meteorology) are available, with lots of charts, graphs and satellite photos available to decode. The signal is fairly easy to spot with its tone-infused-with-ticking, kinda like a Geiger counter. There’s a tone for white and a tone for black. The ticking on each black pixel will be familiar to owners of dot matrix printers in the past. Solid black, horizontal lines will come in with a full black tone for the duration.

Essentially you get a start tone first of all to help with IDing the start of the image, then some phasing tones to help with timing and where the lines start and end. Then comes the image itself, with the tones for black and white pixels (you'll mostly hear the white of the image coming through, and the black pixels will sound like the aforementioned ticking, unless there's a lot of black like a solid line. Then you may hear a full black tone as it draws that line. You'll get a feel for it soon enough.

There are two main catches for European listeners, DWD out of Germany and the Royal Navy here in the U.K. Sadly the skip zone seems to prevent decent reception of the U.K. stations within the U.K. itself, except for late at night sometimes (which may well be groundwave reception in the end - I’m learning a lot about propagation lately). Its signals are noisy and inconsistent. The German Deutscher Wetterdienst station broadcasting from a site near Hamburg is an easier catch. This tends to boom in, allowing very good quality images to be received, sometimes with an indoor antenna.

A rarer catch, and one I’ve never been able to reproduce again, was the NWS station in Boston. This was a combination of being there at the right time of day on the right frequency for that time. I had some great images with a bit of noise and fading - which is to be expected perhaps, given the distance. Subsequent catches have had the wire set up at different angles, which may not help, but I’ve never been able to get this station again.

And I thought that was my lot, just the three stations. And then, scrolling through a later recording made on the 9th December 2021, I heard a noisy but feint signal with the telltale sound of Wefax bursting in. You get to recognise what signals sound like after a while, and I could hear the ticking of the black pixels coming through with the white tone behind it, even through the noise floor.

Incredibly, after looking at what came through, it seemed to be a Chinese station, with the South China Sea text and the outline of Taiwan clearly visible. I’ve never been able to receive any other stations, so I’ll have to try and receive some of the other ones closer to me now. In future I may try aiming the wire broadside towards some of these stations and try to pick them up at appropriate times for each frequency.

Here's an image gallery from the 22nd Nov with two stations found between 9 and 15 MHz: 22nd Nov 2021 - part 1. Also one single image so far from the 29th - I haven't really finished decoding everything yet, because sometimes it is hard to know where to start.

A full list of all of the world’s Marine Wefax stations is available here from the NWS website: https://www.weather.gov/media/marine/rfax.pdf

I’m not really sure how many users are left who need this technology, but it is envisaged that they won’t last much longer. Years? Decades? Who knows… But for the time being they remain a fun target for HF listening and decoding, with strong signals (stronger than amateur stations, not as strong as broadcast stations) allowing for a fairly easy copy. In terms of software there’s a few choices, but for Windows the easiest is probably FLDigi, although Sorcerer is good too.

Take note of the different parameters for the signal you're decoding - DWD seem to use a different shift, for instance, of 425 hz for each colour, rather than 400. Apart from that it's a really easy mode to decode, and they make nice printouts if you've got a big plotter. If you can read Japanese you can even get news delivered to you in this format, although that is in IOC-288 mode rather than IOC-576, for the increased clarity of the Japanese characters.

Now I need to pick another type of signal to investigate. I'm tempted to look at the weather and marine navigation text broadcasts next as they're another easy decode and plentiful across the bands.