Project: The VOLMET Audio Gallery

Utility stations on HF are rather interesting to me and some readers may have seen my recent series on HF radiofax which is nearing its conclusion pending some more monitoring. But there are other types of utility stations out there, some of which are interesting to me as well.

Seeing that HF utility stations could be at risk of “fading away” in the not-too-distant future, I decided to get my act together and start on another project while I still can.

What is a VOLMET?

VOLMET, short for vol (flight) and meteo (weather) in French, is an broadcast of meteorological information for aircraft in flight. Aimed at long-haul flights traversing vast oceanic expanses, it does not require the use of satellite SATCOMS facilities, instead using ionospheric propagation of HF radio signals to reach over the line-of-sight and is thus low-cost. It is sent at scheduled intervals (often five minute slots repeated half-hourly) in the aeronautical bands on HF, using upper-sideband (USB) modulation. Many of them use automated text-to-speech synthesis synthesis to generate the audio reports, with each VOLMET station sounding slightly different. Some systems include STR-SpeechTech Starcaster, Frequentis smartATIS/VOLMET, Microstep IMS4 ATIS/VOLMET and Terma ATIS/VOLMET Flex to name a few. Information provided usually relates to the weather conditions at major airports within the nearby region, with some longer VOLMET systems broadcasting half-hour blocks or even continually.

VOLMET transmissions and their time schedules have been compiled and kept up to date by William Hepburn of DX Info Centre. This resource has been extremely useful for me, as much of the information online can be rather dated and unreliable, whereas this list was last updated on 14th January 2019 (at the point I am writing this article).

Received International VOLMET Recordings

In order to build this VOLMET audio gallery, I have used a wide range of publicly available KiwiSDR receivers and kiwirecorder.py to record audio from those receivers and isolate recording of the relevant VOLMET transmissions. Due to the use of a Python script for recording, it became possible to survey multiple frequencies on multiple SDRs simultaneously, hence this project was completed (for the most part) within around a week with most successful recordings occurring within the first few days.

The location of the receivers was chosen to be as close as possible to the transmitter to hopefully improve audio quality, however, HF propagation effects (such as noise, fading, selective fading and interference) persist. Several attempts were made at receiving most of the VOLMET stations, with the best result shown (with the reception date, time and credit to the KiwiSDR used). Frequencies used were based on the list by William Hepburn linked earlier, with all frequencies monitored for a good quality recording or a minimum of 24 hours to ensure that I had an opportunity to collect all VOLMETs possible given my resources.

Recordings were made with USB demodulator, with a passband of 0Hz to 2700Hz and may thus sound a little more “noisy” and “bassy” than an average receiver which may have a 300-2700Hz passband. I chose to do it this way as some VOLMET transmissions do have significant bass components which affect the timbre of the voice.

All recordings are provided for documentary purposes, to illustrate the nature of the VOLMET transmissions, the speech synthesis/voices in use, the content transmitted, the format of the transmission and as proof of its transmission, propagation and reception at the date, time and locations mentioned.

This first section covers International VOLMETs, which are those that broadcast in English for the use of international flights on major world air routes. A spectrogram image (0-3000Hz) generated by Audacity and the text that begins/ends the transmission are shown for each station with a link to the audio file, details of reception and comments.

Australian VOLMET

“All stations, Australian VOLMET.” / “Australian VOLMET out.”

Australian VOLMET Audio – received via VK2DDS (Newcastle, NSW, Australia) on 20/02/2019 at 0900UTC on 11387kHz. As my “local” VOLMET station, it’s one I am familiar with. The voice is a female voice, but there is a sub-harmonic which makes it sound “gloomy” and as if the receiver is de-tuned.

Trenton Military (Canada) VOLMET

“Aircraft Stations. Aircraft Stations. This is Trenton Military. Trenton Military.” / “Break. Trenton Military out.”

Trenton Military (Canada) VOLMET Audio – received via VE3HOA (Ottawa, Canada) on 23/02/2019 at 0610UTC on 6754kHz. This one sounds almost as if a male voice with purpose is making an announcement – I wonder if this system is based on recorded speech samples rather than synthesis as it is very natural-sounding and easy to understand. This is a rather long 50-minute VOLMET.

Gander (Canada) VOLMET

“This is Gander Radio.”

Gander (Canada) VOLMET Audio – received via SCPDNY (Hampton Bays, New York, USA) on 20/02/2019 at 1050UTC on 3485kHz. Another male voice, with a recorded message insert of “For thunderstorm METAR contact Gander radio.” A 10-minute block rather than the more usual 5-minute block, always repeats the location name and fairly easy to copy. The hum in the background appears to be interference with the SDR and may not be from the transmitter.

Beijing (China) VOLMET

“Beijing Meteorological Information for Aircraft in Flight.”

Beijing (China) VOLMET Audio – received via North East Asian Broadcasting Institute (Seoul, South Korea) on 21/02/2019 at 1245UTC on 5673kHz. Due to a lack of KiwiSDRs around the China area, there wasn’t many chances to get a better copy. This system starts with the full meaning of VOLMET, with a very “rhythmic” female voice synthesis system that seems to be assembling blocks of speech. The broadcast seems to end in a few “glitchy” noises as if the voice synthesis system is either glitching or skipping through some empty entries.

Guangzhou (China) VOLMET

“Guang[zhou] Meteorological Information for Aircraft in Flight.”

Guangzhou (China) VOLMET Audio – received via North East Asian Broadcasting Institute (Seoul, South Korea) on 20/02/2019 at 0230UTC on 13285kHz. The name of Guangzhou seems to be clipped at the beginning of the transmission, but the system is identical to that of Beijing, sharing the same frequency. Unfortunately, it seems that time-scheduling is not their strong point, occasionally overlapping.

Hong Kong VOLMET

“Aaaaaaaaaal stations. This is Hong Kong VOLMET.” / “Hong Kong VOLMET out.”

Hong Kong VOLMET Audio – received via VR3BG (Hong Kong) on 19/02/2019 at 1015UTC on 8828kHz. This one has a very recognizable male voice with a very exaggerated “all” at the beginning of the broadcast and proper radio pronunciation of four (fower). The intonation is a little unsmooth and again, it has a rather rhythmic pace to the report.

Mumbai (India) VOLMET

“Mumbay VOLMET broadcast. Mumbay VOLMET broadcast.” / “All stations, this is Mumbay VOLMET broadcast.”

Mumbai (India) VOLMET Audio – received via VUHams (Kerala, India) on 19/02/2019 at 1025UTC on 11387kHz. This is a male voice which sounds like it’s in a hurry, perhaps too much of a hurry as it can be slightly difficult to understand. It seems nobody programmed any exceptions, as Mumbai is pronounced “mum-bay” by the system, which seems at odds with what I know. Chennai is pronounced “chen-ay” by the system … so maybe it was recently upgraded. The other Indian system (Karachi) was not heard, but this VOLMET covers Karachi, so maybe Karachi is no longer?

Shannon (Ireland) VOLMET

“This is Shannon VOLMET, Shannon VOLMET.”

Shannon (Ireland) VOLMET Audio – received via EmeraldSDR (Carlow, Ireland) on 20/02/2019 at 0900UTC on 3413kHz. This one is voiced by an electronic, sultry female voice … talking perhaps a little slowly with a slight second “ticking” in the background. The “flight level” words seem to have been re-recorded, as another system that assembles speech snippets. At the moment, it seems to be on a half-hour slot repeating the same pattern at odds to the previous one hour loop.

Tokyo (Japan) VOLMET

“All Stations. This is Tokyo.” / “Tokyo Out.”

Tokyo (Japan) VOLMET Audio – received via ZL/KF6VO (Takeradio) (Aichi, Japan) on 21/02/2019 at 0140UTC on 8828kHz. Voiced by a female voice with a recognisably Asian accent, this is rather easy to understand and clear, although the pronunciation of “no sig” sounds a bit strange to me.

Military One (UK) VOLMET

“Tid(?) Military One Timeslot One Information Broadcast”

Military One (UK) VOLMET Audio – received via G0LUJ-1 (Lytham, England) on 20/02/2019 at 0730UTC on 5450kHz. Voiced by a female voice synthesis with a noticeably English accent, it always starts timeslots with a noise that sounds like “tid”? No idea what that is supposed to be. Some snippets seem to be voiced at different tempos resulting in a speech which seems to go from rapid to slow all of a sudden … and back again.

Auckland (New Zealand) VOLMET

“All stations, this is Auckland VOLMET.” / “Auckland VOLMET out.”

Auckland (New Zealand) VOLMET Audio – received via Marahau (Tasman District, New Zealand) on 21/02/2019 at 0030UTC on 8828kHz. If there was an award to be given for the most monotonic voice synthesis system for VOLMETs, I’d have to award it to New Zealand’s male voice that is almost completely devoid of intonation. It sounds like someone that might have had too much to drink and is way too tired reading out some numbers to a metronome – consistent, but boring and very bassy.

Singapore VOLMET

“All stations, this is Singapore Radio, time …” / “Singapore Radio out.”

Singapore VOLMET Audio – received via SWLOI33 (Jakarta, Indonesia) on 21/02/2019 at 0250UTC on 11387kHz. This is voiced by a young-sounding male voice at a moderate pace which is fairly easy to understand but there was some fading in the reception.

Bangkok (Thailand) VOLMET

“All stations, this is Bangkok VOLMET, at …” / “Bangkok VOLMET out.”

Bangkok (Thailand) VOLMET – received via VR3BG (Hong Kong) on 20/02/2019 at 0940UTC on 6676kHz. This particular transmission was affected by frequency-selective fading, but at most places I could hear the transmission, either it sounded like it was “clipping” from being played too loudly or lost in the noise. The male voice is understandable, but the transmission has a harsh tone.

Russian VOLMETs

Although I was made aware of the existence of “national” VOLMETs, I have never heard any Russian VOLMETs until I embarked on this project. While it seems their international VOLMETs are out of service, their national VOLMETs still seem to be operating but perhaps more like an ATIS with recorded rather than synthesized audio in some cases. Unfortunately, as I don’t understand Russian and the signals are a little weak, I’ll leave you with the recordings instead. In some cases there are several recordings to illustrate some behavioural traits.

Rostov (RLAP) VOLMET

Rostov (RLAP) VOLMET Audio 1 – received via RA3TKH, R3TIO (Afonino, nr. Nizhny Novgorod, Russia) on 21/02/2019 at 1055UTC on 11297kHz. This one was prefixed by a strange frequency “chirp”. The voice sounds a little mechanical – perhaps synthesized.

Rostov (RLAP) VOLMET Audio 2 – received via RA3TKH, R3TIO (Afonino, nr. Nizhny Novgorod, Russia) on 22/02/2019 at 1025UTC on 6617kHz. Correlating with the above recording, perhaps the three tones at the beginning are used by some selective calling system but incorrectly transmitted on 11297kHz. The voice follows the same timing patterns, suggesting it is an automated system in use at Rostov. The voice sounds female but lacking in variation.

Samara (RQCI) VOLMET

Samara (RQCI) VOLMET Audio – received via RA3TKH, R3TIO (Afonino, nr. Nizhny Novgorod, Russia) on 22/02/2019 at 0945UTC on 6693kHz. This sounds like a recorded female voice with rather natural sounding pacing and intonation.

Sivkar (UBB-2) VOLMET

Sivkar (UBB-2) VOLMET Audio – received via RA3TKH, R3TIO (Afonino, nr. Nizhny Novgorod, Russia) on 22/02/2019 at 1200UTC on 8888kHz. This sounds like a recorded male voice.

Peterburg (UHD) VOLMET

Peterburg (UHD) VOLMET Audio 1 – received via RA3TKH, R3TIO (Afonino, nr. Nizhny Novgorod, Russia) on 22/02/2019 at0535UTC on 6617kHz.

Peterburg (UHD) VOLMET Audio 2 – received via RA3TKH, R3TIO (Afonino, nr. Nizhny Novgorod, Russia) on 22/02/2019 at 0705UTC on 6617kHz. This transmitter seems to be rather hummy, with visible constant-frequency tones going right through the transmission. It looks that it could actually be a recorded female voice rather than synthesis. I know I’ve heard coughing and sneezing before … but I’m not sure if it was Peterburg.

Novosibirsk (UNNN) VOLMET

Novosibirsk (UNNN) VOLMET Audio – received via RA3TKH, R3TIO (Afonino, nr. Nizhny Novgorod, Russia) on 21/02/2019 at 1140UTC on 11318kHz. Unfortunately Novosibirsk was hard to hear, but this does serve as proof of its activity nonetheless, a Russian female voice.

Unheard VOLMETs

Despite all the planning and a concerted monitoring effort involving at least 24 hours monitoring each frequency where nothing was heard, I was not able to receive any copy from the following stations on the DX Info Centre list:

• Antananarivo
• Brazzaville
• Carrasco
• Comodoro
• Cordoba
• Cyprus
• Ezeiza
• Honolulu
• Karachi
• Khabarovsk
• Kiev
• Kolkata
• La Paz
• Luanda
• Moscow
• New York
• Resistencia
• Tashkent
• Viper (Mount Pleasant)

The greyed stations are those which are noted to be decommissioned or out of service on that list, but that still leaves a large portion of mostly South American and South African VOLMETs not heard (a total of ten stations). I suspect the reasoning is partly due to the sparse coverage of reliable KiwiSDRs with good signals in those areas – while I have not heard them despite efforts, it cannot be concluded that they are off the air.

The only one I would have good reason to believe is off the air is Kolkata, as the Mumbai system sounds “new” and it does include information for Kolkata – suggesting that the Kolkata system may have been retired. Multiple attempts from different locations were not able to hear it, but again, it’s hard to conclude things as a definite although I suspect pilots with access to NOTAMs and flight charts would know which NAVAIDs/etc are in service at any given time.

Conclusion

The publicly available KiwiSDRs have been invaluable for these kinds of “virtual DXes”, allowing me to achieve a personal goal of mine to hear as many VOLMET systems from around the world as possible. In the process, I was able to document, with audio samples, these systems before they disappear. I thank the operators of all of the KiwiSDRs used (and credited above) for opening their receivers for public use and making this possible.

Speaking of which, between the two projects so far, I’ve spent almost 220GB of LTE data over the past three or so months in monitoring all the frequencies, sometimes extensively. In a country where we pay for our data allowance, this hasn’t exactly been a cheap exercise. But it’s been worthwhile, from my perspective – bringing my gear along and taking a holiday to all these locations would cost me a lot more and perhaps … not even be possible.

Unfortunately, while HF is still useful, aviators often dislike the HF radio for communications as signals can be noisy, difficult to read making it easy to miss key information. Increasingly, commercial operators have and use SATCOM technology with new D-VOLMET systems to distribute the VOLMET text nearly instantaneously, unambiguously and free of errors. As a result, as VOLMET stations fall into disrepair, it seems unlikely they will be maintained for too much longer into the future.

Somehow, I have a feeling that New York VOLMET and Honolulu VOLMET may never return, as their return dates seem to forever postpone. Likewise, the other VOLMETs listed as “Out of Service” may also never return. I guess I’m too late for those.