I started life as we know it as an Engineer in Winnipeg at Bristol Aerospace in 1975. I was green and full of schematics from every article of Designer's Corner from EDN since 1960, so I had some idea how to be creative. My job, design a radio to track a floating wind powered weather station so that its weather data and moving position could be sent up to GOES #1 (no GPS) and downlink via landline to Tuktoyutuk where offshore oil riggers could get forecasts on icebergs to pull out on bad weather or icebergs.
My assumption was that if my boss's boss said it was ok. It should work. He flew up there to measure Navy VLF signals and so I thought if I did it right, it should work. I had no idea.
Well it did, for as long as I worked there even with Polar Bear-proof springy whip antenna. But then 4 years later after I moved onto a better high tech job, I learnt it stopped working.
My boss successfully navigated the movement of 10 miles per day on an ice flow using the magnetometer x,y data.
I should never have assumed my design was good enough because it worked and asked, how much margin of signal loss do I need to ensure it will always work and asked for help on getting a spec for the minimum SNR. The concept of triangulation was used to detect at least 3 out of 5 parallel channels at VLF from the Megawatt transmitters around the world used to communicate with submarines. All carriers are cesium time sync'd.. e,g, Cutler Maine, Jim Creek, Hawaii. Norway etc.
I heard later the wind generator ended up in Siberia.
So if any Ruskie friends remember a 60 ft egg-beater turbine on a tripod with battery packs and electronics in the pilons, let me know.
Because as we know now, ice-flows up there, tend to flow in the opposite direction of the earth's spin whereas Jet Streams prevail from the west.
I'm not sure "weather" this fits as a wrong conclusion or an unexpected result., it was the 1st one to exist.