I wish I had a solid source on this. The best I can find is a 6-year-old Reddit comment:
This particular jump with the BT-7 cavalry tank was performed by a tank test pilot Evgeniy Kulchitsky, he mentions in his memoirs that it was a very challenging test to perform (15m height, 42m length of the jump), requiring challenging practice and long preparation. You had to adjust the rotational speed of the tracks mid-air to not outright lose them, assume a (relatively) safe position and then continue to pilot the tank after landing to get it to the shore.
The reason for this testing was to make sure that similar jumps were a viable tactical option (fording rivers at speed, for example) and that both the crew and the tank could manage it and remain effective.
Would it be unsurvivable? The force of the impact on the water would be great, but on the other hand tanks are heavy, and the only thing that matters for the driver (pilot?) is acceleration. Doing absolutely no math the main thing I'd be worried about is the impact with the lake bed afterwards breaking the treads. If all goes well and the engine has some kind of supply of air, you just drive it out again.
It depends on how watertight the tank is and depth of water I suppose, but given the cramped area for drivers, even with a helmet there's probably a concussion and possibly broken bones. You are going to hit the top and front
Bring some padding and brace for impact, I guess? Clearly someone did it (unless this is doctored).
It's my impression that modern tanks are built to operate under an appreciable depth of water, which makes sense since airtightness is also watertightness. I have no idea when and where this was taken, and what the case was then. It would suck if the bruised and shaken driver had to swim out.