Well, let’s see. Stress of combat, low-resolution thermal imagers, seeing the vehicle shoot (Often in what appears to be your direction) at about 2-3 kilometers distance…
That’s why it happened in Desert Storm, and that’s why it’s happening now. If you want to be somewhere safe and controlled, stay off the battlefield. If you want a place where mistakes don’t happen, you’re just shit out of luck.
To be honest, after seeing the quality of the M1A1 thermal imager output, I’m surprised friendly-fire happens as infrequently as it does. You try to identify a T-64 from a M1A1 or Challenger 2 when it’s a bright green blob at 3 klicks.
Well, that’d be true, if they were firing in the same direction. Or appeared to be, anyway. I understand that at least one M1A1 was hit in Desert Storm by a friendly M1A1 because a T-64 nailed it in the rear with a HEAT round, and another tank saw the flash on the rear of the tank and thought it was firing at HIM (The flash from the HEAT round looked like a muzzle flash).
As for radar vs. planes, that’s tricky too. IFF can fail, and radars that can distinguish friendly from enemy aircraft can’t always do so.
In any case, lessons have been learned, and more and more effort has been going into improving the problems experienced in Desert Storm. There is more use of thermal identification markings, to make vehicles easier to ID in thermal imagers. Improved communication and GPS systems make it easier to determine who is where. The newer tanks (M1A2, etc) include thermal imagers with improved clarity and magnification, making ID much easier (As well as seperate systems for the TC to look through). AWACS and J-STARS information is easier to relay and quicker to update to the troops in the field. Reconisance is pinning down exactly where enemy groups are.
Desert Storm lasted about 100 hours, right? This has lasted a few times as long, so far. How do the casualty numbers from friendly fire stack up between the two?