I appreciate you doing the math on energy and luminous flux density (seriously), but the concern is beyond distraction.
I was hit at 2500 feet near Portland a few years back and it illuminated my left wing first (high wing), causing me to close one eye but the night vision in the open eye was seriously hampered for >5 minutes and even the eye that I quickly closed was modestly affected for a short time. I don't recall the height above terrain that 2500 MSL represented there, but I have to think it was better than 2000' distance, and that was over a populated (read: lit) area. In a darker environment, a laser's effect on night vision would be more pronounced.
That seems like a very drastic effect from diffuse exposure to laser light at long distance.
Partial adaptation from full sunlight to darkness takes about 5 minutes [1], and the diffuse laser light you were exposed to is orders of magnitude weaker than sunlight.
I would be curious to see some more experimental research in this area. We know a lot about energy thresholds for eye damage, but we don't have as much precise data for adaptation thresholds.
Sure, but the pupils would be much more open than they would be in sunlight.
That is the crux of the problem. In sunlight, your pupils are so constricted that negligible light hits the retina. But at night, they're fully dilated and a sudden burst of laser can cause harm.
I was about 40 when this incident occurred; eyes in generally good health (astigmatic, stable prescription, well-corrected, no color deficiency, nothing else significant)
You can just do the math yourself to prove that most of the panicked news we hear is bullshit.
For any given laser, we know the divergence (the given figures are usually smaller than the actual divergence). Let's look at the wicked lasers arctic, a popular (but low quality) 445nm blue laser with a power output of up to 2000mW. This is one of the more powerful consumer handheld lasers out there.
Wicked claims a divergence of 1.5mRad. I think that's optimistic. It's probably really above 2 mRad.
Let's say you get within 300 meters of a plane. That's pretty tricky; you'd be lucky to do that even near an airport (and the closer you are to an airport, the easier it is to get arrested.
At 300 meters, the beam is (very optimistically) about 450mm. Because of the multimode properties of the 445nm diode, even with this very optimistic spec, it's probably more like a 1m x .45m ellipse. But let's just assume a circle.
pi.45.45 = .63 meters squared.
That's a power density of 2W/.63m^2, or about .00037W/cm^2.
According to the IEC and SPIE, that's a low enough power density that at visible wavelengths, you could be exposed to it for long periods with no eye damage. [1][2]
This is assuming ideal conditions, including a perfect beam, no atmospheric interference, and no reflection/filtering on the aircraft.
There is no physical danger to the pilot; the risk is that the pilot gets distracted. This risk is not substantial, as it's very difficult to keep a laser trained at a pilot long enough to be dangerously distracting, and you have to be very close for the laser light to even be noticeable.
By the way; There are a few videos floating around of lasers shining into cockpits. Most of these look really bad, because A)they filmed with the intention of making lasers look dangerous and B)lasers saturate camera CCDs very easily, so they can look incredibly bright on video even if our eyes handle them pretty well.
[1] http://en.wikipedia.org/wiki/File:IEC60825_MPE_W_s.png [2] http://spie.org/Documents/Publications/00%20STEP%20Module%20...