...
| band | central wavelength | extinction (mag per airmass) | seeing degradation compared to r band at zenith | |
|---|---|---|---|---|
| u | 350 nm | 0.40*a | (1.1)*a0.6 | |
| g | 450 nm | 0.18*a | (1.07)*a0.6 | |
| r | 650 nm | 0.10*a | (1.0)*a0.6 | |
| i | 750 nm | 0.08*a | (0.97)*a0.6 | |
| z | 850 nm | 0.05*a | (0.94)*a0.6 | |
| y | 1000 nm | 0.04*a | (0.91)*a0.6 |
...
plot of seeing degradation vs. airmass, and polynomial fit:
...
Decent approximation to seeing degradation vs. airmass a is
S=0.35+0.72a^2-0.07a
Signal to Noise degradation vs. airmass:
SNR scales as source flux in the numerator and (for unresolved objects) as seeing in the denominator. Flux at an airmass "a" is reduced by a factor f(a)=10^(x*(a-1)/2.5) where x is the extinction coefficient listed in the table above. So SNR vs. airmass at fixed exposure time for unresolved point sources scales as SNR(a)~10^(x*(a-1)/2.5)/(0.35+0.72a^2-0.07a)
Some references
...