I used the Fresnel equation from Wikipedia (and "translated" it into Excel):

Rs=(($B$1*COS(C5)-$B$2*(1-($B$1/$B$2*SIN(C5))^2)^0.5)/($B$1*COS(C5)+$B$2*(1-($B$1/$B$2*SIN(C5))^2)^0.5))^2 Rp=(($B$1*(1-($B$1/$B$2*SIN(C5))^2)^0.5-$B$2*COS(C5))/($B$1*(1-($B$1/$B$2*SIN(C5))^2)^0.5+$B$2*COS(C5)))^2 

With

$B$1 = N1 $B$2 = N2 C5, C6.. = the respective angle /180 * PI() 

And then there is this very useful site that gives you information on the optical properties of water.

Now if I enter a wavelength of "1" I get a refractive index of n = 1.327. In my calculation I use this as N2 = 1.327 with N1 = 1 (for air). If you scroll down on this page there is a section called "Reflection calculator" that gives you the calculated result of light reflected on the surface of water, at the given wavelength for any given angle.

The great thing is, that when I compare these results (those in my excel sheet and those on this site) they perfectly match, for every angle, for at least 6 digits. Naturally that gave me the confidence that I did everything right.

Just recently I stumbled over a problem. For instance I can enter a wavelength of 16.039, which incidentally also has a refractive index of n = 1.327. In my excel sheet the outcome will be the same, as the refractive index is the only relevant input. But now the "Reflection calculator" gives me a very different result which does not match my excel results.

So the logic suggests the formula they use there is after all different the formula I described above and includes another parameter. I assume that parameter might be the "extinction coefficient" k, which is very small at a wavelength of 1, but quite substantial at 16.039.

Do you have any idea what the formula used on this site might actually look like?