Fun facts:
Water ice doesn't absorb a lot of light in the visible; pure water ice is almost entirely transparent, although with slightly stronger absorption in the red (hence why very large amounts of mostly pure ice, such as glacial ice, has blue shading). What makes water ice much more visible comes from light scattering off impurities, air bubbles, or cracks (e.g. why ice cubes are sometimes clear and sometimes white, despite both being water ice).
It gets more interesting in the near-ultraviolet; between 200 and 400 nm, water ice is so transparent that we literally cannot measure the amount of light it absorbs. Even using deep boreholes in ultra-pure Antarctic ice, any loss of transparency is apparently dominated by scattering off impurities at levels of parts-per-billion. At the wavelength where absorption is lowest (~390 nm), our best estimate is that the mean free path of a photon in pure ice is over a kilometer; this makes water ice in the near-UV the most transparent solid known.
From Warren and Brandt (2008), and in comparison to an earlier compilation by Warren (1984).
The imaginary index of refraction on the y-axis is a measure of how strongly a material absorbs light. The higher up on the graph, the stronger the absorption (and note that this is a logarithmic scale).
In the near-UV (200-400 nm), water ice absorbs light so weakly that they just omit the data entirely here.
"We leave a gap in the compilation between 200 and 390 nm (Figure 1). In this domain, mim < 2 x 10-11 . For most applications this extremely low value of mim, indicating an absorption length ka-1= 1.5 km at wavelength = 390 nm, will be indistinguishable from zero."
