Tough to say form the pictures, but here goes with some typical aspects:
It being very cold will not have helped. If it was just freezing and you had some speed the extra cooling due to wind chill takes it to -23 degC or so. see this link: http://www.eol.ucar.edu/homes/rilling/wc_table.html
Not sure if that is enough to cause trouble though, need to know more about the alloy and it's treatment. But I doubt the cold alone could cause a complete brittle failure.
Also because I see some rough surface with shadows as if there are "peaks" in the far left in the second picture. This could be gross deformations from over stressing. Such deformations are commonly seen on fractures that have cracked over a large percentage of the surface area when the last bit is just torn apart. The first picture is more difficult to judge, but there appears to be a colour difference with the far right of it being a lighter shade (this would be corresponding to the "peaks"area in the other picture?)
Now if the larger remaning area of the fracture is indeed not just darker but also smoother, dull in appearance or even with water/dirt or corrosion products on it that would make sense. It would indicate that it started with a small crack some time ago which grew over time, until it was so far the remaining ligament gave up under your braking action, the cold not helping at that fatal event.
The critical elements to observe being that the long time frame of the growing crack meant water/dirt and corrosion had access to the already cracked surface, and that the original crack had to strat somewhere. Otherwise this to me would be a plausible scenario, though it takes a lot of extrapolating from hints in the two pictures.
Do you see any of the features supporting this scenario clearly, or not at all?
Common sense was never common
28h silver rear Suntour hub
32h Ritchey champagne WCS front hub