Pavement Markings on Bridges, or Lack Thereof

[vtk]

So, why is it that pavement markings seem to wear away on bridges faster than on solid-ground pavement?  This can lead to dangerous confusion, particularly where the bridge is curved and/or the lanes aren't strictly parallel. 

One example is US 62 EB over I-270, Gahanna, OH.  The 3 lanes entering the bridge split into 2+2 exiting the bridge, as the whole thing simultaneously curves left.  While the stripes were adequate the last few times I drove this, they had been problematic in the past. 

Another example is OH 104 EB(NB) over the Scioto River on the south side of Columbus.  While the bridge is essentially straight, the lanes shift leftwards.  The stripes are badly worn away, making it difficult to properly track your lane across the bridge. 

There are of course many other examples of pavement markings wearing away quickly on bridges, but these two stick in my mind because of the challenging roadway geometry made worse by inadequate markings.  I'll ask again, why does this happen?

[Alps]

Concrete vs. asphalt could be part of it, but I bet a large factor is the air under the bridge. Road surfaces tend to stay warmer than the surrounding ground (darn hot in summer, even), whereas bridges reflect the ambient temperature. My thought is that the stripe particles don't adhere as well the colder things get.

[vdeane]

Possibly icing?  Also I've noticed that pavement marking seem to fade faster on concrete than asphalt, and in contrast to roads, more bridges seem to be concrete than asphalt.

[SteveG1988]

I would say, it might be due to the chemicals used to keep the bridges from icing up. It might be a little more....solvent like compared to the stuff used on normal roads that stay warmer than the surrounding area. Also,it may be due to the deck flexing more due to heating and cooling cycles, the paint may not be flexible enough to handle it.