I seem to remember it backs by 30 degrees and speed decreases by 50% at the surface (10 degrees and 30% over water due to less friction) but as usual happy to be corrected.
It's to do with the surface friction reducing the speed and therefore reducing the Coriolis effect I think, so doubtful that it could be any different at night.
EDIT: Just found this which explains it:
http://ww2010.atmos.uiuc.edu/(Gh)/gu...r/fw/fric.rxml