The rate of sea-level rise has increased in the period from 1993 to the
present (Figure 1), largely due to the growing contribution of ice loss from
Greenland (Box 1) and Antarctica. However, models of the behaviour of
these polar ice sheets are still in their infancy, so projections of sea-level
rise to 2100 based on such "process models" are highly uncertain. An
alternative approach is to base projections on the observed relationship
between global average temperature rise and sea-level rise over the
past 120 years, assuming that this observed relationship will continue
into the future. New estimates based on this approach suggest a sealevel
rise of around a metre or more by 2100

The above recently doubling of the estimate to around 1 meter by 2100 is based on a linear extrapolation from previous data. Hansen argued a couple of years ago that the linearity assumption should be questioned, and that if rapid collapse of the arctic and antarctic glaciers were to occur, quick rises of sea level beyond the 1 meter range could result.
http://www.newscientist.com/article/mg19526141.600-huge-sea-level-rises-are-coming--unless-we-act-no w.html?page=1

Many civic projects related to flood control are financed and designed with 100 year time frames, so large uncertainty in this area is difficult from both the technical and the political viewpoints...