Predicting Storm Surge

Storm surge is often the most devastating part of a hurricane.
Mathematical models used to predict surge must incorporate the
effects of winds, atmospheric pressure, tides, waves and river
flows, as well as the geometry and topography of the coastal ocean
and the adjacent floodplain. Equations from fluid dynamics describe
the movement of water, but most often such huge systems of
equations need to be solved by numerical analysis in order to
better forecast where potential flooding will occur. Much of the
detailed geometry and topography on or near a coast require very
fine precision to model, while other regions such as large open
expanses of deep water can typically be solved with much coarser
resolution. So using one scale throughout either has too much data
to be feasible or is not very predictive in the area of greatest
concern, the coastal floodplain. Researchers solve this problem by
using an unstructured grid size that adapts to the relevant regions
and allows for coupling of the information from the ocean to the
coast and inland. The model was very accurate in tests of
historical storms in southern Louisiana and is being used to design
better and safer levees in the region and to evaluate the safety of
all coastal regions. For More Information: A New Generation
Hurricane Storm Surge Model for Southern Louisiana, by Joannes
Westerink et al.