164
result with more iterations (three times more). If these constant values are too important or
small, the solution explodes.
Adjustment of the eddy viscosity :
To adjust this parameter, we have some experimental statements such as the major
whirlwinds and the order of the extent of the velocities. In case of the test problem, we
referred to the results of Liege. We checked that if we got good results for a A value included
between 10-2 and 10- 3 . If A is smaller, the velocity modulus becomes infinite. Physically, we
are in the case where the dissipation does not compensate any more the energy brought by the
wind. When A is higher than 10- 2 the dissipation is too important and the speed field
obtained is too weak.
II. 3.3 Some results in the case of real spots:
Actually, we worked on the sites of the Thau's pond and the Bonifacio's nozzles. The
first site does not present real adjustment difficulties as we have an important number of
measures. This is not the same case for the second spot. We give in figure I and 2 the
circulation and the elevation for a north-east wind (10 mls) in the case ofthe Thau's pond.bb