Changeset 559 for trunk/MESOSCALE/LMD_MM_MARS/SIMU/DEF/TESTCRATER/ze_hill

MESOSCALE: debugged idealized crater case to solve problems found by JBM. the problems appeared because crater case forced the model to extrapolate at initialization ; hence the crater was not actually here. solved the problem by setting the input sounding at bottom of crater. at the same time introduced coriolis terms. it is now important to use the pert_coriolis keyword at T to avoid impacting background winds. updated all setting files for TESTCRATER, which now works fine.