Is land surface processes representation a possible weak link in current Regional Climate Models?