Abstract
Discussed is the errors of height tendency and vertical velocity caused by assumptions of fictitious boundary conditions, i.e. the height tendency or the vertical velocity vanishes on the side-boundary or at the lower boundary respectively. Generally speaking, errors penetrate farther into the domains with increase of the scales of errors on and at boundaries. The distributions of errors in the domains are given for various type of boundary errors. Comparison between the exact solution and the result of prediction by the barotropic model is made for a disturbance of Haurwitz's type.
