摘要: A fundamental activity of numerous numerical weather predictions centers is objective analysis of the global atmosphere. Typically, the objective analysis is a product of a four-dimensinal global data assimilation system. Such a s... 展开 A fundamental activity of numerous numerical weather predictions centers is objective analysis of the global atmosphere. Typically, the objective analysis is a product of a four-dimensinal global data assimilation system. Such a system includes objective analysis, initialization, and global forecast model components. It is common for the objective analysis to be performed on constant pressure surfaces, while the forecast model operates on the standard sigma coordinate or a close variant. This necessitates vertical interpolations between pressure surfaces and sigma-coordinate surfaces before and after the forecast. This study shows that the choice of moisture variable and interpolation method in the vertical interpolation can be crucial to preserving a reasonable moisture distribution in humidity assimilations. The use of relative humidity, dewpoint (T sub d), and specific humidity are tested in four two-point interpolation methods including linear, logarithmic, exponential, and power law. The standard use of logarithmic interpolation, is shown to be a suitable choice when vertically interpolating RH or T sub d, but a singularly poor choice when interpolating Q. 收起