Abstract
Algorithms, based on Magnus's form equations, are described that minimize the difference between several relationships between temperature and water vapor pressure at saturation that are commonly used in archiving data. The work was initiated in connection with the development of a unified upper-air dataset that will use measurements gathered from the late 1930s to the present and archived in several data centers. The conversion of field measurements to archived humidity values within the databases that are being used have not been consistent and in some cases are unknown. A goal of this work was to develop a uniform and accurate method to convert these data to various humidity variables without regard to the equations used in archiving the original data. Archived temperature values are recorded to 0.1°C. This precision creates a temperature dependent range in uncertainty in saturation vapor pressure. A procedure was developed to take this into account when the error minimizing equations were derived.
