Abstract
From analytical and empirical calculations with upper-air pressure, temperature, density and height, it is concluded that: (a) mean density can be estimated closely from mean pressure and temperature, (b) the standard deviation of density can be estimated from mean pressure and temperature and the standard deviation of temperature, (c) the correlation of temperature at two pressure surfaces is a very good estimate of the correlation of density at these two surfaces, but a similar relation does not exist at two constant height surfaces, (d) a linear relation exists between the standard deviation of height of a pressure surface and the standard deviation of pressure at a height equal to the mean height of the pressure surface, (e) there is also some relationship between the correlation of the heights of two pressure surfaces and the correlation of pressures at the monthly mean heights of the pressure surfaces.
These relations provide some short-cut procedures for estimating the statistics from other available statistics. While only three stations and the four midseason months were used to check the relationships, the results seem to substantiate the conclusions.
