The conservative shape of temperature profiles in the lower tropospher
e above the Antarctic ice sheet makes the surface temperature, t(s), a
n effective independent variable for estimating the height of the 500
hPa surface above automatic weather stations (AWS), In place of regres
sion equations for the thinning of the surface/500 hPa layer (produced
by negative layer mean temperatures, (t) over bar) as a dependent var
iable, an alternative approach examined here uses the layer mean tempe
rature itself as a dependent variable, It is shown that this eliminate
s systematic differences between the regression parameters for radioso
nde stations at different elevations, and should validate a single set
of parameters for a wide range of AWS sites.