Abstract
A large sample of temperature observations was statistically analyzed to estimate horizontal temperature variability as a function of distance. These estimates were determined at several altitudes from the surface to 15 km and are applicable to horizontal distances up to 175 km, from a data set collected at White Sands Missile Range,. N. M.
The results were used to assess the amount of disagreement one should expect when comparing radiosonde temperature measurements with corresponding measurements derived from satellite radiometric observations. The conclusion was that horizontal temperature variations over the radiometrically observed area contribute approximately 1 K rms disagreement between such comparisons. When the separation distance between the radiosonde and satellite observation approaches 200 km, rms differences of greater than 2 K are to be expected at this location.
