Abstract
A technique was developed to calculate the radiosonde temperature error as a function of altitude under different environmental conditions. The environmental conditions analyzed include the surface (or cloud) temperature, the atmospheric gaseous constituents, the aerosol and thermodynamic structure of the atmosphere, the solar elevation angle, the solar albedo, the rise rate of the balloon, and the atmospheric density.
The technique was validated by comparing calculated results with data from flights of experimental radiosondes containing the NWS radiosonde's rod thermistor and three other thermistors with different radiative coatings. The experimental measurements were compared with that predicted by the modeling technique. Comparisons were made between eight flights: four at night, three daylight, and one twilight, which occurred during all seasons of the year and under various surface conditions. The comparisons showed good agreement. For the flights analyzed the temperature error at nighttime was small, below 20 km, and increased negatively above this altitude. At 30 km the error generally exceeded −1°K. During the daytime the temperature error was positive and sometimes took on its maximum value as low as 20 km. At altitudes near 30 km and above, the error often decreased due to influences of an increasing atmospheric temperature. Results from this study suggest that the radiosonde temperature error is likely to differ at different latitudes and solar elevation angles because of differing radiative fluxes to the thermistor and because of differing atmospheric temperature profiles.
