A scheme for correcting Arcasonde-1A temperatures, taking into account the thermal coupling between the bead thermistor and its Mylar-film mount, is examined by a combination of theoretical and experimental considerations using recently available information on heat-transfer processes in rarefied airflow. Total applicable corrections estimated from the few (six) temperature profiles considered were found to be about 1.2K at 30 km, 1.8K at 40 km, 4.4K at 50 km, rising sharply to 12.OK at 60 km (only two flights) and 22.7K at 63 km (one flight). Total uncertainties in the measured temperatures after correction were estimated to be about ±5.5K at 60 km, ±2.9K at 60 km, ±2,9K at 50 km, ±1.4K at 40 km, and 1.0K at 30 km and lower.
It was observed that temperature measurements made during the winter months required larger corrections than those made in summer months, especially at the higher altitudes.
The corrected thermistor-measured temperatures, in most cases, agree well with temperatures from grenade and pitot tube measurements even up to the 60-km level. Although aerodynamic heating and solar radiation are major problems, it appears feasible to use the flat-plate, thin-film sensors of the Arcasondc-IA type, or the loop-film sensors of the Datasonde type (which are replacing the Arcasondes) to obtain realistic temperature profiles up to mesospheric levels.