Abstract
An analysis of tropical radiosonde temperature measurements made during the Line Island Experiment suggests that conventional radiosonde preflight procedures are inadequate in a remote tropical environment. Temperatures computed from conventional and modified baseline techniques are compared at five pressure surfaces, 1000, 800, 600, 400 and 200 mb. Temperatures obtained from the two baseline techniques showed an average deviation at 1000 mb of 0.96C for 62 soundings. These comparisons indicate that a careful examination of radiosonde calibration techniques is needed before large investments are made in future global experiments.
