In this study, we investigate the global distributions of TLC by developing a java- based CALIPSO INteractive Digitization Interface (CINDI) to identify TLC from CALIPSO Level 1 backscatter data. We then apply CINDI to all CALIPSO profiles in January and July of 2009 to retrieve physical parameters (e.g., height, thickness, average ice water content, etc.) from each of the TLC that is identified. This level of detail cannot be realized by the current CALIPSO Level-2 Vertical Feature Mask (VFM) algorithm. The analysis is extended to 36S and 36N to include the mid-latitude TLC.
In a preliminary comparison between the semi-automatously generated CINDI dataset and the Version 4 VFM dataset, we found the majority of sampled TLC points were classified as clear air in the latter dataset, approximately 40% were classified as cirrus clouds, and the majority of the remaining were identified as sulfate or dust aerosols. While the disparity between the two classification schemes implies a nebulous ground truth, analysis of TLC morphology in the CINDI dataset reveals a significant difference in the TLC location and height between January and July. The correspondence between collocated Aura Microwave Limb Sounder water vapor retrievals strongly suggests a seasonal difference in global TLC distributions and supports TLC as an effective dehydration mechanism.