An Improved CALIPSO Lidar Level 2 Cloud Data Set (Version 3)

The Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) satellite was launched in April 2006 to better understand the roles that aerosols and clouds play in climate. The primary instrument on¬board the CALIPSO payload is a two-wavelength (532 nm and 1064 nm) polarization-sensitive backscatter lidar. Since observations began in June 2006, the CALIPSO lidar has acquired a global, nearly continuous record of vertical-resolved measurements of clouds and aerosols.

The CALIPSO Lidar Level 2 data set was first released in December 2006 and provided information primarily on the layer structure of clouds and aerosols. A 2nd version was released in December 2007 that expanded on the original data release and included a new set profile products as well as optical properties such as optical depth.

The third Level 2 data release is expected in Spring 2010 (Version 3) and includes several significant advancements over the previous data sets. One feature is the use of a new five-dimensional (5D) PDF-based Cloud and Aerosol Discrimination (CAD) algorithm instead of a 3D approach used in Version 2. This algorithm enhances separation between aerosols/clouds and, thus, provides much improved classification of very dense dust layers which had been misclassified as clouds in earlier data sets. Another feature is the addition of a new algorithm for cloud ice-water phase that permits the identification of horizontal oriented ice (HOI) crystals. Preliminary analysis indicates that the occurrence of clouds with HOI crystals in polar regions is far more prevalent than previously believed. A revision was also made to the algorithm that misreported the frequency of marine trade cumulus at 5 km resolution. Among other changes, Version 3 adds uncertainty estimates for all retrieved parameters and provides restructured aerosol and cloud profile products.

This paper will describe improvements to the Version 3 Level 2 data set and will provide sample comparisons between the Version 2 and Version 3 data sets to highlight the improvements.