Cloud Statistics from Calipso Lidar Data for the Performance Assessment of a Methane Space Lidar

Kurzfassung

In this thesis a performance assessment for the future German-French climate monitoring initiative, Methane Remote Sensing Lidar Mission (MERLIN), proposed by DLR and CNES in 2010 was undertaken. A general space lidar performance issue is the obstruction by optically dense clouds. For this purpose cloud free statistics, the global cloud top flatness and global cloud top distributions were derived from the Cloud-Aerosol Lidar and Infrared Path�nder Satellite Observation (CALIPSO) level 2, 333 m and 5 km lidar cloud-layer products between 01 January 2007 and 01 January 2008. Merging both data sets together thereby allowed the best possible simulation of near global and seasonal real world atmospheric conditions that a spaceborne Integrated Path Differential Absorption (IPDA) lidar like MERLIN will encounter. With 40.5 % overall global cloud free fraction, a cloud gap distribution which is following a power-law distribution with exponent 1.51 +-� 0.01 together with a mean cloud gap length of 7.41 km and about 200 daily global cloud top flatness events, the analysis reveals a dominance of small cloud gaps which is confirmed by a low median cloud gap length of only 1 km. While the cloud free fraction results were compared and confirmed with Aqua Moderate Resolution Imaging Spectroradiometer (MODIS) seasonal and annual cloud fraction data, the power-law distribution of cloud gaps was confirmed by an extensive statistical analysis using maximum likelihood estimation, Kolmogorov-Smirnov statistics and likelihood ratio tests. Taking 6.05 x 10e8 individual CALIPSO measurements of the year 2007 with a horizontal resolution of 333 m and computing cloud gap and cloud free statistics for 2 �x 2� latitude/longitude grid points thereby identified regional and seasonal changes in the probability of spaceborne lidar surface detection. The analysis reveals that MERLIN will be able to perform near global methane mixing ratio column retrievals.