Study on retrieval algorithms for a backscatter lidar - Final report

In this study a systematic approach is used to find the most suitable algorithm for the inversion of space borne backscatter lidar signals. First a review of existing lidar algorithms is presented. The widely used analytical inversion method appeared to be the most promising one. For this algorithm a detailed error analysis is performed including propagation of all rel— evant systematic and statistical errors. Computer code is developed and tested using synthetic data as well as experimental ones. All tests proved the code to work well. For general use it is recommended to use the analytical inversion algorithm in the forward integration mode in conjunction with absolute calibration of the lidar signals. In special situations better stability can be achieved by using the backward integration mode, provided that suitable calibration targets at the far end of the lidar measurement range can be found. It is recommended to initiate a systematic investigation of objects in the atmosphere and at the surface which could be used as calibration targets. Additionally, further investigations should be performed concerning a modi— fied inverse modeling approach. This algorithm appears to have a high potential for avoiding some of the instabilities due to the rather large noise contribution to the expected ATLID signals