Noise Extraction for Raman Lidar Signal Processing

摘要

The return Raman Lidar signal contains a strong elastically scattered component (at λ_0) that is useful for profiling clouds and aerosols and also weaker inelastically scattered components that provide chemical-specific information. For profiling water vapor, we use components produced by vibrational Raman effect that produces energy shifts characteristic of the molecules in the atmosphere (3652 cm~(-1) for water vapor, 2331 cm~(-1) for nitrogen). The aim of this paper is to process lidar backscattered signal that contains water vapor and aerosol information in order to improve their recovery. Since they are affected by different kinds of noise, an appropriate filtering, with an improved recovery, represents a way to get good estimates of the above components. Water vapor and aerosols are two significant atmospheric components that are generally detected for a better knowledge of weather and climate. In spite of optical filters included in the experimental apparatus used for this paper, there is a need of further filtering, by using signal digital filtering. To discriminate noises from the main signal that is backscattered from sky, we are investigating on the use of appropriate digital filtering to be utilized in order to retrieval a noiseless signal. This approach is different from the current one that uses a poissonian averaging of collected data. In our investigation, we prefer to employ filters that preserve either amplitude information or phase one. Different kinds of filtering procedures have been used in order to isolate the main signal from noise.