Data processing and inversion interfacing the upc elastic-raman lidar system

Resumen

In the increasing environmental regulatory framework LIDAR (Light Detection and Ranging, a synonym of laser radar) offers unprecedented capabilities for atmospheric remote sensing. The RSLAB Barcelona lidar station of the Universitat Politècnica de Catalunya is part of the European Aerosol Research Lidar Network (EARLINET) whose main aim is the observation of the temporal and spatial distribution of atmospheric aerosols (particles) at continental scale. The international context is GALION (Global Atmospheric Watch Atmospheric Lidar Observation Network)-GEOSS (Global Earth Observation System of Systems). The core of this Ph.D. is the range-resolved retrieval of the atmospheric optical parameters, namely, aerosol atmospheric extinction, backscatter and their lidar ratio, by using advanced data processing techniques and the 2+1-channel elastic/Raman RSLAB lidar instrument. These optical parameters play an essential role to improve environmental and forecast models and, ultimately, our knowledge of the atmosphere. Though lidar inversion algorithms have been under discussion over 3 decades, a systematic computer-oriented formulation of them including a thorough treatment of the noise and a detailed conceptual formulation of the lidar signal processing practicalities and trade-offs involved is missing. The inversion techniques studied in this Ph.D. encompass both elastic algorithms (i.e., without wavelength shift in reception) and the inelastic (Raman-shifted) lidar inversion technique. Specifically, this Ph.D. focuses on: One- and two-component elastic inversion algorithms: Impact of a noisy calibration and error mitigation, independent inversion of the aerosol lidar ratio by assimilating co-operative information from a sun-photometer or a two-point calibration. Automation. Real-data GUI inversion platform (KLTmenu). Raman-lidar inversion: Data treatment using variant FIR (Finite Impulse Response) smoothing and variant linear fitting. Assessment of the spatial inversion resolution. Real-data GUI inversion platform (SIMRAM). Estimation of inversion errors: Signal-to-noise ratio (SNR) estimators for elastic and Raman lidar channels. One-component elastic-inversion backscatter error bars due to backscatter calibration, observation noise, and lidar-ratio error sources. Combined elastic/Raman lidar inversion error bars and extinction error bars under moderate-to-high and low SNRs. Maximum-likelihood lidar-ratio estimation. Application results of the inversion platforms and methodologies developed in this Ph.D. encompass a CALIPSO-satellite calibration/validation case example and the 2007-2009 Barcelona aerosol climatology. From the data-processing point of view, variant spatial filtering suggests to be a low-distortion, noise-robust technique for the combined elastic/Raman inversion method. Error bars and SNR estimators are key tools to assess the quality of the inversion results.