Articles | Volume 9, issue 10
https://doi.org/10.5194/amt-9-4935-2016
https://doi.org/10.5194/amt-9-4935-2016
Research article
 | 
07 Oct 2016
Research article |  | 07 Oct 2016

Assessment of lidar depolarization uncertainty by means of a polarimetric lidar simulator

Juan Antonio Bravo-Aranda, Livio Belegante, Volker Freudenthaler, Lucas Alados-Arboledas, Doina Nicolae, María José Granados-Muñoz, Juan Luis Guerrero-Rascado, Aldo Amodeo, Giusseppe D'Amico, Ronny Engelmann, Gelsomina Pappalardo, Panos Kokkalis, Rodanthy Mamouri, Alex Papayannis, Francisco Navas-Guzmán, Francisco José Olmo, Ulla Wandinger, Francesco Amato, and Martial Haeffelin

Abstract. Lidar depolarization measurements distinguish between spherical and non-spherical aerosol particles based on the change of the polarization state between the emitted and received signal. The particle shape information in combination with other aerosol optical properties allows the characterization of different aerosol types and the retrieval of aerosol particle microphysical properties. Regarding the microphysical inversions, the lidar depolarization technique is becoming a key method since particle shape information can be used by algorithms based on spheres and spheroids, optimizing the retrieval procedure. Thus, the identification of the depolarization error sources and the quantification of their effects are crucial. This work presents a new tool to assess the systematic error of the volume linear depolarization ratio (δ), combining the Stokes–Müller formalism and the complete sampling of the error space using the lidar model presented in Freudenthaler (2016a). This tool is applied to a synthetic lidar system and to several EARLINET lidars with depolarization capabilities at 355 or 532 nm. The lidar systems show relative errors of δ larger than 100 % for δ values around molecular linear depolarization ratios (∼ 0.004 and up to ∼  10 % for δ = 0.45). However, one system shows only relative errors of 25 and 0.22 % for δ = 0.004 and δ = 0.45, respectively, and gives an example of how a proper identification and reduction of the main error sources can drastically reduce the systematic errors of δ. In this regard, we provide some indications of how to reduce the systematic errors.

Download
Short summary
This work analyses the lidar polarizing sensitivity by means of the Stokes–Müller formalism and provides a new tool to quantify the systematic error of the volume linear depolarization ration (δ) using the Monte Carlo technique. Results evidence the importance of the lidar polarizing effects which can lead to systematic errors larger than 100 %. Additionally, we demonstrate that a proper lidar characterization helps to reduce the uncertainty.