Constantino Muñoz

Effects of noise on lidar data inversion with the backward algorithm.

The lidar data-inversion algorithm widely known as the Klett method (and its more elaborate variants) has long been used to invert elastic-lidar data obtained from atmospheric sounding systems. The Klett backward algorithm has also been shown to be robust in the face of uncertainties concerning the boundary condition. Nevertheless electrical noise at the photoreceiver output unavoidably has an impact on the data-inversion process, and describing in an explicit way how it affects retrieval of the atmospheric optical coefficients can contribute to improvement in inversion quality. We examine formally the way noise disturbs backscatter-coefficient retrievals done with the Klett backward algorithm, derive a mathematical expression for the retrieved backscatter coefficient in the presence of noise affecting the signal, and assess the noise impact and suggest ways to limit it.

3D Scanning Portable Backscatter Lidar Platform for Atmospheric Remote Sensing: Performance and Architecture overview

This article is aimed at describing the technology, system architecture and specifications of a new 3D Nd:YAG scanning lidar. Main features of the system are interspersed low-range and far-range exploration, open user-configuration scanning tools and a specific architectural design based on parallel CPU control, a LabView user interface and a digitally controlled optoelectronic receiver. The latter provides key advantages to the whole system architecture such as calibration of lidar returns in terms of absolute power and repeatability. Issues concerning system responsivity calibration, receiver gain self-calibration, automatic gain control and synchronization offset-drift zeroing and the like, all of which are of prime importance for the lidarist, are presented. As far as we know, these contributions are new to the state-of-the-art of the community of optical and electronic lidar system designers.

Six-channel polychromator design and implementation for the UPC elastic/Raman LIDAR

A 6-channel dichroic-based polychromator is presented as the spectrally selective unit for the U.P.C. elastic/Raman lidar. Light emission is made at 355-nm (ultraviolet, UV), 532-nm (visible, VIS) and 1064-nm (near infrared, NIR) wavelengths. In reception, the polychromator is the spectral separation unit that separates the laser backscattered composite return into 3 elastic (355, 532, 1064-nm wavelengths) and 3 Raman channels (386.7, 607.4 and 407.5-nm (water-vapor) wavelengths). The polychromator houses photo-multiplier tubes (PMT) for all the channels except for the NIR one, which is avalanche photodiode (APD) based. The optomechanical design uses 1-inch optics and Eurorack standards. The APD-based receiver uses a XY-axis translation/elevation micro-positioning stage due to its comparatively small active area and motorised neutral density filters are used in all PMT-based channels to avoid detector saturation. The design has been specially optimized to provide homogeneous spatial light distribution onto the photodetectors and good mechanical repeatability. All channels are acquired in mixed analog and photon-counting mode using Licel® transient recorders, which are controlled by means of a user friendly LabVIEWTM interface. The paper focuses on the main polychromator optical design parameters, that is, light collimation trade-offs, end-to-end transmissivity, net channel responsivity, light distribution and spot size onto the photodetectors. The polychromator along with the rest of the U.P.C. lidar system has successfully been tested during a recent lidar system intercomparison campaign carried out in Madrid (Spain) during Oct. 2010.

Separation of aerosol fine‐ and coarse‐mode radiative properties: Effect on the mineral dust longwave, direct radiative forcing

An improvement of the estimation of mineral dust longwave, direct radiative forcing is presented. It is based on recent developments that combine Sun photometer and multiwavelength lidar data to retrieve range-resolved coarse- and fine-mode extinction coefficients. The forcings are calculated separately for each mode, and their sum is compared to the classical approach in which only the total extinction is considered. The results of four cases of mineral dust intrusion in Barcelona, Spain, show that when the coarse mode predominates, the longwave forcings calculated with the classical approach are underestimated up to 20% near the surface. In all cases the strong coarse-mode predominance near the surface has also an effect on the forcing in the upper layers.

Power budget and performance assessment for the RSLAB multispectral elastic/raman lidar system

The need of a multi-spectral lidar has widely been experienced in last few years with a view to invert the optical and microphysical properties of aerosols and their impact on the climate change. As a part of the EARLINET-GALION objectives, a joint effort has already been made by the European Aerosol Research Lidar Network (EARLINET). The EARLINET advanced standard of 3+2-channel configuration for lidar instruments (3+2 standing for 3 elastic channels and 2 respective Raman channels) enables retrieval of aerosol microphysical properties. An overview of the new RSLAB 3+2+1 multispectral lidar system, therefore, is presented in terms of power budget estimation for all the reception channels and overall system performance, that is, signal-to-noise ratio (SNR) and maximum sounding range achieved.

The European Aerosol Research Lidar Network (EARLINET): An Overview

The European Aerosol Research LIdar NETwork (EARLINET) is the first aerosol lidar network on a continental scale with the main goal to provide a comprehensive, quantitative, and statistically significant database for the aerosol distribution over Europe. Next, we present EARLINET along with the main network activities.

Piece-wise variance method for signal-to-noise ratio estimation in elastic/Raman lidar signals

A straightforward signal-to-noise ratio (SNR) estimator for elastic/Raman lidar channels and related noise-induced errorbars is presented. The estimator is based on piece-wise estimation of the mean signal power and noise variance component under analog detection. The piece-wise estimator results are compared with those obtained from a previously published SNR parametric estimator under high and low SNR scenarios.

A wind speed and fluctuation simulator for characterizing the wind lidar correlation method

Aerosol distribution in the atmosphere is used as a wind-tracer by lidars since it is drifted by the wind and respond to its changes. Two methods are used: the correlation and the Doppler method. This first one is easier and cheaper to implement than the latter. It makes it competitive for retrieving wind speed profiles or estimating wind turbulence. However, its accuracy decreases significantly as the distance from the optical radiation source increases, hence the need to characterize the method by means of wind field profile simulations and validate it by comparing the retrieval of real wind velocities with that of cooperative instruments such as radiosoundings. In this paper, the bases of a 2D lidar signal simulator are presented. The relationship between wind fields and the aerosol concentration dynamics, and the way they relate to lidar signal returns is explained. The first results of the application of the correlation method for the retrieval of wind velocities from real data at the UPC are presented and compared to radiosoundings measurements.

Engineering of a water-vapour, Raman, elastic-backscatter Lidar at the Technical University of Catalonia (Spain)

Implementation of the pure-vibrational Raman spectra lidar method for simultaneous measurements of atmospheric water-vapour, aerosol extinction and backscatter coefficients is reported. A Q-switched Nd:YAG laser provides the three elastic wavelengths of 1064, 532 and 355 nm while the return signal is collected by a 40-cm aperture telescope. A spot-to-spot fiber bundle conveys the light from the telescope focal plane to a specific polychromator especially simulated and designed with care on minimizing optical losses and physical dimensions. The reception field of view, which is limited by the fiber bundle characteristics, is the same for all wavelengths. By means of four customised dichroic filters and beam splitters, light is separated into the three elastic wavelengths (355, 532, 1064 nm) as well as the 386.7- and 607.4-nm N2-Raman-shifted wavelengths, and the 407.5-nm H2O-Raman-shifted wavelength. Signal detection is achieved by using avalanche photodiodes at 1064 and 532 nm and analog acquisition while photomultiplier tubes and fast photon counting acquisition at the rest of the wavelengths. A specific design of the optoelectronics of the receiving channels is controlled by a distributed CPU thanks to a user-friendly LabViewTM interface. User-configurable scanning tools are built-in, but can also be customized. In this work an overview of the system though particularly geared to the polychromator unit is presented as well as a power link-budget assessment, which is to include simulation of end-to-end transmissivities, will be discussed for the main channels involved. The first measurements have already been made at 1064, 532, and 607.4 nm.

Statistical considerations on the Raman inversion algorithm: data inversion and error assessment

Lidar (radar laser) systems take advantage of the relatively strong interaction between laser light and aerosol/molecular species in the atmosphere. The inversion of optical atmospheric parameters is of prime concern in the fields of environmental and meteorological modelling and has been (and still is) under research study for the past four decades. Within the framework of EARLINET (European Aerosol LIdar NETwork), independent inversions of the atmospheric optical extinction and backscatter profiles (and thus, of the lidar ratio, as well) have been possible by assimilating elastic-Raman data into Ansmann et al.’s algorithm [the term “elastic-Raman” caters for the combination of one elastic lidar channel (i.e., no wavelength shift in reception) with an inelastic Raman one (i.e., wavelength shifted)]. In this work, an overview of this operative method is presented under noisy scenes along with a novel formulation of the algorithm statistical performance in terms of the retrieved-extinction mean-squared error (MSE). The statistical error due to signal detection (Poisson) is the main error source considered while systematic and operational-induced errors are neglected. In contrast to Montercarlo and error propagation formulae, often used as customary approaches in lidar error inversion assessment, the statistical approach presented here analytically quantifies the range-dependent MSE performance as a function of the estimated signal-to-noise ratio of the Raman channel, thus, becoming a straightforward general formulation of algorithm errorbars.