Lidar remote sensing of atmospheric aerosols
Theme led by A. Miffre
The novelty of this work lies in the extreme sensitivity and precision of the experiments carried out.
This sensivity and accuracy results from that of the LIAPS-instrument light backscattering polarization detector, a world first for atmospheric aerosols.
Systematic biases being circumvent in laboratory, these biases are hence minimized in our field lidar instrument, which allow revealing the intrinsic vertical layering of the atmosphere.
Main findings
Revisiting the historical approach by Ångström (Optics Letters, 2020), selected as iLM spotlight 2020
In 1929, while studying the ability of light to reveal the size of mineral dust in the atmosphere, A. Ångström published a paper in which he introduced a quantity , describing the spectral behaviour of optical scattering, noting that "the larger the particle size, the smaller the value obtained for it". In contrast, this parameter takes the well-known value of 4 for molecular optical scattering.
At the iLM, by taking advantage of the polarisation of light, the authors revisited A. Ångström's historic approach to establishing the molecular scattering of light. Ångström's historical approach to establish experimentally the sensitivity of the parameter to the shape, spherical or non-spherical, of the particles studied. Using numerical simulation, the authors were able to reveal the parameter's dependence on the size and chemical composition of the spherical and non-spherical particles that make up the atmosphere. This result allows the atmospheric science community, and in particular aerosol physics and optical remote sensing, to interpret the values of that are measured every day, beyond the historical interpretation of A. Ångström's historical interpretation based on size speciation.
Fig. 1 : Left : Refractive index versus particles effective radius for spherical then non-spherical particles as numerically simulated. Right : The Angstroem parameter can serve to evaluate the particles size as a function of its refractive index, hence going beyond the traditional conclusion of a size indicator.
Revealing the vertical extent of photocatalytic nucleation process in the atmosphere with lidar (World first, PNAS, 2012), selected as iLM spotlight 2013
Optics Express 2014, OSA Spotlight Award 2014, Remote Sensing 2019
The formation of new particles in the atmosphere is key for cloud formation and therefore climate change, but is the result of a complex chain of reactions and processes that are not yet fully understood. This work is a world first, with lidar observation of nucleation in the atmosphere (OSA Spotlight Award), and therefore an understanding of the spatio-temporal extension of this process, which has yet to be revealed. It should be remembered that this process can now be detected at a distance of several kilometres. The Pi-polarimeter has revealed the underlying physico-chemical processes, and in particular the nucleation of aerosols in the atmosphere, coupled with photo-catalytic processes on the surface of particles of desert origin (PNAS 2012, Optics Express 2014, OSA Spotlight Award).
Read more about nucleation by lidar
The detection of aerosol nucleation presents a major difficulty, due in particular to the very small size of the new particles formed and their low concentration. A. Miffre and P. Rairoux have developed a new optical method, making it possible for the first time to measure remotely (lidar) the backscatter coefficient of the nucleation centres responsible for cloud formation in the atmosphere. In the presence of non-spherical particles of desert and volcanic origin, the authors used numerical simulation (T-matrix) to identify the optical requirements in terms of spectra and polarisation for the remote detection (lidar) of nucleation centres. The observation of this process in Lyon was made possible by optimising the sensitivity and precision of the ILM's lidar instrument. The extreme sensitivity and precision of the polarimeter's measurements also enabled unexpected phenomena to be observed, such as nucleation coupled with photo-catalytic processes on the surface of nanoparticles [PNAS], highlighting a new pathway in the nucleation process under normal temperature and pressure conditions. This result was published in the prestigious journal PNAS, after being submitted to Nature and Science. It led to a new publication in Optics Express (G. David et al., 2014, OSA Spotlight June 2014), in which we identified the link between optics and fundamental physics (nucleation processes). Figure 5 below shows the distribution in space and time of the backscatter coefficient of spherical nanoparticles in the atmosphere (beta_s). This coefficient, if used in the UV and in the presence of a very precise analysis of the polarisation of the backscattered wave, can be used as a tracer of the nucleation process in the atmosphere. This work opens the way to observe the vertical extent of nucleation.
Fig. 2 Physico-chemical mechanism for the formation of sulphuric acid droplets from desert sand particles. The optical signature of this nucleation can be observed in the atmosphere thanks to an extremely precise (10-5) UV polarisation measurement carried out by the ILM team. Lidar observation of particle formation in the atmosphere (Optics Express 2014, OSA Spotlight). First remote observation by lidar of the formation of new particles in the atmosphere, visualised by the increase, in the UV, of the βs backscatter coefficient specific to spherical particles, here between 2 and 3 km altitude, in the presence of Saharan desert sand dust.
Planck radiation emitted by atmospheric soot detected remotely by lidar (World first, Optics Express 2015)
The carbon aerosol is now recognized as a major uncertainty on climate change and public health, and specific instruments are required to address the time and space evolution of this aerosol, which efficiently absorbs light. In this paper, we report an experiment, based on coupling lidar remote sensing with Laser-Induced-Incandescence (LII), which allows, in agreement with Planck’s law, to retrieve the vertical profile of very low thermal radiation emitted by light-absorbing particles in an urban atmosphere over several hundred meters altitude. Accordingly, we set the LII-lidar formalism and equation and addressed the main features of LII-lidar in the atmosphere by numerically simulating the LII-lidar signal. We believe atmospheric LII-lidar to be a promising tool for radiative transfer, especially when combined with elastic backscattering lidar, as it may then allow a remote partitioning between strong/less light absorbing carbon aerosols.
 |
Fig. 3 Experimental set-up for LII-lidar detection then successive zooms enhancing the Planck radiation on the last figure on the left.
On the complex and roughned shape of mineral particles as detected by lidar and DDA-numerical simulations (Atm. Research 2018)
Our understanding of the contribution of mineral dust to the Earth's radiative budget is limited by the complexity of these particles, which present a wide range of sizes, are highly-irregularly shaped, and are present in the atmosphere in the form of particle mixtures. To address the spatial distribution of mineral dust and atmospheric dust mass concentrations, polarization lidars are nowadays frequently used, with partitioning algorithms allowing to discern the contribution of mineral dust in two or three-component particle external mixtures. In this paper, we investigate the dependence of the retrieved dust backscattering (βd) vertical profiles with the dust particle size and shape. For that, new light-scattering numerical simulations are performed on real atmospheric mineral dust particles, having determined mineralogy (CAL, DOL, AGG, SIL), derived from stereogrammetry (stereo-particles), with potential surface roughness, which are compared to the widely-used spheroidal mathematical shape model. For each dust shape model (smooth stereo-particles, rough stereo-particles, spheroids), the dust depolarization, backscattering Ångström exponent, lidar ratio are computed for two size distributions representative of mineral dust after long-range transport. As an output, two Saharan dust outbreaks involving mineral dust in two, then three-component particle mixtures are studied with Lyon (France) UV–VIS polarization lidar. If the dust size matters most, under certain circumstances, βd can vary by approximately 67% when real dust stereo-particles are used instead of spheroids, corresponding to variations in the dust backscattering coefficient as large as 2 Mm− 1·sr− 1. Moreover, the influence of surface roughness in polarization lidar retrievals is for the first time discussed. Finally, dust mass-extinction conversion factors (ηd) are evaluated for each assigned shape model and dust mass concentrations are retrieved from polarization lidar measurements. From spheroids to stereo-particles, ηd increases by about 30%. We believe these results may be useful for our understanding of the spatial distribution of mineral dust contained in an aerosol external mixture and to better quantify dust mass concentrations from polarization lidar experiments.
Fig. 4 Examples of smoothened and roughned versions of mineral particles which are computed with DDA.
Optical partitionning of two, three-component particle external mixtures (Atmos. Chem. Phys., 2013, Atmos. Res. 2018)
The spatio-temporal distribution of atmospheric nanoparticles of various origins has thus been analysed with record precision and sensitivity, thanks to the sensitivity and precision of the Pi-Polarimeter: urban particles [AtmEnv2010], particles contained in clouds of volcanic ash [GRL 2011, JTECH 2011, AtmEnv 2012], or desert sand [ACP 2013, Atm Res 2018, Rem Sens 2019, Opt. Lett 2020], whose distinctive feature is that they are transported and dispersed in the atmosphere over several thousand kilometres. These publications are the result of the record sensitivity achieved in measuring polarisation (3×10-4 at 4 km altitude), which is close to molecular depolarisation. Following a robust calibration procedure for the polarisation detector, the analysis of the lidar signals obtained highlighted the link between the backscatter coefficient measured by lidar and the formalism of the scattering matrix. Thanks to a very precise analysis of the polarisation of the backscattered wave, we were able to carry out an optical partitioning between spherical and non-spherical nanoparticles contained in an external mixture formed by these two types of nanoparticles. In the atmosphere, this corresponds, for example, to a mixture of volcanic ash (non-spherical) and sulphates (spherical), such as the filament of ash from the Icelandic volcano observed by lidar in Lyon.
Calibration of the Lyon polarization lidar instrument (Appl. Phys. B 2012, Remote Sensing 2019)
Derived from the aerosol labratory Pi-polarimeter, this remote sensor is unique in its accuracy and sensitivity, particularly in UV depolarisation where molecular scattering is strong. Particular care has therefore been taken to systematically analyse the biases affecting the measurement (Applied Phys B, 2012): contribution of solar radiation, limitation of molecular scattering to the Cabannes line, polarisation and spectral cross terms cancelled to better than 10-7. Thus, the originality of our work lies in its sensitivity and accuracy: we measure depolarisations as low as 0.4% (close to molecular depolarisation), over two orders of magnitude (40% depolarisation for volcanic ash), with high accuracy (a few percents), and over several kilometres (typically from 500 metres to 5 km altitude). Figure 1 shows a photograph of our lidar detector, the calibration of which, carried out in the UV and VIS in the real atmosphere, is made robust by the cancellation of polarisation and spectral cross-terms.
 |
 |
|
| Fig. 5 Lyon lidar remote sensor: (a) Real atmosphere, (b) Photograph of the UV-VIS sensor, (c) UV-VIS polarisation calibration method and curve (Remote Sensing 2019), (d): Lidar remote sensing observation of the cloud of volcanic ash from the Icelandic volcano in the Lyon atmosphere (Spring 2010, Miffre et al., AtmEnv, 2012). |
||
