LIDAR RATIOS OF DUST OVER WEST AFRICA MEASURED DURING “SHADOW”

The lidar ratios of Saharan dust at 355 and 532 nm (LR355 and LR532) measured over West Africa during SHADOW field campaign are analyzed. Results demonstrate that even for pure dust, the lidar ratio may present strong height dependence. The possible reasons of height dependence of lidar ratios during strong dust events are considered.


INTRODUCTION
Atmospheric dust provides significant impact on the Earth system and this impact still remains highly uncertain. The Sahara desert is the largest source region of mineral dust, so quantification of the Saharan dust optical and microphysical properties is an important contribution to the study of the Earth radiative balance. In modeling the direct aerosol effect, the vertical profiles of the extinction coefficient belong to the basic inputs. When the extinction profiles are derived from the measurements of elastic backscatter lidar, the lidar ratio (extinction to backscattering ratio) is a key parameter in data inversion. The great advantage of Raman and HSRL lidars is in their ability to provide independent measurements of aerosol backscattering and extinction coefficients. The numerous lidar stations regular profile the properties of the Saharan dust layers during their transport to Europe and North America.
The properties of dust particles are modified during transport in atmosphere, so it is important to study the dust properties in the region of origin. Such observations were performed during SAMUM-1,2 experiments in 2006, 2008 (in Morocco and Capo Verde respectively), as well as during SHADOW campaign in Senegal in 2015-2016 in Senegal [1][2][3][4][5]. However, the dust is a mixture of various minerals whose relative abundance depends on dust origin, so even near the source of origin the dust properties may present significant variations. Atmospheric dust in Africa can also contain the products of biomass burning and the local pollutions. As a result, the comparison of lidar ratios observed during different campaigns my look controversy. The averaged values of LR observed during SAMUM-2 campaign, are LR355=52.7±10.2 and LR532=54.2±9.6 [3]. Corresponding averaged value of depolarization ratio at 532 nm is 31%, indicating that heavy dust episodes were predominant. In contrast to SAMUM, where lidar ratios at 355 and 532 nm were close, SHADOW measurements demonstrate that in many dust episodes the lidar ratio at 355 nm (LR355) significantly exceeded LR532, which was related by Veselovskii et al. [2] to increase of the imaginary part of the refractive index in UV. In our presentation we analyze vertical variation of the lidar ratios and consider relationship between LR355 and LR532 for different dust episodes.  Detailed information regarding the SHADOW campaign and the instruments is presented in [2]. The LILAS lidar is based on a tripled Nd:YAG laser with a 20 Hz repetition rate, and pulse energy of 90/100/100 mJ at 355/532/1064 nm. The aperture of the receiving telescope is 400 mm. The lidar allows evaluation of three particle backscattering (355 nm, 532 nm, 1064 nm) and two extinction coefficients (355 nm, 532 nm). To improve the system performance at 532 nm, the rotational Raman channel was used instead of the vibrational one. For period 1-4 April 2015 a strong dust episode occurred. Fig.1. shows vertical profiles of aerosol backscattering β and extinction α coefficients at 355 nm and 532 nm together with particle depolarization ratio 532, obtained during two measurements sessions on the nights 1-2 and 2-3 April. For both sessions 532 exceeds 30%, and depolarization presents small decrease with height. The values α355 and α532 are close, while β355 exceeds β532, which can be related to increase of the imaginary part mI of the refractive index of dust at 355 nm, as discussed in [2].  One of possible explanations of observed height dependence of the particle intensive parameters is that at high altitudes the dust particles are mixed with another type of aerosol. The MERRA-2 aerosol transport model predicts, that for period considered the particles of organic carbon (main component of biomass burning aerosol) may occur above 2000 m [4]. Organic carbon (OC) is characterized by increased absorption in UV, which could explain the increase of LR355. Presence of small OC particles could also explain the increase of the Angstrom exponent. However, for explanation of LR532 decrease with height we should assume that LR532 of OC is low. This may look unexpected, because lidar measurements usually provide high values of LR532 for the smoke (above 60 sr). We should keep in mind however, that our measurements were performed at condition of low relative humidity (below 30%).

Dust-smoke episodes
To analyze the values of lidar ratio of OC and their dependence on the relative humidity (RH), we have considered several smoke episodes during the campaign. The maximal value of RH in these episodes varied from 30% to 90%. The results obtained on 15 December 2015 are shown in Fig.3. The figure presents the lidar ratios at 355 and 532 nm together with profile of RH, calculated from lidar derived water vapor mixing ratio and radio sonde measured temperature. Maximal value of RH is below 40%. The LR532 and LR355 are close at 1000 m, but LR532 decreases with height from 50 sr to 40 sr, while LR355, in opposite, increases up to 70 sr. Assuming, that depolarization ratio of dust and smoke are known (35% and 7% correspondingly) the backscattering β532 can be decomposed for dust and smoke contributions [5], as shown in Fig.3c. The contribution of smoke to β532 becomes predominant above 2000 m, and corresponding value of LR532 is about 40 sr. For other dustsmoke episodes characterized by higher RH, the values of LR532 and LR355 are higher. Fig.4

Dust episode on 10-24 April 2015
Strong dust episode occurred also for period 10-24 April. Fig.5 shows extinction coefficients and lidar ratios at 355 and 532 nm for two sessions on 15-16 and 23-24 April. The depolarization ratio for both days is about 30%, so particles can be considered as a pure dust. Still, in contrast with 1-4 April episode, LR355 and LR532 are close, which can be an indication that absorption in UV is lower. To analyze the difference in dust origin, Fig.6

CONCLUSIONS
Presented results demonstrate that lidar ratio at 532 nm during dust episodes may decrease with height, which can be due to two reasons: presence of smoke near the top of the dust layer and height variation of the dust properties (in particular height variation of the imaginary part). Analysis of smoke episodes show that the lidar ratio of smoke depends on RH and varies from 40 to 70 sr at 532 nm. This confirms that smoke can contribute to the decrease of observed LR532 at condition of low humidity.
Observations demonstrate also that for some dust events the lidar ratios at 355 nm and 532 nm coincide, while for other episodes LR355 significantly exceeds LR532. Such difference, in principle, can be the result if different mineralogy at the region of dust origin.