Investigating emission sources and transport of aerosols in Siberia using airborne and spaceborne LIDAR measurements

Airborne backscatter lidar measurements at 532 nm were carried out over Siberia in July 2013 and June 2017. The Russian Tu-134 flew over major Siberian cities (Novosibirsk, Tomsk, Krasnoyarsk, Yakutsk), the gas flaring fields of the Ob valley and Siberian Taiga in order to sample several kinds of Siberian aerosol sources. Aerosol types are derived using the Lagrangian FLEXible PARTicle dispersion model (FLEXPART) simulations, Moderate Resolution Imaging Spectrometer (MODIS) Aerosol Optical Depth (AOD), Infrared Atmospheric Sounding Interferometer (IASI) CO total column and AOD at 5 10 μm. Forest fire detection is based on NASA Fire Information for Resource Management System (FIRMS) from MODIS and the Visible Infrared Imaging Radiometer Suite (VIIRS) observations and airborne in-situ measurements when available. Six aerosol type could be identified in this work: (i) Dusty aerosol mixture (ii) Ob valley industrial emission (iii) fresh boreal forest fire plumes (iv) aged forest fire plumes (v) pollution over the Tomsk/Novosibirsk region (vi) long range transport of Chinese pollution over Yakutsk. The backscatter to extinction ratio and then the corresponding lidar ratio (LR) were derived 10 for each of these 6 identified aerosol type, using an iterative method based on the Fernald forward inversion constrained by the 10 km MODIS collection 6 AOD distribution closed to the airborne lidar observation. The LR analysis showed that the lowest LR range was obtained for the "Dusty Mix" case (26-40 sr) and the highest for the urban and industrial pollution from the Tomsk/Novosibirsk area (71-90 sr). The comparison is good with previous estimate of LR according to the aerosol classification. The range of lidar ratio obtained for gas flaring emission (43-60sr) was lower than the high values encountered 15 in the Tomsk/Novosibirk urban area and has never been characterized using lidar observations. Airborne lidar backscatter ratio vertical structure, aerosol types and integrated LR derived from the airborne data analysis were compared to nearby CALIOP overpasses. These comparisons showed three main differences with the CALIOP LR and aerosol type classification over Siberia: (i) CALIOP aerosol layer can be classified as Elevated smoke instead of Polluted continental and vice versa, but with little influence on the LR value (ii) aging and transport of aerosol layers effect on the CALIOP LR value is not always 20 properly accounted for even when the CALIOP classification is correct (iii) the lack of discrimination between fresh and old fire plume leads to an overestimation of the optical depth for the fresh fires in the CALIOP AOD over the fire source region. 1 https://doi.org/10.5194/acp-2020-195 Preprint. Discussion started: 21 April 2020 c © Author(s) 2020. CC BY 4.0 License.


Introduction
Atmospheric aerosols play a very important role in many meteorological, radiative and chemical processes taking place in the atmosphere such as cloud formation, scattering and absorption of incident solar (short-wave) and thermal (long-wave) radiation 25 from the Earth, as well as affect the air quality (Chỳlek and Coakley, 1974). Due to the variety of the optical, microphysical and chemical properties of atmospheric aerosols, closely depending on the formation processes of particulate matter and their subsequent aging processes occurring in the atmosphere, and the poor knowledge of their spatio-temporal distribution as well, they have been identified by the Intergovernmental Panel on Climate Change (IPCC) as one of the main uncertainty sources when assessing radiative forcing and the climate change (Stocker et al., 2013). The aircraft performed various legs at low altitude (≈ 0.6 km) and between 4 and 8 km in order to sample different atmospheric layers. Only lidar data collected during flights above 4 km are used for the aerosol layer characterization. 90 2.2 Airborne lidar system The lidar system installed on board the Tu-134 aircraft is based on the LOSA aerosol lidar developed at the IAO SB RAS (Balin et al., 2011;Penner et al., 2015). The transmitter module is based on a solid state Nd-YAG laser emitting 8 ns laser pulses at 1064 nm and 532 nm. The maximum output energy at 532 nm is 100 mJ with a repetition rate of 10 Hz and a beam divergence of 2.5 mrd. The optical receiver is a 150 mm diameter reception lens coupled with a 1nm filter and two reception channels (co-95 and cross-polarization). The full geometrical overlap is obtained between 80m and 150m. In practice the first 200 m are not used to reduce the errors when estimating the overlap function correction in clear air region below the aircraft. The detection unit is composed of a photomultiplier coupled to an analog-to-digital converter (ADC) electronic system with a sampling rate range of 25 -100 MHZ (i.e. a 1.5-6m vertical resolution) and a resolution of 12 bits. The cross-polarization calibration is not sensitive enough to characterize the aerosol type and is mainly used to discriminate cloud and aerosol layers. The near infrared 100 channel was not available during the aircraft campaigns. Attenuated backscatter signal in decimal logarithm > -2.3 with a signal above detection threshold in the depolarization channel is considered as cloud. The initial lidar data temporal averaging Dust outbreaks from the Eastern Asia desert are only taken into account if the IASI 10 µm AOD is greater than 0.08 in the area with elevated PES (Peyridieu et al., 2013;Capelle et al., 2014Capelle et al., , 2018 and when CALIOP aerosol depolarization ratio is greater than 15% in the same area (Tesche et al., 2011;Groß et al., 2011Groß et al., , 2013. Urban pollution aerosol sources are considered only when large cities (> 500000 inhabitants) are included in the high PES area. The location of flaring sources is based on 170 the anthropogenic emissions ECLIPSEv4 data-set (Evaluating the Climate and Air Quality Impacts of Short-Lived pollutants) described in Klimont et al. (2017). This inventory includes in particular the gridded methane emissions from gas flaring in the Russian Arctic at a 0.5 o x 0.5 o horizontal resolution. A threshold of 50 moles.km −2 /hour has been applied to the methane emissions to select areas that could potentially be defined as flaring sources. Elevated CO tropospheric column (> 2.10 18 molecule.cm −2 ) is also mandatory to check the contribution of industrial and urban combustion aerosol sources (Wang et al.,175 2018).
In addition to this first guess for the aerosol type identification, in-situ aircraft measurements of CO concentration, black carbon (BC) mass concentration and aerosol size distribution are also analyzed for aircraft ascent or descent across the aerosol layer observed by the airborne lidar. Excess of CO (∆CO), i.e. the difference with the background CO concentration taken as the minimum of CO measured during the two campaigns in the lower troposphere (0-5 km), must reach 30 ppb for biomass burning 180 aerosol and gas flaring emission (Paris et al. (2009b)). Black carbon mass concentrations is also used to identify combustion aerosol: BC > 0.5 µg.m −3 and BC maximum correlated with elevated ∆CO. The ratio of the aerosol concentration in the cloud condensation nuclei (CCN) mode between 15 nm and 80 nm over the aerosol concentration in the Aitken nuclei (AN) mode between 80 nm and 200 nm is used to identify the aerosol aging (Willeke and Whitby, 1975;Bäumer et al., 2008;Furutani et al., 2008). The complete methodology for the aerosol type identification is summarized in the upper panel of Fig. 2.

Aircraft campaign data analysis
Six flights have been selected during the 2013 and 2017 campaigns, because they correspond to different aerosol sources. The lidar vertical cross-section of the calibrated attenuated backscatter (PR2) have been used to identify the horizontal and vertical extent of the aerosol layers (532 nm scattering ratio larger than 1.5). For each aerosol plumes, the AOD and the integrated LR are calculated, while the aerosol type is retrieved using the methodology described in section 3.    The role of dust emission is also significant for this case study when looking at the 4-day average of the 10 µm AOD measured 205 by IASI which is in the range 0.15-0.35 above Kazakhstan (Fig. 5). A CALIOP overpass on June 16 also shows high aerosol depolarizing ratio (15%-20%) up to 3 km altitude at 50 o N,70 o E above Kazakhstan (Fig. 5). An aerosol depolarization ratio less than 20% is consistent with polluted dust aerosol (Tesche et al.  Groß et al. (2013)). In the flight area, IASI also detected a 10 µm AOD between 0.08-0.1, which is smaller than above Kazakhstan, but reaches the dust detection threshold of 0.08. The aerosol layer observed by the airborne lidar at 57 o N,80 o E can be considered 210 as a mixture of dust, industrial pollution the Ob Valley and aged smoke. When looking at the in-situ measurement made by the aircraft, ∆CO range is between 20 and 30 ppbv up to 2.5 km and the BC mass concentration increases from 0.2 µg.m −3 at 2 km to 0.4 µg.m −3 at 500 m. These moderate values of ∆CO and BC compared to other flights (Table 1) are also consistent with a dusty aerosol mixture.

215
On June 18, 2017, the aircraft flew again between Novosibirsk and Surgut and aerosol layers have been observed at higher latitudes above the gas and oil extraction field of the Ob Valley at 59 o N. he PR2 latitudinal cross-section when the aircraft is flying at 4.5 km, shows an aerosol layer in the 0 and 3.5 km altitude range. The AOD and LR are calculated for a 6 min average profile shown by the red rectangle in Fig. 6. The lidar ratio range is between 43 sr and 60 sr when using 0.2 and 0.26 for respectively the 25 th and 75 th percentiles of the nearby MODIS 550 nm AOD. The corresponding 532 nm AOD for the 220 airborne lidar is then 0.23 ± 0.06 (Table 1)  and 0.075-0.081 respectively when calculating the lidar ratio at these two longitudes. For the three selected profiles, we obtain the same LR range of 32-39 sr, while the AOD measured by the lidar at 532 nm is 0.115 ± 0.01 for this biomass burning plume.  The lidar ratio range is between 71 sr and 90 sr when using 0.13 and 0.21 for respectively the 25 th and 75 th percentiles of the 280 nearby MODIS 550 nm AOD (Table 1). The corresponding 532 nm AOD for the airborne lidar is then 0.17 ± 0.05 (Table 1).
The strong PES values obtained with the FLEXPART backward simulation for the aerosol layer observed by the lidar (

Long range transport of Northern China emissions
The last case study corresponds to the lidar observations near the city of Yakutzk (62 o N, 129 o E) on July 19, 2013. A PR2 vertical cross-section when descending to Yakutsk shows several aerosol layers in the 0 and 5 km altitude range (Fig. 17). The AOD and LR are calculated for a 40 sec average profile shown by the red rectangle in Fig. 17 when the aircraft is high enough 295 to sample the entire aerosol layer between 0 and 4 km. The lidar ratio range is between 41 sr and 51 sr when using 0.1 and 0.13 for respectively the 25 th and 75 th percentiles of the nearby MODIS AOD at 550 nm (Table 1). The corresponding AOD for the airborne lidar at 532 nm is then 0.12 ± 0.04 (Table 1). China is in the lower range (40-50 sr) as expected for aged aerosol and southerly flow with many clouds and high humidity over Eastern Siberia (Fig. 18) Table 2). The aerosol type and the range of the LR distribution for the CALIOP profile is then taken from the Version 4.1 level 2 CALIOP aerosol data product Kim et al. (2018).
The comparisons of the total backscatter ratio vertical profiles for the six case studies presented in Table 1 are shown in Fig.   20. The positions and dates of the CALIPSO profiles chosen for this comparison are given in Table 2 along with the spatial and 365 temporal differences between the aircraft measurements and the CALIOP profiles. Only one case above the Ob valley on June 18 ( Fig. 20d) corresponds to a distance less than 150 km and a time lag less than 12 h. For the other cases, a strong sensitivity to a spatial distance larger than 200 km is only expected on June 16 (Fig. 20a) when the air mass transport direction is not parallel to the line connected the aircraft and CALIOP profile positions, and when the differential advection of the aerosol plume may change the vertical structure of the total backscatter. The range of the selected CALIOP backscatter ratio profiles in 370 Fig. 20 differs from that of airborne lidar by less than 20% and the thicknesses of the aerosol layers are in good agreement. The comparison for the case with long range transport from Northern China even shows a surprisingly good agreement (Fig. 20f) considering the large distance (576 km) between the two measurements mainly because the transport pathway is parallel to the line connected the CALIOP footprint and the aircraft position. Because the sensitivity to spatial and temporal mismatches is expected to be large for the biomass burning cases, two CALIOP vertical profiles are selected in Fig. 20b, c provided that the 375 0.8·PES max criteria is still true. For the fresh fire case (Fig. 20b), a layer with similar structure and backscatter ratio magnitude was seen by CALIOP 60 hours earlier and 500 km further north, and should be also considered in the comparison. For the aged forest fires, the differences between the two CALIOP vertical profiles are quite small and the CALIOP profiles are then fairly representative of the airborne lidar observations. The 532 nm AOD calculated for the aircraft and CALIOP profiles ( Table  Figure 20. Comparison of aircraft and CALIOP averaged backscatter ratio vertical profile. 2) are very similar except for the aged urban pollution layers transported from northern China where the AOD differences 380 between CALIOP and the airborne lidar is 0.16 (+57%). This is mainly due to large difference of lidar ratio (70 sr instead 45 sr) between the CALIOP retrieval and the aircraft data (see hereafter). Therefore the selected CALIOP profiles are suitable to discuss the differences between the CALIOP V4.2 lidar ratio values compared to the airborne lidar data analysis.
The aerosol composition for the six averaged CALIOP profiles are given in Table 2 using the CALIOP level 2 aerosol type and taking into account the thickness of the aerosol layers. The corresponding lidar ratio is obtained using the fraction of the for the two CALIOP profiles selected for the analysis of the aged forest fire plume (less polluted dust fraction for the CALIOP profile least distant from the Baikal lake fire area).

390
Regarding the aerosol type found for the CALIOP observations, there is a good agreement with our detailed analysis of the aerosol sources for three cases (dusty mixture, fresh smoke and aged pollution plume from Northern China) while the classification is not appropriate for the aged Siberian fires from Lake Baikal region and for the urban and gas flaring emissions from Russia. For the aged forest fire case and the gas flaring emissions, the fraction of polluted dust aerosol is too high. These two misclassifications led to moderate underestimates of the lidar ratio of 5 sr and 7 sr respectively for the case of old fires and 395 flaring gas emissions. In fact the misclassification even compensates for the bias on the value of the lidar ratio of the polluted continental type which is too high for gas flaring emissions. The misclassification of the CALIOP profile related to urban and industrial Russian emissions seen as an elevated smoke type occure mainly because the polluted PBL thickness is frequently higher than 2.5 km in summer above Siberia (i.e. the upper limit to ascribe an elevated smoke type to an aerosol layer in the CALIOP V.4). This misclassification does not impact the value of the lidar ratio.

400
Eventhough the aerosol type classification is correct for the dusty mixture and the transport of pollution plume from Northern China, the lidar ratio found for CALIOP is too high by 25 sr. Although this difference is not far from the CALIOP lidar ratio uncertainty, it is likely that aerosol plume aging and mixture with background aerosol cannot be properly taken into account and lead to a positive bias when deriving the lidar ratio from the aerosol type. Regarding the fresh forest fire case, CALIOP classification is correct but since the age of the fire is not taken into account in the CALIOP data processing the lidar ratio 405 is two times larger than the estimated value by the airborne lidar. This will lead to an overestimation of the aerosol AOD by CALIOP when sampling biomass burning plume very close to the fire region.

Conclusions
Two airborne lidar campaigns were carried out over Siberia in July 2013 and June 2017. Aerosol types and optical properties were derived using FLEXPART, satellite data and airborne in-situ measurements when available. Six aerosol type could be 410 identified in this work: (i) Dusty aerosol mixture (ii) Ob valley industrial emission (iii) fresh boreal forest fire plumes (iv) aged forest fire plumes (v) pollution over the Tomsk/Novosibirsk region (vi) long range transport of Chinese pollution over Yakutsk.
The aircraft in-situ measurement, mainly ∆CO and BC have been useful to validate the identification of the aerosol origin using FLEXPART and the satellite observations ; namely large BC concentrations in the fresh forest fire plume and large ∆CO for the long range transport of Eastern Siberian forest fires and of polluted plumes from Northern China. The lidar ratio (LR) 415 analysis shows that the lowest LR range is obtained for the "Dusty Mix" case (26-40 sr) and the highest for the urban and industrial pollution from the Tomsk/Novosibirsk area (71-90 sr). We found a good agreement of this work analysis of the LR values according to the aerosol classification with previous studies (e.g., Burton et al., 2013). The range of lidar ratio obtained for gas flaring emission (43-60 sr) is lower than the high values encountered in the Tomsk/Novosibirk urban area and has never been characterized using lidar observations. 420 Airborne lidar backscatter ratio vertical structure, aerosol types and integrated LR derived from the airborne data analysis (section 4) were compared to nearby CALIOP overpasses. We found three main differences with the CALIOP LR and aerosol type classification over Siberia: (i) layer can be classified as Elevated smoke instead of Polluted continental and vice versa, but with little influence on the LR value (ii) aging and transport of aerosol layers effect on the LR value is not always properly accounted for even when the classification is correct (e.g. the dusty mixture is properly identified but with a lidar ratio too high)

425
(iii) the lack of discrimination between fresh and old fire plume leads to an overestimation of the optical depth for the fresh fires. Constrained LR CALIOP with an independent AOD could be another alternative to alleviate some of these limitations discussed in this paper. Such an independent AOD value could be given by co-localized MODIS observations especially for daytime observations in summer. Surface lidar reflectance observations on homogeneous surfaces such as the Siberian taiga or artic tundras could be also a very good alternative as discussed by (Josset et al., 2018) especially for nighttime observations.