UTLS wildfire smoke over the North Pole region, Arctic haze, and aerosol-cloud interaction during MOSAiC 2019/20: An introductory

An advanced multiwavelength polarization Raman lidar was operated aboard the icebreaker Polarstern during the MOSAiC (Multidisciplinary drifting Observatory for the Study of Arctic Climate) expedition, lasting from September 2019 to October 2020, to contiuously monitor aerosol and cloud layers in the Central Arctic up to 30 km height at latitudes mostly between 85◦N and 88.5◦N. The lidar was integrated in a complex remote sensing infrastructure aboard Polarstern. Modern aerosol lidar 5 methods and new lidar techniques and concepts to explore aerosol-cloud interaction were applied for the first time in the Central Arctic. The aim of this introductory article is to provide an overview of the observational spectrum of the lidar products for representative measurement cases. The highlight of the lidar measurements was the detection of a 10 km deep wildfire smoke layer over the North Pole area from, on average, 7 km to 17 km height with an aerosol optical thickness (AOT) at 532 nm around 0.1 (in October-November 2019) and 0.05 from December to mid of March 2020. The wildfire smoke was trapped 10 within the extraordinarily strong polar vortex and remained detectable until the beginning of May 2020. Arctic haze was also monitored and characterized in terms of backscatter, extinction, and extinction-to-backscatter ratio at 355 and 532 nm. High lidar ratios from 60-100 sr in lofted mixed haze and smoke plumes are indicative for the presence of strongly light-absorbing fine-mode particles. The AOT at 532 nm was of the order of 0.025 for the tropospheric haze layers. In addition, so-called cloud closure experiments were applied to Arctic mixed-phase cloud and cirrus observations. The good match between cloud 15 condensation nucleus concentration (CCNC) and cloud droplet number concentration (CDNC) and, on the other hand, between ice-nucleating particle concentration (INPC) and ice crystal number concentration (ICNC) indicated a clear influence of aerosol particles on the evolution of the cloud systems. CDNC was mostly between 20 and 100 cm−3 in the liquid-water dominated cloud top layer. ICNC was of the order of 0.1-1 L−1. The study of the impact of wildfire smoke particles on cirrus formation revealed that heterogeneous ice formation with smoke particles (organic aerosol particles) as INPs may have prevailed. ICNC 20 values of 10-40 L−1 were clearly below ICNC levels that would indicate homogeneous freezing. 1 https://doi.org/10.5194/acp-2020-1271 Preprint. Discussion started: 29 December 2020 c © Author(s) 2020. CC BY 4.0 License.

radiative transfer studies. The lidar ratio of aerosol particles is an important quantity in aerosol typing efforts because it is sensitive to the ratio of absorption to scattering by the particles (Müller et al., 2007) and is low for marine particles (around 15-20 sr) and can be very high for light-absorbing particles (100-120 sr) (Ohneiser et al., 2020). The linear depolarization ratio is defined as the cross-polarized-to-co-polarized backscatter ratio and allows us to sensitively distinguish spherical particles showing particle linear depolarization ratios (PLDR) close to zero from non-spherical aerosol particles showing PLDR values 5 of typically 0.1-0.3. In the case of clouds, liquid-droplet layers show PLDR≈ 0 at layer base where light depolarizing multiple scattering is low, and PLDR of 0.4-0.6 in, e.g., cirrus layers. "Co" and "cross" denote the planes of polarization parallel and orthogonal to the plane of linear polarization of the transmitted laser pulses, respectively.
The retrieval of aerosol microphysical properties such as particle volume, mass, and surface area concentration and estimates of cloud-relevant properties (aerosol-type-dependent cloud condensation nuclei, CCN, and ice-nucleating particles, INPs) is 10 performed by means of the POLIPHON (Polarization Lidar Photometer Networking) approach (Mamouri andAnsmann, 2016, 2017;Ansmann et al., 2019Ansmann et al., , 2020. Hofer et al. (2020) exemplary shows the full set of POLIPHON aerosol products in the cases of an 18-month Polly campaign in Dushanbe, Tajikistan, for central Asian aerosol. Alternatively to the POLIPHON method, we used the multiwavelength lidar inversion technique (Müller et al., 1999(Müller et al., , 2014Veselovskii et al., 2002Veselovskii et al., , 2012 to derive microphysical properties of aerosols including the particle size distribution for detected pronounced aerosol layers. Details of 15 the retrieval of microphysical properties of liquid-water cloud layers can be found in Jimenez et al. (2020a, b). Regarding the determination of height profiles of the water-vapor-to-dry-air mixing ratio from the Raman lidar observations of water vapor and nitrogen vibrational Raman signals we follow the procedure described by Dai et al. (2018). Relative humidity is obtained from the mixing-ratio measurements and temperature profiles measured regularly (four times a day) aboard Polarstern. Quality checks of the continuously obtained water vapor fields is based on comparisons with the relative-humidity profiles obtained 20 with radiosondes.
Radiosonde temperature and pressure profiles are used in the lidar data analysis to correct for Rayleigh backscattering and attenuation effects on the measured lidar signal profiles. Temporal averaging and vertical smoothing of signal profiles were applied to reduce the impact of signal noise. Corrections of the incomplete overlap between laser beam and receiver FOVs in the lidar near-range are needed to extend the observational range to very low heights above the lidar, down to 120 m. 25

Observations
We begin with a few impressions (snapshots) of typical aerosol and cloud scenes observed with our lidar during the winter months. According to Fig. 1, the Polarstern moved very slowly with the pack ice in December, January, and February and was mostly located between 86 • N and 88 • N. The exceptionally strong polar vortex of 2019-2020 was well established to that time. Figure 3 shows a 10-day measurement sequence (2-12 Dec 2019). Complex features of aerosol layering, cirrus evolution The measured linear depolarization ratio in the right panels of Fig. 3 allows us to precisely distinguish cirrus from layered mixed-phase clouds as explained above. Ice crystals cause large depolarization ratios (green to red colors in Fig. 3b, d, f), and, in contrast, liquid-water layers produce rather low depolarization ratios around zero in Fig. 3f. The increase of the depolarization values with increasing penetration of the laser beam into the water cloud layer is caused by multiple scattering by the cloud droplets. This aspect is explained in more detail in Sect. 3.3. The depolarization ratio of aerosol particles was found to be 5 generally small (Fig. 3h) in the free troposphere and stratosphere which indicates spherical haze and smoke particles.

Wildfire smoke layer in the UTLS regime
According to Stohl (2006) the aerosol conditions in the Arctic troposphere up to UTLS heights are controlled by aerosol long-range transport. Stohl (2006) identified three major pathways. The winter transport of aerosols (regime 1), mainly from Asian and European sources towards the high Arctic occurs at low heights and long-transported aerosols may reach heights 10 up to the middle troposphere (5-7 km height). In summer, warm, moist and polluted air massed mainly originating from North America and Asia are forced to ascend during long range transport and thus reach the colder Arctic at heights in the middle and upper troposphere. Two mechanisms are then active. Either, uplift of air masses occurs at the Arctic front after low-level northward transport (regime 2) or the mid-latitude air masses are lifted over the source regions (such as wild fire smoke) followed by high-altitude transport in northerly directions (regime 3). All this is described by Stohl (2006) . A fourth mechanism (quite similar to the third one) became relevant during the last few years and is characterized by a rather fast ascent of wildfire smoke up to the tropopause and occasionally into the lower stratosphere via pyro-cumulonimbus (pyro-Cb) convection over areas with intense and long-lasting fires (Fromm and Servranckx, 2003;Fromm et al., 2010;Peterson et al., 2018;Khaykin et al., 2018;Ansmann et al., 2018;Hu et al., 2019;Zuev et al., 2019), immediately followed by further ascent due to self-lifting processes caused by absorption of solar radiation 20 and heating of the smoke-containing air layers (Boers et al., 2010;de Laat et al., 2012;Torres et al., 2020;Ohneiser et al., 2020;Kablick et al., 2020;Khaykin et al., 2020). The light-absorption-related lifting occurs during the spread of the smoke over the respective hemisphere and continues as long as the smoke layers are optically dense enough (aerosol optical thickness AOT>1-2 at 500 nm) with the consequence that the smoke reaches, e.g., the Central Arctic at heights up to 5-10 km above the tropopause. Law and Stohl (2007) already mentioned the potentially strong impact of pyro-Cb activity on the vertical smoke 25 distribution, but to that time this transport pathway was of minor importance. This fourth mechanism is responsible for the   Fig. 3g and h). The smoke 30 layer extended from 8 to more than 18 km height. The internal vertical structures were rather smooth and indicate an aged smoke layer. A clear indication and striking feature for the dominance of smoke particles is the observed spectral dependence of the extinction-to-backscatter ratio (Fig. 4d). The 532 nm lidar ratio is much larger than the 355 nm lidar ratio. This is an unambiguous indication for the presence of aged biomass burning smoke. No other aerosol type (or cloud type) produces an inverse spectral behavior in terms of the particle lidar ratio (Müller et al., 2005;Haarig et al., 2018;Ohneiser et al., 2020).
The unusual spectral behavior of the lidar ratio is related to the strong wavelength dependence of the backscatter coefficient and the weak spectral dependence of the extinction coefficient ( Fig. 4a and c). These specific optical properties are linked to the narrow size distribution of absorbing smoke particles which form a well-defined accumulation mode as shown in Fig. 5. The 5 size distributions of the smoke particles were obtained from the Polly observation by applying the lidar inversion method to the layer-mean backscatter and extinction information (Veselovskii et al., 2012). In agreement with many in-situ observations of long-transported aged smoke (Fiebig et al., 2003;Petzold et al., 2007;Dahlkötter et al., 2014), an accumulation mode with a comparably large mode radius of 250-300 nm (for the volume size distribution) was found in October and November 2019.
A distinct coarse mode was absent.

10
As already mentioned, the particle depolarization ratios at 532 nm (around 1%) and at 355 nm (1-2%) were very low and indicated spherical particles (Fig. 4b). Assuming a core-shell structure of the smoke particles, slight deviations from the spherical shape (probably caused by an irregular black-carbon-containing core structure) can lead to depolarization ratios of up to 0.2 at 532 nm (Haarig et al., 2018;Gialitaki et al., 2020;Ohneiser et al., 2020). With increasing age the core structure obviously collapses, gets compact, and the particles become more and more spherical with time (Baars et al., 2019).

15
An overview of the smoke conditions during the MOSAiC winter half year (October to April) is presented in Figure 6.
Usually, one observation per day is considered. Gaps in the observations are caused by long-lasting low level fog and lowcloud events. The maximum smoke concentration was found around or just below the tropopause (Fig. 6a). In the computation of the tropopause height from the radiosonde data we used a refined version of the World Meteorological Organization approach (WMO, 1992;Klehr, 2012). Most of the time, the smoke layer was observed between 7 and 17 km height. A trend of downward 20 motion is not visible. The maximum extinction coefficients (532 nm) around the tropopause slowly decreased with time from values >10 Mm −1 in October and November to <5 Mm −1 in April 2020 (Fig. 6a).
In terms of the aerosol optical thickness (AOT) for 532 nm, shown in Fig The layer-mean 532 nm smoke extinction coefficients in Fig. 6c (obtained from the ratio of AOT divided by the layer geometrical depth in Fig. 6a) were on the order of 10 Mm −1 in October, around 4-5 Mm −1 during the main winter months  Probably as a consequence of the missing aerosol information in the CALIPSO data base, the potential impact of the Arctic wildfire smoke on the strong ozone reduction (in the height range from 12-22 km height) in early spring 2020 remained completely unconsidered in the discussion of the reasons for the observed record-breaking ozone depletion (DeLand et al., 2020;Wohltmann et al., 2020;Innes et al., 2020;Manney et al., 2020). Only polar stratospheric clouds (PSC) were taken into account in the ozone-hole analysis. However, in contrast to PSCs, which developed from December to April (according  A PSC observation is shown in Fig. 7. Note that we corrected our stratospheric smoke observations in Fig. 6 for PSC effects. According to the PSC classification scheme (Achtert and Tesche, 2014), we observed a type Ib PSCs. This type is made up of supercooled liquid ternary solutions that consist of H 2 SO 4 , HNO 3 , and H 2 O. In contrast to type Ia and II PSC particles At the end of this section, we need to discuss a potential contribution of volcanic aerosol to the UTLS aerosol pollution.
Besides the record-breaking fires at high northern latitudes, a strong eruption of the Raikoke volcano in the Kuril Islands

Arctic haze vertical structures
The original and primary goal of the shipborne MOSAiC lidar measurements was to provide, for the first time, a seasonally and According to the backward trajectory analysis shown in Fig. 9 (HYSPLIT, 2020), the aerosol pollution in the pronounced aerosol layers around 5 km (4 February, 11 UTC, Fig. 8a) and 4 km height (4 March, 11 UTC, Fig. 8b) originated from central 30 and western European regions (4 February) and from Russia and the Black Sea area (4 March). However, a clear identification of the source region was impossible. The height-resolved trajectory analysis indicated that most of the aeorosol circled around in the Arctic (at latitudes > 70 • N) for more than a week before crossing the Polarstern.
In Fig. 10, the optical properties of Arctic haze for the two cases are illustrated. 12-hour (4 February) and 18-hour (4 March) mean height profiles of the basic lidar products (backscatter, extinction, extinction-to-backscatter ratio) are shown. In both cases, we found a near surface layer up to about 2.5 km height and a separated lofted aerosol layer up to 5 and 7 km height. A clear wavelength dependence of the backscatter and extinction coefficients was found on 4 March, as expected for fine-mode dominated particles in the Arctic (Quennehen et al., 2012) consisting of a mixture of anthropogenic haze and fire smoke (Wang 5 et al., 2011). The Ångström exponent for the extinction coefficient was around 1.7 in the lofted layer above 3 km height. The lidar ratios were high with values close to 100 sr in the lofted layer on 4 March. This is indicative of the presence of small, strongly light-absorbing particles (Müller et al., 2007).
Our findings are in agreement with the results previously published and obtained during major field activities such as POLARCAT-IPY (Polar Study using Aircraft, Remote Sensing, Surface Measurements, and Models, of Climate, Chemistry,  Agricultural fires during spring (Europe, Asia) and flaring of natural gas (Russian oil industry) are important sources for black carbon particles. Eurasian fires mix with anthropogenic haze mainly from Asia during the long-range transport towards the central Arctic, probably beginning in late winter and with peak occurrence in the spring season .
Agricultural and forest fires produce BC/OA ratios of typically <0.1 whereas anthropogenic haze may cause ratios >0.15. It was observed during the POLARCAT-IPY and ACTRAS that the impact of fire aerosol increases with height.

25
On 4 Februay 2020 (Fig. 10a), the aerosol optical properties were less well defined, the extinction coefficients were almost equal at both wavelengths (355 and 532 nm), and the noisy lidar ratios at 532 nm were larger than at 355 nm in the lofted layer above 3 km height which is typical for aged wildfire smoke as discussed in the foregoing section. The lidar ratios were lower than on 4 March, i.e., less absorbing particles were present. The volume depolarization ratios were rather low in both cases and indicated the dominance of aerosol pollution (spherical particles). The 532 nm AOT was 0.024 (4 February, for the lowest  It should be mentioned at the end that, in contrast to Polly, the CALIPSO lidar is a standard backscatter lidar and needs 5 to assume an appropriate lidar ratio in the conversion of the backscatter profiles into respective extinction coefficient profiles.
Lidar ratios are between 40-70 sr in the CALIPSO data analysis of Arctic observations. Thus, the MOSAiC Polly data analysis will provide valuable lidar-ratio information for CALIPSO lidar studies of Arctic aerosols. It is planned to analyze all favorable MOSAiC lidar observations without fog and low-level clouds to obtain a seasonally and height-resolved aerosol data set in terms of backscatter, extinction, lidar ratio and depolarization ratio profiles for the Central Arctic.

Mixed-phase cloud evolution in Arctic haze
The final two subsections give an overview of aerosol-cloud interaction studies that can be performed by using the MOSAiC remote sensing data base. We begin with a mixed-phase-cloud study and continue in Sect. 3.4 with a case of cirrus evolution in the smoke layer above 8 km height. Persistent mixed-phase clouds with rather constant thermodynamic, micro-, and macrophysical properties have been observed over hours, in few cases over even more than 40 hours, especially during the MOSAiC In order to better understand the role of aerosol particles in cloud evolution processes we need to combine methods that allow us to measure or retrieve the cloud-relevant aerosol parameters, i.e., the cloud condensation nucleus concentration (CCNC) and ice-nucleating particle concentration (INPC) in the aerosol around the developing cloud layer, together with the cloud droplet number concentration (CDNC) and ice crystal number concentration (ICNC) in the observed cloud layers. In the following, we 25 introduce two recently developed concepts to investigate aerosol-cloud interaction solely based on active remote sensing and explain how we derive the required information of CCNC, INPC, CDNC, and ICNC.
The mixed-phase cloud system, studied here, was shown in Fig. 3e-f. The altocumulus layer was present over the Polarstern for nine hours. Favorable conditions with cloud top temperatures around −28.5 • C at 2.6 km height (at 03:00 UTC) were given for heterogeneous ice formation via immersion freezing, i.e., ice nucleation on INPs immersed in the water droplets. After The reasons for the longevity of the shallow cloud top layer mostly consisting of liquid-water droplets are continuously occurring updrafts which permanently cause water supersaturation in the liquid-water layer and droplet nucleation (Ansmann et al., 2009) so that the liquid-water cloud layer can never become completely glaciated. As discussed below in detail, there were always 20-200 droplets per cm 3 in the altocumulus top layer, but only 0.1 to 1 ice crystals per liter, i.e., the droplet-tocrystal number ratio was of the order of 20000 to 200000 and thus too high to convert the liquid water layer into an ice cloud 5 by heterogeneous ice nucleation and riming processes within seconds to minutes before the next updraft led to new droplet formation.
We analyzed the dual-FOV polarization lidar observations of the liquid-water cloud layer. The results are shown in Fig. 11.
The applied dual-FOV polarization lidar technique was recently introduced (Jimenez et al., 2020a, b) and successfully tested in the case of water cloud experiments in southern Chile. The new method was originally designed for pure liquid-water 10 cloud observations but can be applied to mixed-phase clouds as long as backscattering by ice crystals is negligible compared to droplet backscattering in the droplet-dominated cloud top layer. This condition holds here with ice crystal backscatter coefficients of 5-10 Mm −1 sr −1 in the virga and thus also in the cloud top layer and droplet backscatter coefficients of the order of 700 Mm −1 sr −1 .
As can be seen in seen Fig. 11, the CDNC values were about 20 cm −3 in the beginning and around 100 cm −3 later on.

15
Obviously, updraft velocity was weak and correspondingly water super saturation levels were below 0.2% so that fewer particles were activated to become cloud droplets as predicted (CCNC values in Fig. 12). Later on, the updrafts became obviously stronger, and supersaturation levels exceeded 0.2% so that more CCN nucleated cloud droplets as predicted by the retrieved CCNC values. With increasing CDNC the effective radius (characterizing the typical droplet size) decreased and vice versa for constant water vapor conditions. The cloud extinction coefficient showed typical values from 10-20 km −1 most of the time in 20 the droplet-dominated cloud layer.
CCNC was estimated from the particle extinction coefficient profile measured with lidar before the altocumulus appeared over the lidar site. The particle extinction profile is shown in Fig. 12 All obtained numbers for the aerosol and cloud particle number concentrations are shown in Fig. 13 together with the relativehumidity field obtained from the Raman lidar water-vapor mixing ratio observations (combined with radiosonde temperature measurements). The relative-humidity height time display in Fig. 13b indicates that the altocumulus layer developed in a moist 15 air mass and was present over the Polarstern until a dry air mass approached, leading to a strong decrease in relative humidity and dissolution of the stratiform cloud deck. As a result of strong ice crystal evaporation in the virga zone, the relative humidity increased strongly below the liquid-water cloud deck after 3:00 UTC and caused reduced crystal evaporation later on in the virga zone so that ice crystals could partly reached the ground as precipitation.
The good match between CCNC and CDNC (liquid-water cloud closure) and between INPC and ICNC (ice cloud closure, 20 see numbers in Fig. 13a) during the early phase of the altocumulus development indicates that the aerosol particles controlled the cloud properties and thus had a strong influence of the evolution of the observed altocumulus cloud system as long as the humidity conditions were favorable. It should be emphasized that such a closure study with consistent findings is only possible if primary ice and droplet nucleation dominates and secondary ice formation, ice breakup processes, crystal-crystal collision and aggregation processes, as well as droplet collision and coagulation, and strong mixing and entrainment processes 25 are absent. As part of the MOSAiC data analysis, we plan to analyze many winter as well as summer altocumulus events and thus cloud formation under contrasting aerosol conditions to obtain an improved view on the role of long-transported aged aerosol pollution on the evolution of mixed-phase clouds in the high Arctic.

Cirrus evolution in the UTLS wildfire smoke layer
MOSAiC offers the unique opportunity to learn more about the role of wildfire smoke layers on the evolution of Arctic Complications in our understanding arises due to the fact that OA particles can occur as glassy, semi-glassy and liquid particles as a function of temperature and relative humidity, and thus can trigger deposition nucleation (as glassy particles), 5 immersion freezing (as semi-glassy particles), and homogeneous freezing (when liquid) (Koop et al., 2000(Koop et al., , 2011Berkemeier et al., 2014;Wang et al., 2012;Knopf et al., 2018). Aerosol particles serving as INPs usually provide an insoluble, solid surface that can facilitate the freezing of water. Deposition ice nucleation is defined as ice formation occurring on the INP surface by water vapor deposition from the supersaturated gas phase. When the supercooled smoke particle takes up water or its shell deliquesces, immersion freezing can proceed, where the INP immersed in an aqueous solution can initiate freezing. Finally, 10 if the smoke particle become completely liquid (and contain no solid soot fragments), homogeneous freezing will take place at temperatures below 235 K. However, in reality, at given air mass lifting conditions, the ice nucleation process can be very complex. The time that solid organic material (OM) needs for transition to a more liquid state, termed as humidity-induced amorphous deliquescence, can range from several minutes to days at temperatures low enough for ice formation (Mikhailov et al., 2009;Berkemeier et al., 2014;Knopf et al., 2018). Thus the phase change (as function of T and RH) can be longer than 15 typical cloud activation time periods (governed by the updraft velocity), potentially inhibiting full deliquescence and allowing the OA or the organic coating to serve as INP. When amorphous OA or OM are involved in ice nucleation, the condensed-phase diffusion processes within OA particles will most probably govern the ice nucleation pathway (Wang et al., 2012).
To demonstrate that the observed wildfire smoke particle were able to control cirrus evolution and life time we present the results of a first MOSAiC case study here. The observation is from 6 December 2019 ( Fig. 3c and d). HYSPLIT backward 20 trajectory analysis (for arrival heights at 8-9.5 km and arrival times of 6, 10, and 14 UTC) indicate air mass transport within the Arctic at latitudes > 70 • N and heights between 7-10 km during the last seven days before crossing the MOSAiC field site (86 • N, 122 • E). The air masses originated from the remote northern Pacific, and transported water vapor and probably unpolluted marine air. The ascent of the moist air masses occurred 8-10 days before arrival in the Central Arctic. During the 7-day travel in the Arctic the Pacific airmass mixed with the smoke above 7 km. These smoke particles then served as ice nuclei The full cirrus lifetime was about 36 hours. The wildfire smoke is best visible as a light blue layer above 9.3 km height, i.e., above the main cirrus layer, between 6:00 and 12:00 UTC in Fig. 3c. This part of the smoke layer (above 9.3 km) can be regarded as the main reservoir of INPs. The bow-like feature of the lower boundary of the extended virga zone (between 30 00:00 to 18:00 UTC) was caused by a high pressure ridge (with dome-like temporal evolution of the influenced height range).
The ridge crossed RV Polarstern during this day and was characterized by a warmer and very dry air mass in the lowest few kilometers of the troposphere. After entering this dry air mass the ice crystals immediately evaporated.
The data analysis with respect to ice nucleation on smoke particles is explained in Fig. 14 and 15. We applied the wateractivity-based immersion freezing model (ABIFM) that allows predicting of immersion freezing under cirrus conditions (Knopf and Alpert, 2013). In addition, we used the parameterization for deposition ice nucleation (DIN) of Wang and Knopf (2011).
Ice nucleation from the water vapor phase is denoted as deposition nucleation. Figure 14a shows the cirrus layers observed from 6-12 UTC (red) and 12:30-13:30 UTC (blue) on 6 December 2019 and the temperature and relative humidity profiles of radiosondes launched aboard Polarstern at 5, 11 and 17 UTC on this day (Fig. 14b).
Whereas the parameterization for DIN uses ambient temperature and humidity (i.e., ice supersaturation S i , see Fig. 14c) 5 to derive a nucleation rate coefficient that is used to calculate INPC, ABIFM requires the derivation of the water-activity criterion ∆a w (Koop et al., 2000). ∆a w describes the difference between the ice melting conditions (melting temperature) and the observed freezing conditions in terms of temperature and humidity. Homogeneous ice nucleation is characterized by Also in the case of DIN, laboratory results for Leonardite are used (Wang and Knopf, 2011). Ice supersaturation conditions are usually given or produced during updrafts (e.g., during the ascending period of a gravity wave) that could, in principle, be detected and measured with the AMF-1 Doppler radar.

15
It is noteworthy to mention here, that equilibrium at ice supersaturation conditions as observed in the extended virga zone over the whole day (Fig. 14c) is a sign for a low crystal number concentration (<35 L −1 ) (Murray et al., 2010). Such a low amount of ice crystals is not able to quench the supersaturation which is in turn indicative for heterogeneous ice nucleation.
Homogeneous freezing would produce crystal concentrations of >500 L −1 so that equilibrium at ice saturation level would occur within a short time period. 20 Figure 15 provides an overview of all retrieval products necessary to evaluate the potential of smoke particles to serve as INP. We follow here the strategy of the ICNC-INPC closure study (Ansmann et al., 2019). The lidar-radar combination (with TROPOS lidar and the 35-GHz AMF-1 cloud radar KAZR) delivers ICNC values (n ICE ) (Bühl et al., 2019) as given in Fig. 15a and c. The best and most trustworthy radar observations are usually taken in the lower part of the cloud deck where ice crystals are large and thus radar reflectivity is usually very strong and accurate. However, we have to keep in mind that these ICNC 25 values may be lower than the ones at cloud top caused by crystal-crystal collision and aggregation processes (Mitchell et al., 2018). Other ICNC influencing effects, e.g., secondary ice formation can be ignored in the case of cold cirrus clouds (Field et al., 2006;Korolev and Leisner, 2020). From the aerosol lidar observations (at cloud free conditions, for example on 2 and 7 December) we can derive the range (or amount) of available smoke particles and related particle surface area s (in Fig. 15b, upper axis). In this figure, the number concentration of large smoke particles n 250 (with radii > 250 nm, lower axis) is shown  Fig. 15c. 600 s may represent here a typical time period of the lifting phase of a gravity wave. As can be seen the n ICE and n INP,I values (blue and red bars) are in the same range of values which suggests that organic particles may be able to control the evolution of the cirrus layer via the immersion freezing mode (i.e., when the smoke particles have a liquid shell). The impact of deposition INP n INP,D (cyan and orange bars) is comparably weak in this example. We assume here that smoke 5 particles have a solid core (containing black carbon) and a shell of organic material (here Leonardite is assumed) which can liquefy at certain temperature and humidity conditions. The deposition-nucleation and immersion-freezing ability depends on the surface material of the smoke particles. In the MOSAiC data analysis later on we will consider different organic substances (more details in Knopf and Alpert (2013)). It has been shown that humic and fulvic matter can act as deposition nucleation and immersion freezing INPs (Wang and Knopf, 2011;Rigg et al., 2013;Knopf and Alpert, 2013;Knopf et al., 2018). Furthermore, 10 these macromolecules can undergo amorphous phase transition under typical tropospheric conditions (Wang et al., 2012;Slade et al., 2017) similar to the processes we assume the organic coating of the smoke particles experiences.
To sum up, heterogeneous ice nucleation is a complex process, especially in the case of organic aerosol particles. However, the successful closure, indicated by a reasonable match between n INP,I and n ICE , indicates that the wildfire smoke was able to trigger cirrus formation (before homogeneous freezing can take place on stratospheric background or even liquid smoke 15 particles) and control of the further evolution of the ice cloud system. It is clear that many more closure studies are needed to obtain a statistically trustworthy view on smoke and the role in cirrus nucleation. The respective extended data analysis will be part of our MOSAiC data analysis.

Conclusion and outlook
The goal of this introductory article was to provide an overview of the capabilities of modern lidar methods to contribute The highlight of our observations was the detection of the long lasting UTLS wildfire smoke layer which was present over the North Pole region for seven month until spring. More details to the smoke layer and the potential impact on the record-breaking ozone reduction is given by Ohneiser et al. (2021). Besides the smoke, we presented two days with typical Artcis haze layering 25 features and properties. The results agree well with foregoing studies, e.g., performed in the framework of POLARCAT-IPY and ARCTAS.