Beschreibung: 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 to continuously monitor aerosol and cloud layers in the central Arctic up to 30 km height. The expedition lasted from September 2019 to October 2020 and measurements were mostly taken between 85 and 88.5∘ N. The lidar was integrated into a complex remote-sensing infrastructure aboard the Polarstern. In this article, novel lidar techniques, innovative concepts to study aerosol–cloud interaction in the Arctic, and unique MOSAiC findings will be presented. The highlight of the lidar measurements was the detection of a 10 km deep wildfire smoke layer over the North Pole region between 7–8 km and 17–18 km height with an aerosol optical thickness (AOT) at 532 nm of around 0.1 (in October–November 2019) and 0.05 from December to March. The dual-wavelength Raman lidar technique allowed us to unambiguously identify smoke as the dominating aerosol type in the aerosol layer in the upper troposphere and lower stratosphere (UTLS). An additional contribution to the 532 nm AOT by volcanic sulfate aerosol (Raikoke eruption) was estimated to always be lower than 15 %. The optical and microphysical properties of the UTLS smoke layer are presented in an accompanying paper (Ohneiser et al., 2021). This smoke event offered the unique opportunity to study the influence of organic aerosol particles (serving as ice-nucleating particles, INPs) on cirrus formation in the upper troposphere. An example of a closure study is presented to explain our concept of investigating aerosol–cloud interaction in this field. The smoke particles were obviously able to control the evolution of the cirrus system and caused low ice crystal number concentration. After the discussion of two typical Arctic haze events, we present a case study of the evolution of a long-lasting mixed-phase cloud layer embedded in Arctic haze in the free troposphere. The recently introduced dual-field-of-view polarization lidar technique was applied, for the first time, to mixed-phase cloud observations in order to determine the microphysical properties of the water droplets. The mixed-phase cloud closure experiment (based on combined lidar and radar observations) indicated that the observed aerosol levels controlled the number concentrations of nucleated droplets and ice crystals.

Beschreibung: The MOSAiC (Multidisciplinary drifting Observatory for the Study of Arctic Climate) expedition was the largest Arctic field campaign ever conducted. MOSAiC offered the unique opportunity to monitor and characterize aerosols and clouds with high vertical resolution up to 30 km height at latitudes from 80 to 90 N over an entire year (October 2019 to September 2020). Without a clear knowledge of the complex aerosol layering, vertical structures, and dominant aerosol types and their impact on cloud formation, a full understanding of the meteorological processes in the Arctic, and thus advanced climate change research, is impossible. Widespread ground-based in situ observations in the Arctic are insufficient to provide these required aerosol and cloud data. In this article, a summary of our MOSAiC observations of tropospheric aerosol profiles with a state-of-the-art multiwavelength polarization Raman lidar aboard the icebreaker Polarstern is presented. Particle optical properties, i.e., light-extinction profiles and aerosol optical thickness (AOT), and estimates of cloud-relevant aerosol properties such as the number concentration of cloud condensation nuclei (CCN) and ice-nucleating particles (INPs) are discussed, separately for the lowest part of the troposphere (atmospheric boundary layer, ABL), within the lower free troposphere (around 2000 m height), and at the cirrus level close to the tropopause. In situ observations of the particle number concentration and INPs aboard Polarstern are included in the study. A strong decrease in the aerosol amount with height in winter and moderate vertical variations in summer were observed in terms of the particle extinction coefficient. The 532 nm light-extinction values dropped from 〉50 Mm-1 close to the surface to 〈5 Mm-1 at 4-6 km height in the winter months. Lofted, aged wildfire smoke layers caused a re-increase in the aerosol concentration towards the tropopause. In summer (June to August 2020), much lower particle extinction coefficients, frequently as low as 1-5 Mm-1, were observed in the ABL. Aerosol removal, controlled by in-cloud and below-cloud scavenging processes (widely suppressed in winter and very efficient in summer) in the lowermost 1-2 km of the atmosphere, seems to be the main reason for the strong differences between winter and summer aerosol conditions. A complete annual cycle of the AOT in the central Arctic could be measured. This is a valuable addition to the summertime observations with the sun photometers of the Arctic Aerosol Robotic Network (AERONET). In line with the pronounced annual cycle in the aerosol optical properties, typical CCN number concentrations (0.2 % supersaturation level) ranged from 50-500 cm-3 in winter to 10-100 cm-3 in summer in the ABL. In the lower free troposphere (at 2000 m), however, the CCN level was roughly constant throughout the year, with values mostly from 30 to 100 cm-3. A strong contrast between winter and summer was also given in terms of ABL INPs which control ice production in low-level clouds. While soil dust (from surrounding continents) is probably the main INP type during the autumn, winter, and spring months, local sea spray aerosol (with a biogenic aerosol component) seems to dominate the ice nucleation in the ABL during the summer months (June-August). The strong winter vs. summer contrast in the INP number concentration by roughly 2-3 orders of magnitude in the lower troposphere is, however, mainly caused by the strong cloud temperature contrast. A unique event of the MOSAiC expedition was the occurrence of a long-lasting wildfire smoke layer in the upper troposphere and lower stratosphere. Our observations suggest that the smoke particles frequently triggered cirrus formation close to the tropopause from October 2019 to May 2020.

Beschreibung: The dataset contains daily nc-files of the Cloudnet liquid water content during Polarstern cruise PS106. The data is retrieved using the instrument synergystic approach Cloudnet (Illingworth, 2007, doi:10.1175/BAMS-88-6-883). This variable was calculated for the profiles where the \"categorization\" data has diagnosed that liquid water is present and liquid water path is available from a coincident microwave radiometer. The model temperature and pressure were used to estimate the theoretical adiabatic liquid water content gradient for each cloud base and the adiabatic liquid water content is then scaled so that its integral matches the radiometer measurement so that the liquid water content now follows a quasi-adiabatic profile. If the liquid layer is detected by the lidar only, there is the potential for cloud top height to be underestimated and so if the adiabatic integrated liquid water content is less than that measured by the microwave radiometer, the cloud top is extended until the adiabatic integrated liquid water content agrees with the value measured by the microwave radiometer. Missing values indicate that either 1) a liquid water layer was diagnosed but no microwave radiometer data was available, 2) a liquid water layer was diagnosed but the microwave radiometer data was unreliable; this may be because a melting layer was present in the profile, or because the retrieved lwp was unphysical (values of zero are not uncommon for thin supercooled liquid layers), or 3) that rain is present in the profile and therefore, the vertical extent of liquid layers is difficult to ascertain.

Beschreibung: The dataset contains daily nc-files of the Cloudnet liqiud particles effective radius during Polarstern cruise PS106. The data is retrieved using the instrument synergystic approach Cloudnet (Illingworth, 2007, doi:10.1175/BAMS-88-6-883). This variable was calculated for the profiles where the \"categorization\" data has diagnosed that liquid water is present the cloud droplet effective radius is calculated after Frisch et al 2002 (doi:10.1175/1520-0426(2002)019〈0835:TROSCD〉2.0.CO;2), relating Z with r_eff by assuming a lognormal size distribution, and its width. The number concentration required to represent the size distribution is derived by scaling the LWP measured with microwave radiometer over the observed (single) cloud layer.

Beschreibung: The dataset contains daily nc-files of the Cloudnet ice water content during Polarstern cruise PS106 based on: - 35.5 GHz Cloud radar MIRA. The data is retrieved using the instrument synergystic approach Cloudnet (Illingworth, 2007, doi:10.1175/BAMS-88-6-883). This variable was calculated from the 35.5-GHz radar reflectivity factor after correction for gaseous attenuation, and temperature taken from a forecast model, using the following empirical formula: log10(iwc[g m-3]) = 0.000242Z[dBZ]T[degC] + 0.0699Z[dBZ] + -0.0186T[degC] + -1.63 (Hogan, 2006, doi:10.1175/JAM2340.1). In this formula Z is taken to be defined such that all frequencies of radar would measure the same Z in Rayleigh scattering ice. However, the radar is more likely to have been calibrated such that all frequencies would measure the same Z in Rayleigh scattering liquid cloud at 0 degrees C. The measured Z is therefore multiplied by |K(liquid,0degC,35GHz)|^2/0.93 = 0.9441 before applying this formula. The formula has been used where the \"categorization\" data has diagnosed that the radar echo is due to ice, but note that in some cases supercooled drizzle will erroneously be identified as ice. Missing data indicates either that ice cloud was present but it was only detected by the lidar so its ice water content could not be estimated, or that there was rain below the ice associated with uncertain attenuation of the reflectivities in the ice. Note that where microwave radiometer liquid water path was available it was used to correct the radar for liquid attenuation when liquid cloud occurred below the ice; this is indicated a value of 3 in the iwc_retrieval_status variable. There is some uncertainty in thisprodedure which is reflected by an increase in the associated values in the iwc_error variable. When microwave radiometer data were not available and liquid cloud occurred below the ice, the retrieval was still performed but its reliability is questionable due to the uncorrected liquid water attenuation. This is indicated by a value of 2 in the iwc_retrieval_status variable, and an increase in the value of the iwc_error variable.

Beschreibung: The dataset contains daily nc-files of the attenuated backscatter coefficient and volume depolarization from the OCEANET multiwavelength Raman and polarization lidar PollyXT (Engelmann et al., 2016, doi:10.5194/amt-9-1767-2016; Baars et al., 2017, doi:10.5194/amt-10-3175-2017) during Polarstern cruise PS106. The data is calibrated within the Cloudnet (Illingworth, 2007 doi:10.1175/BAMS-88-6-883) processing scheme.

Beschreibung: The dataset contains daily nc-files of the Cloudnet ice particles effective radius during Polarstern cruise PS106 based on: - 35.5 GHz Cloud radar MIRA. The data is retrieved using the instrument synergystic approach Cloudnet (Illingworth, 2007, doi:10.1175/BAMS-88-6-883). This variable was calculated from visible extinction coefficient and ice water content using the following analytical formula: r_eff[\\mum] = 3*IWC[g cm-3] / (2 rho_i \\alpha[m-1]) 10**6 with rho_i=0.917g cm-3 the desity of solid ice. This retrieval is based on an analytical relationship between IWC and \\alpha. However both parameters have been retrieved following Hogan 2006 (doi:10.1175/JAM2340.1) by using two different empirical formulas based on the same measured quantity, the 35.5-GHz radar reflectivity factor.

Beschreibung: The dataset contains daily nc-files of the Cloudnet target categorization during Polarstern cruise PS106. The data is retrieved using the instrument synergystic approach Cloudnet (Illingworth, 2007 doi:10.1175/BAMS-88-6-883 ). This dataset is an aggregation of data from cloud radar, lidar, a numerical forecast model and optionally a rain gauge and microwave radiometer. It is intended to facilitate the application of synergistic cloud-retrieval algorithms by performing a number of the preprocessing tasks that are common to these algorithms. Each of the observational datasets has been interpolated onto the same grid, although the model data are provided on a reduced height grid. Radar reflectivity has been corrected for attenuation, where possible, and two additional fields have been added: \"category_bits\" contains a categorization of the targets in each pixel and \"quality_bits\" indicates the quality of the data at each pixel. Finally, estimates of the random and systematic errors in reflectivity factor and attenuated backscatter are provided.

Beschreibung: The dataset contains daily nc-files of the Cloudnet liquid water content during Polarstern cruise PS106. The data is retrieved using the instrument synergystic approach Cloudnet (Illingworth, 2007 doi:10.1175/BAMS-88-6-883). This variable was calculated for the profiles where the \"categorization\" data has diagnosed that liquid water is present and liquid water path is available from a coincident microwave radiometer. The model temperature and pressure were used to estimate the theoretical adiabatic liquid water content gradient for each cloud base and the adiabatic liquid water content is then scaled so that its integral matches the radiometer measurement so that the liquid water content now follows a quasi-adiabatic profile. If the liquid layer is detected by the lidar only, there is the potential for cloud top height to be underestimated and so if the adiabatic integrated liquid water content is less than that measured by the microwave radiometer, the cloud top is extended until the adiabatic integrated liquid water content agrees with the value measured by the microwave radiometer. Missing values indicate that either 1) a liquid water layer was diagnosed but no microwave radiometer data was available, 2) a liquid water layer was diagnosed but the microwave radiometer data was unreliable; this may be because a melting layer was present in the profile, or because the retrieved lwp was unphysical (values of zero are not uncommon for thin supercooled liquid layers), or 3) that rain is present in the profile and therefore, the vertical extent of liquid layers is difficult to ascertain.

Beschreibung: The dataset contains daily nc-files of the Cloudnet ice particles effective radius during Polarstern cruise PS106 based on: - 35.5 GHz Cloud radar MIRA. The data is retrieved using the instrument synergystic approach Cloudnet (Illingworth, 2007 doi:10.1175/BAMS-88-6-883 ). This variable was calculated from visible extinction coefficient and ice water content using the following analytical formula: r_eff[\\mum] = 3*IWC[g cm-3] / (2 rho_i \\alpha[m-1]) 10**6 with rho_i=0.917g cm-3 the desity of solid ice. This retrieval is based on an analytical relationship between IWC and \\alpha. However both parameters have been retrieved following Hogan 2006 (doi:10.1175/JAM2340.1 ) by using two different empirical formulas based on the same measured quantity, the 35.5-GHz radar reflectivity factor.