Review - acp-2021-926

Abstract. Marine stratocumuli are the most dominant cloud type by area coverage in the Southern Ocean (SO). They can be divided into different self-organized cellular morphological regimes known as open and closed mesoscale-cellular convective (MCC) clouds. Open and closed cells are the two most frequent types of organizational regimes in the SO. Using the liDAR-raDAR (DARDAR) version 2 retrievals, we quantify 59 % of all MCC clouds in this region as mixed-phase clouds (MPCs) during a 4-year time period from 2007 to 2010. The net radiative effect of SO MCC clouds is governed by changes in cloud albedo. Both cloud morphology and phase have previously been shown to impact cloud albedo individually, but their interactions and their combined impact on cloud albedo remain unclear. Here, we investigate the relationships between cloud phase, organizational patterns, and their differences regarding their cloud radiative properties in the SO. The mixed-phase fraction, which is defined as the number of MPCs divided by the sum of MPC and supercooled liquid cloud (SLC) pixels, of all MCC clouds at a given cloud-top temperature (CTT) varies considerably between austral summer and winter. We further find that seasonal changes in cloud phase at a given CTT across all latitudes are largely independent of cloud morphology and are thus seemingly constrained by other external factors. Overall, our results show a stronger dependence of cloud phase on cloud-top height (CTH) than CTT for clouds below 2.5 km in altitude. Preconditioning through ice-phase processes in MPCs has been observed to accelerate individual closed-to-open cell transitions in extratropical stratocumuli. The hypothesis of preconditioning has been further substantiated in large-eddy simulations of open and closed MPCs. In this study, we do not find preconditioning to primarily impact climatological cloud morphology statistics in the SO. Meanwhile, in-cloud albedo analysis reveals stronger changes in open and closed cell albedo in SLCs than in MPCs. In particular, few optically thick (cloud optical thickness >10) open cell stratocumuli are characterized as ice-free SLCs. These differences in in-cloud albedo are found to alter the cloud radiative effect in the SO by 21 to 39 W m−2 depending on season and cloud phase.

Marine stratocumuli are the most dominant cloud type by area coverage in the Southern Ocean (SO). They can be divided into different self-organized cellular morphological regimes known as open and closed mesoscale-cellular convec- tive (MCC) clouds. Open and closed cells are the two most frequent types of organizational regimes in the SO. Using the liDAR- raDAR (DARDAR) version 2 retrievals, we quantify 59 % of all MCC clouds in this region as mixed-phase clouds (MPCs) during a 4-year time period from 2007 to 2010. The net radiative effect of SO MCC clouds is governed by changes in cloud albedo. Both, cloud morphology and phase, have previously been shown to impact cloud albedo individually, but their interac- tions and their combined impact on cloud albedo remain unclear. Here, we investigate the relationships between cloud phase, organizational patterns, and their differences regarding their cloud radiative properties in the SO. The mixed-phase fraction, which is defined as the number of MPCs divided by the sum of MPC and supercooled liquid cloud (SLC) pixels, of all MCC clouds at a given cloud-top temperature (CTT) varies considerably between austral summer and winter. We further find that seasonal changes in cloud phase at a given CTT across all latitudes are largely independent of cloud morphology and are thus seemingly constrained by other external factors. Overall, our results show a stronger dependence of cloud phase on cloud-top height (CTH) than CTT for clouds below 2.5 km in altitude. Preconditioning through ice-phase processes in MPCs has been observed to accelerate individual closed to open cell transitions in extratropical stratocumuli. The hypothesis of preconditioning has been further substantiated in large-eddy simulations of open and closed MPCs. In this study, we do not find preconditioning to primarily impact climatological SO cloud mor- phology statistics. Meanwhile, in-cloud albedo analysis reveals stronger changes in open and closed cell albedo in SLCs than MPCs. In particular few optically thick (cloud optical thickness > 10) open cell stratocumuli are characterized as ice-free SLCs. Theses differences in in-cloud albedo are found to alter the cloud radiative effect in the SO by 12 W m−2 to 39 W m−2 depending on season and cloud phase.

The accurate representation of clouds and their phase is crucial to enable a correct representation of the Earth’s radiation balance. This has been demonstrated by large radiative errors in global models over the Southern Ocean mainly caused by an incorrect representation of supercooled liquid lowlevel clouds. In contrast to lowlevel clouds, midlevel clouds are rarely investigated. To fill this gap, this study investigates active satellite observations of midlevel clouds over the Southern Ocean and the Arctic Ocean from 2007 and 2008 with a comprehensive comparison to observations of lowlevel clouds. Midlevel and lowlevel clouds are distinguished by cloud base height. The DARDAR dataset provides a detailed phase categorization based on CloudSat and CALIPSO measurements. We have analyzed the cloud phase partitioning as a function of cloud top temperature, vertical cloud thickness, and the horizontal cloud extent. A local minimum in the mean liquid fraction within a cloud column can be observed for a cloud top temperature of -15 °C. An exception to this observation occurs for lowlevel clouds over the Arctic Ocean, which feature a plateau instead of a minimum. This hints at processes producing ice at these temperatures, which could be habit dependent vapor growth, secondary ice production, or a combination of both processes, as already discussed in other studies. Furthermore, daily sea ice concentrations from a passive microwave instrument are collocated to investigate their correlation with the cloud phase. At equal cloud top temperature, lowlevel clouds over the Southern Ocean and the Arctic Ocean have a higher liquid fraction, if they occur over sea ice. Midlevel clouds over the Southern Ocean show the same behaviour, while midlevel clouds over the Arctic Ocean show no significant phase dependence on the sea ice concentration. In addition, collocated CAMS reanalysis data are used to investigate the influence of different concentrations of various aerosol types such as sea salt, dust, black carbon, or organic matter on the cloud phase. Preliminary results show a stronger influence of the mixing ratio of sea salt on the phase of lowlevel clouds, but also the phase of midlevel clouds over the Southern Ocean seems to be influenced. Future work will further investigate the influence of different parameters, such as sea ice concentration, aerosol concentration, and cloud top temperature on the cloud phase by applying a machine learning model and exploring the relative importance of these parameters for the cloud phase, as well as their interactions and synergies.

Abstract. Low-level (cloud tops below 2 km) mixed-phase clouds are important in amplifying warming in the Arctic region through positive feedback in cloud fraction, water content and phase. In order to understand the cloud feedbacks in the Arctic region, good knowledge of the vertical distribution of the cloud water content, particle size and phase is required. Here we investigate the vertical extent of the cloud-phase and ice-phase optical properties in six case studies measured in the European Arctic during the ACLOUD campaign. Late spring- and summertime stratiform clouds were sampled in situ over pack ice, marginal sea ice zone and open-ocean surface, with cloud top temperatures varying between −15 and −1.5 ∘C. The results show that, although the liquid phase dominates the upper parts of the clouds, the ice phase was frequently observed in the lower parts down to cloud top temperatures as warm as −3.8 ∘C. In the studied vertical cloud profiles, the maximum of average liquid phase microphysical properties, droplet number concentration, effective radius and liquid water content, varied between 23 and 152 cm−3, 19 and 26 µm, 0.09 and 0.63 g m−3, respectively. The maximum of average ice-phase microphysical properties varied between 0.1 and 57 L−1 for the ice number concentration, 40 and 70 µm for the effective radius, and 0.005 and 0.08 g m−3 for the ice water content. The elevated ice crystal number concentrations and ice water paths observed for clouds, with cloud top temperatures between −3.8 and −8.7 ∘C can be likely attributed to secondary ice production through rime splintering. Low asymmetry parameters between 0.69 and 0.76 were measured for the mixed-phase ice crystals with a mean value of 0.72. The effect of the ice-phase optical properties on the radiative transfer calculations was investigated for the four cloud cases potentially affected by secondary ice production. Generally the choice of ice-phase optical properties only has a minor effect on the cloud transmissivity and albedo, except in a case where the ice phase dominated the upper cloud layer extinction. In this case, cloud albedo at solar wavelengths was increased by 10 % when the ice phase was given its measured optical properties instead of treating it as liquid phase. The presented results highlight the importance of accurate vertical information on cloud phase for radiative transfer and provide a suitable data set for testing microphysical parameterizations in models.

Abstract. We previously developed the Cloud Height Retrieval from O2 Molecular Absorption (CHROMA) algorithm for the Ocean Color Instrument (OCI) on the new NASA Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) mission. Here, we apply CHROMA to observations from the Ocean Land Colour Instrument (OLCI) to guide expectations for PACE, as it will take some time to obtain large-scale validation data for OCI. We use cloud top height (CTH), phase, and (for liquid clouds) cloud optical thickness (COT) data from the ground-based Atmospheric Radiation Measurement (ARM) network to evaluate the OLCI retrievals. We found that OLCI and Moderate Resolution Imaging Spectroradiometer (MODIS) CTH compare similarly well to the ARM reference. OLCI has a tendency to underestimate CTH as CTH increases, and algorithm assumptions about cloud geometric thickness may contribute to this. ARM COT from multifilter shadowband radiometers (MFRSR) and Sun photometers are well-correlated with one another, albeit with a roughly 30 % offset on average; OLCI and MODIS COT agree more closely with the MFRSR data. OLCI retrieval uncertainty estimates show skill at telling low-uncertainty cases from high-uncertainty ones, although CTH uncertainties are underestimated. Additionally, we compare the OLCI data to satellite retrievals based on thermal infrared measurements from MODIS and Sea and Land Surface Temperature Radiometer (SLSTR) data. Differences are broadly consistent with physical expectations based on the A-band vs. thermal techniques, although one key challenge in such aggregated comparisons is different cloud masking sensitivities and algorithm failure rates meaning additional sampling differences are introduced. We conclude by discussing the transition to and possible enhancements for PACE OCI.

Abstract. A newly developed daytime cloud property retrieval algorithm, FAME-C (Freie Universität Berlin AATSR MERIS Cloud), is presented. Synergistic observations from the Advanced Along-Track Scanning Radiometer (AATSR) and the Medium Resolution Imaging Spectrometer (MERIS), both mounted on the polar-orbiting Environmental Satellite (Envisat), are used for cloud screening. For cloudy pixels two main steps are carried out in a sequential form. First, a cloud optical and microphysical property retrieval is performed using an AATSR near-infrared and visible channel. Cloud phase, cloud optical thickness, and effective radius are retrieved, and subsequently cloud water path is computed. Second, two cloud top height products are retrieved based on independent techniques. For cloud top temperature, measurements in the AATSR infrared channels are used, while for cloud top pressure, measurements in the MERIS oxygen-A absorption channel are used. Results from the cloud optical and microphysical property retrieval serve as input for the two cloud top height retrievals. Introduced here are the AATSR and MERIS forward models and auxiliary data needed in FAME-C. Also, the optimal estimation method, which provides uncertainty estimates of the retrieved property on a pixel basis, is presented. Within the frame of the European Space Agency (ESA) Climate Change Initiative (CCI) project, the first global cloud property retrievals have been conducted for the years 2007–2009. For this time period, verification efforts are presented, comparing, for four selected regions around the globe, FAME-C cloud optical and microphysical properties to cloud optical and microphysical properties derived from measurements of the Moderate Resolution Imaging Spectroradiometer (MODIS) on the Terra satellite. The results show a reasonable agreement between the cloud optical and microphysical property retrievals. Biases are generally smallest for marine stratocumulus clouds: −0.28, 0.41 μm and −0.18 g m−2 for cloud optical thickness, effective radius and cloud water path, respectively. This is also true for the root-mean-square deviation. Furthermore, both cloud top height products are compared to cloud top heights derived from ground-based cloud radars located at several Atmospheric Radiation Measurement (ARM) sites. FAME-C mostly shows an underestimation of cloud top heights when compared to radar observations. The lowest bias of −0.3 km is found for AATSR cloud top heights for single-layer clouds, while the highest bias of −3.0 km is found for AATSR cloud top heights for multilayer clouds. Variability is low for MERIS cloud top heights for low-level clouds, and high for MERIS cloud top heights for mid-level and high-level single-layer clouds, as well as for both AATSR and MERIS cloud top heights for multilayer clouds.

Spaceborne lidar observations from the Cloud‐Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) satellite are used to evaluate cloud amount and cloud phase in the Community Atmosphere Model version 5 (CAM5), the atmospheric component of a widely used state‐of‐the‐art global coupled climate model (Community Earth System Model). By embedding a lidar simulator within CAM5, the idiosyncrasies of spaceborne lidar cloud detection and phase assignment are replicated. As a result, this study makes scale‐aware and definition‐aware comparisons between model‐simulated and observed cloud amount and cloud phase. In the global mean, CAM5 has insufficient liquid cloud and excessive ice cloud when compared to CALIPSO observations. Over the ice‐covered Arctic Ocean, CAM5 has insufficient liquid cloud in all seasons. Having important implications for projections of future sea level rise, a liquid cloud deficit contributes to a cold bias of 2–3°C for summer daily maximum near‐surface air temperatures at Summit, Greenland. Over the midlatitude storm tracks, CAM5 has excessive ice cloud and insufficient liquid cloud. Storm track cloud phase biases in CAM5 maximize over the Southern Ocean, which also has larger‐than‐observed seasonal variations in cloud phase. Physical parameter modifications reduce the Southern Ocean cloud phase and shortwave radiation biases in CAM5 and illustrate the power of the CALIPSO observations as an observational constraint. The results also highlight the importance of using a regime‐based, as opposed to a geographic‐based, model evaluation approach. More generally, the results demonstrate the importance and value of simulator‐enabled comparisons of cloud phase in models used for future climate projection.

Abstract. The effects of different aerosol types on cloud albedo are analysed using the linear relation between total albedo and cloud fraction found on a monthly mean scale in regions of subtropical marine stratocumulus clouds and the influence of simulated aerosol variations on this relation. Model experiments from the Coupled Model Intercomparison Project phase 5 (CMIP5) are used to separately study the responses to increases in sulfate, non-sulfate and all anthropogenic aerosols. A cloud brightening on the month-to-month scale due to variability in the background aerosol is found to dominate even in the cases where anthropogenic aerosols are added. The aerosol composition is of importance for this cloud brightening, that is thereby region dependent. There is indication that absorbing aerosols to some extent counteract the cloud brightening but scene darkening with increasing aerosol burden is generally not supported, even in regions where absorbing aerosols dominate. Month-to-month cloud albedo variability also confirms the importance of liquid water content for cloud albedo. Regional, monthly mean cloud albedo is found to increase with the addition of anthropogenic aerosols and more so with sulfate than non-sulfate. Changes in cloud albedo between experiments are related to changes in cloud water content as well as droplet size distribution changes, so that models with large increases in liquid water path and/or cloud droplet number show large cloud albedo increases with increasing aerosol. However, no clear relation between model sensitivities to aerosol variations on the month-to-month scale and changes in cloud albedo due to changed aerosol burden is found.

Abstract. Clouds play a crucial role in the Earth's energy budget and the hydrological cycle. However, differences in the spatiotemporal resolution of satellite sensors and in retrieval algorithms lead to substantial heterogeneity among retrieved cloud products. Based on global geostationary satellite thermal infrared brightness temperature data from the Gridded Satellite (GridSat-B1) project, this study applied the single-layer Cloud retrieval model – Small Attention-UNet (Cloud-SmaAtUNet) algorithm in the DaYu CLoud Analysis System (DaYu-CLAS) to retrieve a global cloud product with a 3 h temporal resolution, 0.07° spatial resolution, and 23 year temporal span (2000–2022). This product is referred to as the DaYu Global Cloud physical properties Products (DaYu-GCP). The DaYu-GCP includes CLoud Phase (CLP), Cloud Top Height (CTH), Cloud Optical Thickness (COT), and Cloud Effective Radius (CER), covering all regions between 70° S–70° N and 180° W–180° E. Evaluation based on the Moderate-resolution Imaging Spectroradiometer (MODIS) official cloud products shows that the annual CLP identification accuracy of DaYu-GCP remains stable at 85 % ± 0.7 %, while the annual RMSE for CTH, COT, and CER stabilize at 1.50 ± 0.03 km, 10.71 ± 0.15, and 6.75 ± 0.10 µm, respectively. The multi-year variations in accuracy are within 2 %, with no evident interannual differences, and the spatiotemporal distributions are continuous. In addition, evaluation based on observations from the Cloud Profiling Radar and the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) indicates that the DaYu-GCP products show reasonable day–night consistency for optically thin cloud. Furthermore, the DaYu-GCP products are compared with other global cloud products. Taking the Northern Hemisphere as an example, the interannual variations of Cloud Cover Frequency (CCF), CTH, COT, and CER retrieved from DaYu-GCP show correlation coefficients of 0.760, 0.486, 0.764, and 0.514 with the ISCCP product, respectively, and 0.444, 0.778, 0.171, and 0.412 with the CLARA-A3 product. The DaYu-GCP dataset, which is stored in the Network Common Data Format (NetCDF), is freely available on the Science Data Bank at https://doi.org/10.57760/sciencedb.26292 (Zhao et al., 2026). The corresponding code can be found at https://github.com/lingxiao-zhao/DaYu-GCP (last access: 25 June 2025).

One can count on significant increase in efficiency of algorithms used for solving the inverse problem of atmospheric total water vapor (TWV) and liquid water content (LWC) microwave radiometric retrieval through introducing into the related computational models such information as cloud top height, ground-level and cloud top temperatures, binary cloud mask or also cloud phase distributions obtained in the radiometer antenna’s field-of-view. This supplementary information, in turn, may be separately pre-reconstructed by means of data on infrared (IR) observations carried out simultaneously and in accordance with microwave ones. But it also makes sense to consider the possibilities and advantages of developing unified (single) algorithms, which involve synchronous IR and microwave data processing (using all available spectrum at once) for solving various inverse problems. We suppose such algorithms could be implemented on the basis of deep convolutional neural networks (CNNs) using data assimilation approach. Herewith, to demonstrate capacities of CNNs in combining estimates made by analyzing data of various wavelength ranges, we consider the problem of cloud phase spatial distribution IR-reconstruction in more detail, because of both near and far IR subranges are sensitive to cloud phase, but two distinct and independent reconstruction algorithms are usually utilized for them.

Supplementary material to "Exploring Relations between Cloud Morphology, Cloud Phase, and Cloud Radiative Properties in Southern Ocean Stratocumulus Clouds"

Cloud and aerosol lidars measuring backscatter and depolarization ratio are most suitable instruments to detect cloud phase (liquid, ice, or mixed phase). However, such instruments are not widely deployed as part of operational networks. In this study, we propose a new algorithm to detect supercooled liquid water clouds based solely on ceilometers measuring only co-polarisation backscatter. We utilise observations collected at Davis, Antarctica, where low-level, mixed phase clouds, including supercooled liquid water (SLW) droplets and ice crystals remain poorly understood, due to the paucity of ground-based observations. A 3-month set of observations were collected during the austral summer of November 2018–February 2019, with a variety of instruments including a depolarization lidar and a W-Band cloud radar which were used to build a 2-dimensional cloud phase mask distinguishing SLW and mixed phase clouds. This cloud phase mask is used as the reference to develop a new algorithm based on the observations of a single polarisation ceilometer operating in the vicinity for the same period. Deterministic and data-driven retrieval approaches were evaluated: an extreme gradient boosting (XGBoost) framework ingesting backscatter average characteristics was the most effective method at reproducing the classification obtained with the combined radar-lidar approach with an accuracy as high as 0.91. This study provides a new SLW retrieval approach based solely on ceilometer data and highlights the considerable benefits of these instruments to provide intelligence on cloud phase in polar regions that usually suffer from a paucity of observations.

Under certain conditions mass transfer limitations on the growth of cloud condensation nuclei (CCN) may have a significant impact on the number of droplets that can form in a cloud. The assumption that particles remain in equilibrium until activated may therefore not always be appropriate for aerosol populations existing in the atmosphere. This work identifies three mechanisms that lead to kinetic limitations, the effect of which on activated cloud droplet number and cloud albedo is assessed using a one-dimensional cloud parcel model with detailed microphysics for a variety of aerosol size distributions and updraft velocities. In assessing the effect of kinetic limitations, we have assumed as cloud droplets not only those that are strictly activated (as dictated by classical Köhler theory), but also unactivated drops large enough to have an impact on cloud optical properties. Aerosol number concentration is found to be the key parameter that controls the significance of kinetic effects. Simulations indicate that the equilibrium assumption leads to an overprediction of droplet number by less than 10% for marine aerosol; this overprediction can exceed 40% for urban type aerosol. Overall, the effect of kinetic limitations on cloud albedo can be considered important when equilibrium activation theory consistently overpredicts droplet number by more than 10%. The maximum change in cloud albedo as a result of kinetic limitations is less than 0.005 for cases such as marine aerosol; however albedo differences can exceed 0.1 under more polluted conditions. Kinetic limitations are thus not expected to be climatically significant on a global scale, but can regionally have a large impact on cloud albedo.

Cold-air outbreaks over high-latitude oceans typically include mixed-phase clouds and precipitation, in particular supercooled liquid clouds that support snow and graupel through ice growth processes. The partitioning of the total water into the liquid and ice phases impacts both weather and climate prediction, but accurate measurements on the phase partitioning remain difficult to acquire, especially near–real time. Here, we present a machine learning approach to retrieve liquid water path (LWP) using passive microwave measurements combined with vertically integrated radar reflectivities. The approach is an extension of Cadeddu et al. (2009), with the novel addition of radar reflectivity. The machine learning models are trained using the Passive and Active Microwave Radiative Transfer (PAMTRA) code applied to output from numerical simulations of three independent cold-air outbreaks sampled during the Cold-Air Outbreaks in the Marine Boundary Layer Experiment (COMBLE) campaign. Brightness temperatures corresponding to the four sidebands of an upward-looking G-band (183 GHz) vapor radiometer, along with the vertically integrated reflectivity from a zenith-pointing 95-GHz Wyoming Cloud Radar, are simulated from the perspective of a near-surface aircraft track. The radar reflectivity helps discriminate the snow contribution to the brightness temperatures. The machine learning models are thereafter tested on a simulation of an independent cold-air outbreak during COMBLE and against measurements from the U.S. Department of Energy Atmospheric Radiation Measurement North Slope of Alaska observatory. This machine learning approach is shown to provide robust, computationally efficient, near-real-time measurements of LWP and water vapor path during the Cold-Air Outbreak Experiment in the Sub-Arctic Region (CAESAR) campaign in February–April 2024. Significance Statement Precipitation from mixed-phase clouds over the high-latitude open waters impacts shipping, fishing, and coastal communities. Mixed-phase clouds are also a modeling challenge, inhibiting weather and climate prediction skill. More comprehensive measurements of the cloud liquid and ice water path occupying the same vertical column improve our understanding and ability to represent this challenging cloud type. Here, we propose a new retrieval to determine the liquid water path within air columns that also contain ice.

Global storm resolving models (GSRMs) represent the next generation of global climate models. One of them is a 5‐km Icosahedral Nonhydrostatic Weather and Climate Model (ICON). Its high resolution means that parameterizations of convection and clouds, including subgrid‐scale clouds, are omitted, relying on explicit simulation but necessarily utilizing microphysics and turbulence parameterizations. Standard‐resolution (10–100 km) models, which use convection and cloud parameterizations, have substantial cloud biases over the Southern Ocean (SO), adversely affecting radiation and sea surface temperature. The SO is dominated by low clouds, which cannot be observed accurately from space due to overlapping clouds, attenuation, and ground clutter. We evaluated SO clouds in ICON and the ERA5 and MERRA‐2 reanalyzes using approximately 2400 days of lidar observations and 2300 radiosonde profiles from 31 voyages and a Macquarie Island station during 2010–2021, compared to the model and reanalyzes using a ground‐based lidar simulator. We found that ICON and the reanalyzes underestimate the total cloud fraction by about 10% and 20%, respectively. ICON and ERA5 overestimate the cloud occurrence peak at about 500 m, associated with underestimated lower tropospheric stability and overestimated lifting condensation level. The reanalyzes strongly underestimate fog and very low‐level clouds, and MERRA‐2 underestimates cloud occurrence at almost all heights. Outgoing shortwave radiation is overestimated in MERRA‐2, implying a “too few, too bright” cloud problem. SO cloud and fog biases are a substantial issue in the analyzed model and reanalyzes and result in shortwave and longwave radiation biases.