G. Cesana

Multimodel evaluation of cloud phase transition using satellite and reanalysis data

We take advantage of climate simulations from two multimodel experiments to characterize and evaluate the cloud phase partitioning in 16 general circulation models (GCMs), specifically the vertical structure of the transition between liquid and ice in clouds. We base our analysis on the ratio of ice condensates to the total condensates (phase ratio, PR). Its transition at 90% (PR90) and its links with other relevant variables are evaluated using the GCM-Oriented Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation Cloud Product climatology, reanalysis data, and other satellite observations. In 13 of 16 models, the PR90 transition height occurs too low (6 km to 8.4 km) and at temperatures too warm (−13.9°C to −32.5°C) compared to observations (8.6 km, −33.7°C); features consistent with a lack of supercooled liquid with respect to ice above 6.5 km. However, this bias would be slightly reduced by using the lidar simulator. In convective regimes (more humid air and precipitation), the observed cloud phase transition occurs at a warmer temperature than for subsidence regimes (less humid air and precipitation). Only few models manage to roughly replicate the observed correlations with humidity (5/16), vertical velocity (5/16), and precipitation (4/16); 3/16 perform well for all these parameters (MPI-ESM, NCAR-CAM5, and NCHU). Using an observation-based Clausius-Clapeyron phase diagram, we illustrate that the Bergeron-Findeisen process is a necessary condition for models to represent the observed features. Finally, the best models are those that include more complex microphysics.

Using in-situ airborne measurements to evaluate three cloud phase products derived from CALIPSO

We compare the cloud detection and cloud phase determination of three independent climatologies based on Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) to airborne in situ measurements. Our analysis of the cloud detection shows that the differences between the satellite and in situ measurements mainly arise from three factors. First, averaging CALIPSO Level l data along track before cloud detection increases the estimate of high- and low-level cloud fractions. Second, the vertical averaging of Level 1 data before cloud detection tends to artificially increase the cloud vertical extent. Third, the differences in classification of fully attenuated pixels among the CALIPSO climatologies lead to differences in the low-level Arctic cloud fractions. In another section, we compare the cloudy pixels detected by colocated in situ and satellite observations to study the cloud phase determination. At midlatitudes, retrievals of homogeneous high ice clouds by CALIPSO data sets are very robust (more than 94.6% of agreement with in situ). In the Arctic, where the cloud phase vertical variability is larger within a 480 m pixel, all climatologies show disagreements with the in situ measurements and CALIPSO-General Circulation Models-Oriented Cloud Product (GOCCP) report significant undefined-phase clouds, which likely correspond to mixed-phase clouds. In all CALIPSO products, the phase determination is dominated by the cloud top phase. Finally, we use global statistics to demonstrate that main differences between the CALIPSO cloud phase products stem from the cloud detection (horizontal averaging, fully attenuated pixels) rather than the cloud phase determination procedures.

Recent Advances in Arctic Cloud and Climate Research

While the representation of clouds in climate models has become more sophisticated over the last 30+ years, the vertical and seasonal fingerprints of Arctic greenhouse warming have not changed. Are the models right? Observations in recent decades show the same fingerprints: surface amplified warming especially in late fall and winter. Recent observations show no summer cloud response to Arctic sea ice loss but increased cloud cover and a deepening atmospheric boundary layer in fall. Taken together, clouds appear to not affect the fingerprints of Arctic warming. Yet, the magnitude of warming depends strongly on the representation of clouds. Can we check the models? Having observations alone does not enable robust model evaluation and model improvement. Comparing models and observations is hard enough, but to improve models, one must both understand why models and observations differ and fix the parameterizations. It is all a tall order, but recent progress is summarized here.

Direct atmosphere opacity observations from CALIPSO provide new constraints on cloud‐radiation interactions

The spaceborne lidar CALIPSO (Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation) directly measures atmospheric opacity. In 8 years of CALIPSO observations, we find that 69% of vertical profiles penetrate through the complete atmosphere. The remaining 31% do not reach the surface, due to opaque clouds. The global mean altitude of full attenuation of the lidar beam (z_opaque) is 3.2 km, but there are large regional variations in this altitude. Of relevance to cloud-climate studies, the annual zonal mean longwave cloud radiative effect and annual zonal mean z_opaque weighted by opaque cloud cover are highly correlated (0.94). The annual zonal mean shortwave cloud radiative effect and annual zonal mean opaque cloud cover are also correlated (A0.95). The new diagnostics introduced here are implemented within a simulator framework to enable scale-aware and definition-aware evaluation of the LMDZ5B global climate model. The evaluation shows that the model overestimates opaque cloud cover (31% obs. versus 38% model) and z_opaque (3.2 km obs. versus 5.1 km model). In contrast, the model underestimates thin cloud cover (35% obs. versus 14% model). Further assessment shows that reasonable agreement between modeled and observed longwave cloud radiative effects results from compensating errors between insufficient warming by thin clouds and excessive warming due to overestimating both z_opaque and opaque cloud cover. This work shows the power of spaceborne lidar observations to directly constrain cloud-radiation interactions in both observations and models.

Development of Algorithm for Discriminating Hydrometeor Particle Types With a Synergistic Use of CloudSat and CALIPSO

We developed a method for classifying hydrometeor particle types, including cloud and precipitation phase and ice crystal habit, by a synergistic use of CloudSat/Cloud Profiling Radar (CPR) and Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO)/Cloud-Aerosol LIdar with Orthogonal Polarization (CALIOP). We investigated how the cloud phase and ice crystal habit characterized by CALIOP globally relate with radar reflectivity and temperature. The global relationship thus identified was employed to develop an algorithm for hydrometeor type classification with CPR alone. The CPR-based type classification was then combined with CALIPSO-based type characterization to give CPR-CALIOP synergy classification. A unique aspect of this algorithm is to exploit and combine the lidars sensitivity to thin ice clouds and the radars ability to penetrate light precipitation to offer more complete picture of vertically resolved hydrometeor type classification than has been provided by previous studies. Given the complementary nature of radar and lidar detections of hydrometeors, our algorithm delivers thirteen hydrometeor types: warm water, supercooled water, randomly-oriented ice crystal (3D-ice), horizontally-oriented plate (2D-plate), 3D-ice+2D-plate, liquid drizzle, mixed-phase drizzle, rain, snow, mixed-phase cloud, water+liquid drizzle, water+rain and unknown. The global statistics of three-dimensional occurrence frequency of each hydrometeor type revealed that 3D-ice contributes the most to the total cloud occurrence frequency (53.8%), followed by supercooled water (14.3%), 2D-plate (9.2%), rain (5.9%), warm water (5.7%), snow (4.8%), mixed-phase drizzle (2.3%), and the remaining types (4.0%). This hydrometeor type classification provides useful observation-based information for climate model diagnostics in representation of cloud phase and their microphysical characteristics.