|Paper title||Understanding radar-lidar blind spots to constrain liquid cloud retrievals using solar radiances|
|Form of presentation||Poster|
The synergy of radar and lidar from ground-based networks such as ARM and CloudNet/ACTRIS and the A-Train constellation of satellites has revolutionised our understanding of the global and vertical distribution of clouds and precipitation. However, while the complementary sensitivities of lidar to small ice crystals and radar to larger snowflakes can provide near-complete coverage of ice clouds and snow, the detection and vertical location of liquid cloud is much less certain. In mixed-phase, layered or precipitating cloud scenes the lidar is often quickly extinguished within the first layer, and while the radar penetrates most scenes its signal is dominated by larger precipitating hydrometeors. We use simulated EarthCARE measurements of midlatitude and tropical cloud scenes from a numerical weather model to show that these synergistic blind spots result in less than 25% of liquid clouds being detected by volume, representing only around 10% of total liquid water content.
As well as biasing global liquid cloud statistics and water budgets from spaceborne active remote sensing, these undiagnosed clouds cannot be ignored from a radiative perspective. In this study we use simulated EarthCARE measurements to evaluate the performance of EarthCARE’s synergistic retrieval of cloud and precipitation (ACM-CAP), which will assimilate a solar radiance channel from EarthCARE’s multi-spectral imager (MSI) as well as the cloud profiling radar (CPR) and atmospheric lidar (ATLID). We show that assuming that liquid clouds are collocated with precipitation improves the forward-modelled solar albedo in many complex cloud scenes. Even without active measurements of liquid cloud, the solar radiance and CPR path-integrated attenuation are sufficient to constrain the retrieval of a simplified profile of liquid water content, which reduces underestimates in retrieved liquid water path without introducing a significant compensating error. When the profiling retrievals at nadir and MSI imagery are used to reconstruct a 3D across-swath scene (ACM-3D), the missing liquid contributes to a mean bias error of almost 40 gm-2 with respect to the model fields, compared to around -5 gm-2 when liquid is included in the synergistic retrieval constrained by solar radiances. Finally the radiative closure assessment (ACMB-DF) against EarthCARE’s broadband radiometer (BBR) identifies shortwave flux deficits of 50 to 100 Wm-2 due to this undiagnosed liquid cloud associated with deep midlatitude cloud scenes, confirming that a simple assumption accounting for radar-lidar blind spots within the synergistic retrieval can result in significant improvements in retrievals of radiatively-important liquid cloud.