Premium
Cloud Top Phase Distributions of Simulated Deep Convective Clouds
Author(s) -
Hoose C.,
Karrer M.,
Barthlott C.
Publication year - 2018
Publication title -
journal of geophysical research: atmospheres
Language(s) - English
Resource type - Journals
eISSN - 2169-8996
pISSN - 2169-897X
DOI - 10.1029/2018jd028381
Subject(s) - cloud fraction , cloud computing , liquid water content , cloud top , cloud height , environmental science , phase (matter) , convection , cloud cover , observable , atmospheric sciences , meteorology , geology , physics , computer science , quantum mechanics , operating system
Space‐based observations of the thermodynamic cloud phase are frequently used for the analysis of aerosol indirect effects and other regional and temporal trends of cloud properties; yet they are mostly limited to the cloud top layers. This study addresses the information content in cloud top phase distributions of deep convective clouds during their growing stage. A cloud‐resolving model with grid spacings of 300 m and lower is used in two different setups, simulating idealized and semiidealized isolated convective clouds of different strengths. It is found that the cloud top phase distribution is systematically shifted to higher temperatures compared to the in‐cloud phase distribution due to lower vertical velocities and a resultingly stronger Wegener‐Bergeron‐Findeisen process at the cloud top. Sensitivity studies show that heterogeneous freezing can modify the cloud top glaciation temperature (where the ice pixel fraction reaches 50%) and ice multiplication via rime splintering is visible in an early ice onset at temperatures around −10°C. However, if the analyses are repeated with a coarsened horizontal resolution (above 1 km, similar to many satellite data sets), a significant part of this signal is lost, which limits the detectability of these microphysical fingerprints in the observable cloud top phase distribution. In addition, variation in the cloud dynamics also impacts the cloud phase distribution but cannot be quantified easily.