Premium
Characterizing and understanding systematic biases in the vertical structure of clouds in CMIP5/CFMIP2 models
Author(s) -
Cesana G.,
Waliser D. E.
Publication year - 2016
Publication title -
geophysical research letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.007
H-Index - 273
eISSN - 1944-8007
pISSN - 0094-8276
DOI - 10.1002/2016gl070515
Subject(s) - environmental science , atmospheric sciences , lidar , satellite , pathfinder , liquid water path , climatology , cloud computing , climate model , meteorology , geology , aerosol , remote sensing , climate change , geography , physics , computer science , oceanography , astronomy , library science , operating system
From a traditional low‐, middle‐, and high‐cloud “layered” perspective as well as a more detailed “level” perspective (40 levels), we compare the vertical distribution of clouds in 12 general circulation models (GCMs) against the GCM‐Oriented Cloud‐Aerosols Lidar and Infrared Pathfinder Satellite Observations Cloud Product (CALIPSO‐GOCCP) using a satellite simulator approach. The layered perspective shows that models exhibit the similar regional biases: an overestimate (underestimate) of high clouds over oceans (continents) in the tropics and a strong underestimate of low clouds over stratocumulus regions. Although high clouds are too infrequent on average, the level perspective reveals that high‐level clouds fill too many upper levels of the column when present (geometrically too thick), suggesting an overestimation of the cloud overlap. Compositing by dynamical regimes and large‐scale relative humidity shows that the models tend to have too many high‐level clouds in moist environments and too few boundary layer clouds in dry environments regardless of dynamical regimes.