On the relationship between cloud optical depth and temperature: inferences from ground-based observations at ARM sites

 
Poster PDF

Authors

Yunyan Zhang — Lawrence Livermore National Laboratory
Stephen Klein — Lawrence Livermore National Laboratory
Neil Daniel Gordon — Scripps Institution of Oceanography

Category

Cloud Properties

Description

The dependence of cloud optical depth on cloud top temperature has been explored using International Satellite Cloud Climatology Project (ISCCP) satellite data by Tselioudis et al. 1992 showing that cloud optical depth increases with cold temperatures and decreases with warm temperatures. There is a growing interest in using this relationship to evaluate global climate modeling results and study long-term cloud feedback on climate change (Gordon and Klein 2012). However there is a lack of systematic investigation of this relationship based on ground-based observations. To extend the approach in Del Genio and Wolf (2000) on using ARM observations, we revisit this relationship using most updated long-term high-quality-controlled data to (1) provide a more accurate quantification of this relationship and (2) explore physical mechanisms that determine the relationship.

We first select single-layer overcast cloud observations at the Southern Great Plains (SGP) and North Slope of Alaska (NSA) sites and then separate the contribution to the change of cloud optical depth with temperature due to different factors, such as cloud physical thickness, cloud water content, and cloud water and ice effective radius by using independent measurement from different instrument and retrieval algorithms. Consistent with model results and previous studies, the change of cloud water content with temperature is rather dominating the change of cloud optical depth with temperature. The results are also compared with phase 3 of the Coupled Model Intercomparison Project (CMIP3)/ Cloud Feedback Model Intercomparison Project (CFMIP) global climate model results output at ARM sites. Multi-model mean is able to reproduce this relationship within the uncertainty range of observations and shows the change of cloud water content with temperature is a dominating factor, while significant spread still exists among models on detail aspects.