Can we use observed cloud overlap in models?

Description

The assumption of subgrid cloud overlap used in global climate model (GCMs) is critical to the computed radiation budget. It has been suggested that cloudy layers that are contiguous have a maximum overlap, and other cloudy layers have a random overlap. Some recent parameterizations also use the decorrelation length scale to parameterize the transition from maximum to random overlap. Cloud overlap parameters are typically computed based on observed cloud masks, regardless of the optical thickness of each cloud layer. This study uses the radar data from the ARM Southern Great Plains (SGP), North Slope of Alaska (NSA), and Tropical Western Pacific (TWP) sites to examine possible biases in calculated radiation when observed cloud overlap is used. It is then proposed that effective cloud overlap parameter (e.g., weighted by optical thickness) or a subgrid Probability Distribution Function (PDF) representation of subgrid clouds (instead of the binary representation) should be used for radiation calculations.