A41E-0106
Evaluation of a GCM with the CERES Flux-by-Cloud Type Simulator
Thursday, 17 December 2015
Poster Hall (Moscone South)
Zachary Eitzen1, Wenying Su2, Kuan-Man Xu2, Norman G Loeb2, David Doelling2, Alejandro Bodas-Salcedo3 and Moguo Sun4, (1)Science Systems and Applications, Inc., Lanham, MD, United States, (2)NASA Langley Research Center, Hampton, VA, United States, (3)Met Office Hadley center for Climate Change, Exeter, United Kingdom, (4)Science Systems and Applications, Inc. Hampton, Hampton, VA, United States
Abstract:
Recently, the CERES FluxByCldTyp data product has been developed. This product sorts the CERES top-of-atmosphere (TOA) outgoing shortwave (SW) and longwave (LW) fluxes by region and cloud type. Here, the cloud types are defined by optical depth and cloud top pressure, with bins similar to those of ISCCP. This observational product has the potential to be a powerful tool to evaluate the clouds produced in climate models by helping to identify physical parameterizations that are problematic (e.g., convective clouds, boundary-layer parameterizations, or processes that involve surface albedo). Also, when the flux-by-cloud type and frequency of cloud types are simultaneously used to evaluate a model, the results can determine whether an unrealistically large or small occurrence of a given cloud type has an important radiative impact for a given region. Here, a simulator of the flux-by-cloud type product is applied to three-hourly data from the UK Met Office’s HadGEM2-A model using the Langley Fu-Liou radiative transfer model to obtain TOA SW and LW fluxes. For the simulator, a cloud generator with a maximum-random overlap assumption is applied to each grid cell that has an Aqua overpass within 1.5 hours, generating 1000 individual subcolumns with binary (0 or 1) cloud fraction. Many of the subcolumns are identical to one another, reducing the number of radiative transfer calculations required by over 90%. When the simulator is applied to the Southeast Pacific, Equatorial Pacific, and Southern Great Plains, it produces TOA fluxes similar to the grid-mean fluxes produced by HadGEM2-A, although the simulator’s LW fluxes are somewhat lower than those from the model. The HadGEM2-A TOA SW albedo over the Southeast Pacific (0.184) is close to that observed (0.193). However, when the simulator output is compared to the CERES flux-by-cloud type and frequency of cloud types in the Southeast Pacific region, we find that the HadGEM2-A albedos are slightly lower than observed for most cloud types, but for clouds with the highest optical depths, albedos are higher than observed.
