Ben Johnson

	Ben Johnson in his office

Ben Johnson (CGD), developed and implemented a new cloud fraction parameterization scheme for the NCAR Community Atmosphere Model (CAM). This scheme predicts the Probability Distribution Function (PDF) of total water (water vapor plus cloud liquid and/or ice) within each gridbox of the model. This PDF is used to derive the fraction of the gridbox that is saturated (i.e. cloudy), and the amount of condensate (cloud liquid and/or ice) that is associated with that cloud. The advantage of the PDF framework is that cloud fraction and cloud condensate are inextricably linked to each other through the PDF, whereas in previous methods the cloud fraction and the cloud condensate were predicted separately, and many inconsistencies arose because of their separate treatments. The new cloud scheme is successfully running in CAM and is expected to replace the previous cloud fraction scheme in the next major release of CAM and the Community Climate System Model (CCSM), in 2-3 years time.

During the year Ben continued to develop and evaluate the new PDF-based cloud fraction scheme using the Single Column Community Atmosphere Model, which iis the single-column version of CAM . I used observational date from the Atmospheric Radiation Measurement (ARM) program to provide the forcing conditions for single-column model experiments and to evaluate the performance of the new cloud scheme. In such case studies the new cloud fraction scheme usually gave a similar cloud fraction to the original cloud scheme. However, with the new cloud fraction scheme there was a much more realistic variation of cloud fraction with the amount of cloud condensate, whereas with the original cloud fraction scheme was prone to generating unphysical variations of cloud fraction and cloud condensate since these two quantities were not sufficiently well linked in the original methods.

This year I have also evaluated the performance of the new cloud fraction scheme in CAM , the global atmospheric model. The new scheme predicts 5-10% less low cloud compared to the original cloud fraction scheme leads to a 5 Wm -2 overestimation of the global-mean absorption of solar radiation. However, through experimentation I have shown that these biases could be remedied by optimizing, or "tuning" one or two of the free parameters in the new cloud scheme. Such "tuning" is a necessary practice practice to ensure that a climate model will simulate a stable climate with a balance between absorbed solar radiation and outgoing longwave radiation.

Funding Sources

This research is supported by the National Science Foundation.