Improvement of a radiation process for the non-hydrostatic model

The Japan Meteorological Agency Non-hydrostatic Model has been developed for research and operational purpose. The 5-km resolution NHM is planed to be operational in March 2006. The radiation scheme is based on Kitagawa (2000). We want to use the predicted clouds in the radiation calculation. But it turns out that the clouds are not suitable for the radiation calculation at present. Therefore the cloud fraction diagnosed from the relative humidity by the conventional method of Ohno and Isa (1984) is used in the radiation calculation. But the clouds have some problems. The clouds tend to spread widely compared with actual cloud distribution observed by the satellite and it is considered to be the main cause of the underestimation of the downward shortwave radiation flux at surface and the negative bias of the surface air temperature. To solve these problems, we carry out some tests on the new cloud fraction diagnostic method of Hack (1998).

The tests revealed that the distribution of the cloud fraction by Hack (1998) was more reasonable than the cloud fraction by Ohno and Isa (1984) judging from the satellite image. In fact, the underestimation of the downward shortwave radiation flux at surface was removed. The cloud fraction by Hack (1998) brought forth most reasonable OLR compared with the NOAA-16 satellite observation. On the vertical profile of the temperature, the temperature below 400 hPa was improved in the simulation that adopted the cloud fraction by Hack (1998).

As mentioned above, it was confirmed that the refinements of the cloud fraction diagnostic scheme used in the radiation calculation improved the radiation flux and the vertical profile of the temperature. Improvement and effective use of model-predicted cloud variables are next subject.