We have little time to discuss problems and shortcomings in current climate models but it is good chance for me to mention a few important points. As Dr. Bin Wang summarized using Dr. Kang's paper there is too strong land heating in current models and current cumulus heating appears to be inappropriate particularly over land. One possible reason might be that the current cumulus parameterization is developed on the basis of observation data over the ocean particularly on the GATE data obtained in the Atlantic Ocean. In the Western Pacific region, as I have shown, there seems to be an unrealistic coupling between the surface boundary layer and convection processes. So more good data relevant to each process are key to solve the problems in modeling. Fortunately FORSGC together with other international programs plans to make special observation over the Western Pacific warm pool region from which good data for modeling will come out. Further over lands nonlinear interaction between sub grid-scale orography and local surface property (including vegetation and others) and local circulations must be properly taken into account, which are missing in current models. This has been clearly shown by Dr. Kimura in his high-resolution regional model experiments. So far it was dealt with only in regional models but when the Earth Simulator is realized the effects can be automatically included by use of high-resolution global models.
One more point I like to mention is that all models, even in mean climate states, have biases in precipitation which may be related with strange heating processes over land and oceans. Too strong convection and rainfall in the Indian monsoon region and too weak convection in the Western Pacific warm pool region as Dr. Bin Wang has pointed out. I also stress that realistic seasonal march is essential for simulating monsoons including Baiu. I don't think that the spatial resolution is the only problem. Maybe more realistic heating with correct seasonal march and also the role of the Tibetan plateau must be assessed. Regarding the effects of land surface processes there are many observational studies to show statistical relations between monsoon rain and snow cover and soil moisture and climate variability relation and etc, but we need to assess real physical effects of snow cover and soil moisture. For that we need good observational data to know what processes are going on between snow cover and the atmosphere. So we have lots of things to do.
Dr. Taroh Matsuno (Moderator):
We have listened to the presentations on the future plans by Program Directors. Perhaps invited panel members have already some ideas or views on the future of FRSGC, so I would like to call on the members to express your views.
