Development design of AGCM6
|
General Circulation Model (GCM), which is a numerical model of Earth type planetary atmosphere, has come to use also for the climate change prediction (ex. the global warming associated with increasing of carbon dioxide). Mars and Venus GCM which are developed as a branch of Earth GCM become to be able to simulate characteristic phenomenon observed in each planets. The reason for succeeding in GCM simulations is that many empirical formulas and techniques peculiar to each planets are fully utilized in GCM. GCM is thoroughly tuned up for each planets, and become to be complicated and specialized model. Therefore, following problems have come out from the viewpoint of understanding calculation results. First, it is difficult to understand the circulation structure of planetary atmosphere by use of the specialized GCMs for each planet. To understand what is the cause of different circulation structure between Earth and Mars or Venus atmosphere by use of GCMs, it is necessary to perform various parameter studies where external boundary condition (ex. solar constant), atmospheric constituents and the grand surface condition are changed, and then investigate key parameter or physical process which result in emergence of these differences. These parameter studies are difficult by use of current GCMs, however, since many empirical formulas and techniques for each planets are, so to speak, "hard coded" in current GCMs. It is not easy to perform numerical simulation with changing parameter continuously, and compare one result with that of another parameter by use of current GCMs. Moreover, current GCMs are not useful for understanding dynamics of atmospheric general circulation. To understand the dynamical structure or mechanism of atmospheric phenomenon, it is necessary to simplify the system under considering gradually, and then reveal key parameter or physical process. However, current GCMs are fully complicated softwares because of its many empirical formulas and techniques are included, and then it takes much efforts to develop a reduced system model of GCM which is reduced spatial dimension or simplified particular physical process. The complicated and specialized GCMs have closed its calculation results to observations, while the calculation results have been kept away from dynamical understanding. Current GCMs are also not useful for education of atmospheric general circulation, since they are developed as a "device" of numerical simulation. In these days when numerical computation becomes natural pass, it is necessary to master fundamental concepts of atmospheric general circulation through personal numerical experiment, as mastering physical concepts or lows of atmospheric dynamics by tracing mathematical expression. This educational method is also effective to train next GCM developers. If one can run GCM on his own computer with ease, he will understand many concepts of atmospheric general circulation intuitively through his personal experiment. However, current GCMs are only executable by supercomputer in many cases. In addition, everyone can not always enjoy GCM simulation in comfort and it is difficult to use the GCM as a textbook of initial education for next GCM developers, since documents of the GCM like a numerical code description are not prepared well. In order to hand down the knowledge of atmospheric general circulation to next generation, GCM which is executable anywhere easily and should just be used like a textbook is required. |
|
The goal of AGCM6 project is design and development of next generation GCM which is a tool for understanding and can be used as a textbook. The feature of AGCM6 is as follows.
|