The atmospheric component of the CGCM used in the present study is the global spectral model that was previously used as a forecasting model at the Japan Meteorological Agency. In the horizontal direction, all variables are truncated at total wave number 213 (T213). There are 21 levels in the vertical from the surface to approximately 10 hPa. The model includes comprehensive physical parameteriza-tions: long wave radiation (Sugi et al. 1990), short wave radiation (Lacis and Hansen 1974), planetary boundary layer (Mellor and Yamada 1974), surface fluxes (Louis et al. 1982), SiB as a land surface model (Sellers et al. 1986; Sato et al. 1989), and gravity wave drag (Iwasaki et al. 1989). The cumulus convection scheme is based on the mass flux scheme proposed by Arakawa and Schubert (1974). The scheme is modified for entrainment rate (Moorthi and Suarez 1992) and for determination of mass flux at cloud base (Randall and Pan 1993; Lord et al. 1982). The cloud work function for each cloud type is relaxed back to a critical value taken from Lord and Arakawa (1980) over a fixed time scale. Detailed descriptions of the physical processes are given in Sugi et al. (1990) and in Kuma (1996).
The ocean component is based on the Southampton-East Anglia model (e.g., Killworth et al. 1991). The model domain covers all oceans except the Arctic Ocean, where the observed climatological sea surface temperatures (SSTs) and sea-ice distributions are prescribed as the boundary conditions of the atmospheric model. The horizontal resolution is 0.5625° in both longitude and latitude. There are 37 levels in the vertical direction, with the upper 400 m divided into 25 levels, making it well suited for investigating the upper layer of ocean. The coefficient of vertical mixing depends on the Richardson number, as in the parameterization proposed by Pacanowski and Philander (1981). The coefficients for both horizontal eddy viscosity and diffusivity are 1 x 107 cm2 s_1. The penetrative solar radiation is parameterized by using the formula of type I given in Paulson and Simpson (1977). The convection scheme introduced by Marotzke (1991) is used to immediately eliminate unstable conditions. The process for sea-ice and runoff is not considered in the ocean model.
The initial conditions for the atmospheric model were taken from the National Center for Environmental Prediction-National Center for Atmospheric Research (NCEP-NCAR) global atmospheric reanalysis data (Kalnay et al. 1996). Those of the ocean model were the state of rest with annual mean temperature and salinity distributions (Levitus 1982). Under these conditions, the atmospheric model and the ocean model were coupled through daily mean SSTs and surface fluxes. The CGCM was integrated for 100 years without flux correction, and we analyzed the results during a period from year 30 to year 100.
In order to validate the relationship between ENSO and TC activities in the CGCM, we use the best track data compiled by the National Hurricane Center, the NCEP-NCAR reanalysis, and the extended reconstructed sea surface temperature (ERSST) ver.2 (Smith and Reynolds 2004).
Was this article helpful?