Japan Meteorological Agency
For Global Warming Projection Vol. 7, an experiment using the non-coupled NPOGCM was conducted to verify continuous variations in ocean conditions in addition to the experiment using the CRCM. This section describes the design of the experiments carried out using the CRCM, the NPOGCM and the CGCM2.3.2 to provide boundary conditions to the CRCM and the NPOGCM.
The CRCM experiment was performed using the CGCM2.3.2 fields. Two kinds of experiment were conducted: a present climate simulation for 1981 - 2000, and future climate simulations using the A1B and B1 SRES scenarios for 2081 - 2100. Although the CGCM2.3.2 experiment involved a five-member ensemble simulation, a single CRCM simulation was conducted due to the huge computational resources required.
The NPOGCM was integrated from 1981 to 2100 separately from the RCM20 using the CGCM2.3.2 atmospheric fields for the whole domain. This experiment produced continuous high-resolution simulation data for the North Pacific, enabling assessment of future climate changes in ocean areas while eliminating short-term natural climate variations through time-series analysis of NPOGCM results. The NPOGCM experiment was conducted with the A1B and B1 SRES scenarios.
The PIcntrl pre-industrial control experiment (a historical experiment for the 20th century (20C3M)) and SRES (IPCC, 2000) scenario experiments for the 21st century using the CGCM2.3.2 were conducted for the global climate.
The pre-industrial control experiment (PIcntrl) simulates climate changes using fixed forcing agents at levels from the year 1850 (e.g. CO2 = 290 ppm). It is used as reference for the following historical and scenario experiments, and also provides the initial conditions for the historical experiment.
The historical experiment (20C3M) simulates climate changes in the 20th century. The forcing agents are the historical record of (or estimated) greenhouse gases, tropospheric sulfate aerosol, volcanoes and solar forcing. 20C3M consists of five-member ensembles of simulation. The initial condition of each simulation member is taken from the different states of the PIcntrl fields at simulation years of 1, 51, 101, 151 and 201. Each member is integrated, including the 151-year period from 1850 to 2000. 20C3M is used to assess how the model reproduces past climate changes, and also provides the initial conditions for the scenario experiments. The one member whose initial condition is provided by the PIcntrl field at the first simulation year provides the boundary conditions for the CRCM and the NPOGCM experiments.
The scenario experiments simulate climate changes in the 21st century for the SRES marker scenarios (IPCC, 2000). The SRES A1B and B1 scenarios are used in this study. Each experiment consists of five-member ensembles of simulation. The initial condition of each simulation member is taken from the respective state in the year 1990 in 20C3M. Each member is integrated for 111 years from 1990 to 2100. The member for A1B (B1), whose initial condition includes the first year of the PIcntrl, provides the boundary conditions for the CRCM and the NPOGCM A1B (B1) experiments.
For further details of the experiments using the CGCM2.3.2, please refer to Yukimoto et al. (2006b).