A new global climate model, MRI-CGCM3, has been developed at the Meteorological Research Institute (MRI). This model is an overall upgrade of MRI's former climate model MRI-CGCM2 series. MRI-CGCM3 is composed of atmosphere-land, aerosol, and ocean-ice models, and is a subset of the MRI's earth system model MRI-ESM1. Atmospheric component MRI-AGCM3 is interactively coupled with aerosol model to represent direct and indirect e¤ects of aerosols with a new cloud microphysics scheme. Basic experiments for pre-industrial control, historical and climate sensitivity are performed with MRI-CGCM3. In the pre-industrial control experiment, the model exhibits very stable behavior without climatic drifts, at least in the radiation budget, the temperature near the surface and the major indices of ocean circulations. The sea surface temperature (SST) drift is sufficiently small, while there is a 1 W m À2 heating imbalance at the surface. The model's climate sensitivity is estimated to be 2.11 K with Gregory's method. The transient climate response (TCR) to 1 % yr À1 increase of carbon dioxide (CO 2 ) concentration is 1.6 K with doubling of CO 2 concentration and 4.1 K with quadrupling of CO 2 concentration. The simulated present-day mean climate in the historical experiment is evaluated by comparison with observations, including reanalysis. The model reproduces the overall mean climate, including seasonal variation in various aspects in the atmosphere and the oceans. Variability in the simulated climate is also evaluated and is found to be realistic, including El Niñ o and Southern Oscillation and the Arctic and Antarctic oscillations. However, some important issues are identified. The simulated SST indicates generally cold bias in the Northern Hemisphere (NH) and warm bias in the Southern Hemisphere (SH), and the simulated sea ice expands excessively in the North Atlantic in winter. A double ITCZ also appears in the tropical Pacific, particularly in the austral summer.

The new Meteorological Research Institute Earth System Model version 2.0 (MRI-ESM2.0) has been developed based on previous models, MRI-CGCM3 and MRI-ESM1, which participated in the fifth phase of the Coupled Model Intercomparison Project (CMIP5). These models underwent numerous improvements meant for highly accurate climate reproducibility. This paper describes model formulation updates and evaluates basic performance of its physical components. The new model has nominal horizontal resolutions of 100 km for atmosphere and ocean components, similar to the previous models. The atmospheric vertical resolution is 80 layers, which is enhanced from the 48 layers of its predecessor. Accumulation of various improvements concerning clouds, such as a new stratocumulus cloud scheme, led to remarkable reduction in errors in shortwave, longwave, and net radiation at the top of the atmosphere. The resulting errors are sufficiently small compared with those in the CMIP5 models. The improved radiation distribution brings the accurate meridional heat transport required for the ocean and contributes to a reduced surface air temperature (SAT) bias. MRI-ESM2.0 displays realistic reproduction of both mean climate and interannual variability. For instance, the stratospheric quasi-biennial oscillation can now be realistically expressed through the enhanced vertical resolution and introduction of non-orographic gravity wave drag parameterization. For the historical experiment, MRI-ESM2.0 reasonably reproduces global SAT change

Simulations of climate, including atmospheric chemistry and carbon cycle, are conducted for the period from 1850 through 2100 with a new earth system model (ESM) of the Meteorological Research Institute (MRI), MRI-ESM1. This model has been developed as an extension of the atmosphere-ocean coupled general circulation model, MRI-CGCM3, by adding chemical and biogeochemical processes. The dynamic and thermodynamic processes are entirely the same in both models. The horizontal resolution of MRI-ESM1 is higher than that of the ozone model that handles chemical processes which require high computational cost. In the control experiment, it is confirmed that the climatic drift of the model is insignificant with regard to surface air temperature (SAT), the radiation budget, and trace gas (carbon dioxide (CO 2 ) and ozone) concentrations. Compared with a control experiment by MRI-CGCM3, SAT is slightly higher because of a higher tropospheric ozone concentration. The performance of MRI-ESM1 is validated by conducting a historical experiment. Overall, MRI-ESM1 simulates well observed historical changes in SAT and trace gas concentrations. However, increases in the SAT and atmospheric CO 2 concentration are underestimated, associated with a positive feedback process through soil respiration. Namely the underestimation of atmospheric CO 2 increase causes weak SAT rise which makes soil respiration inactive, and then the excess of net land surface CO 2 uptake suppresses increase of the atmospheric CO 2 concentration. The simulated present-day climate states of SAT, radiation fluxes, precipitation, and trace gas concentrations are also in good agreement with observations although there are errors of radiations, precipitation, and ozone concentration especially over the southern tropical Pacific in the simulation. These errors appear to be originated from the excess of convective activity: so-called double ITCZ (intertropical convergence zone). Both MRI-ESM1 and MRI-CGCM3 are similarly able to reproduce the present-day climatology. In future projections, the global mean SAT rise at the end of the 21st century relative to the pre-industrial era is 3.4°C in the experiment using the Representative Concentration Pathways 8.5 (RCP8.5) by MRI-ESM1, whereas it is 4.0°C in the MRI-CGCM3 experiment.The atmospheric CO 2 concentration projected by MRI-ESM1 for the end of the 21st century is about 800 ppm, which is lower by 130 ppm than that prescribed in the RCP8.5 experiment with MRI-CGCM3. This difference is consistent with the above-mentioned difference in the SAT rise between MRI-ESM1 and MRI-CGCM3. The global mean total column ozone increases by about 25 DU during the period from 2000 to 2100, which is comparable to that prescribed in the experiment with MRI-CGCM3. It is also investigated how aerosols simulated with a newly introduced aerosol-chemistry process influence weak SAT rise at the end of the 20th century in MRI-ESM1.

A series of ensemble reforecast experiments is conducted to investigate the predictability and the occurrence mechanism of a stratospheric sudden warming occurred in late January 2009, which is a typical polar vortex splitting event. To fully examine the rapid vortex splitting evolution and predictability variation, ensemble forecasts are carried out every day during January 2009. The vortex splitting event is reliably predicted by forecasts initialized after 6 days prior to the vortex breakup. It is also found that the propagating property of planetary waves within the stratosphere is a key to the successful prediction for the vortex splitting event. Planetary waves incoming from the troposphere are reflected back into the troposphere for failed forecasts, whereas they are absorbed within the stratosphere for succeeded forecasts. Composite analysis reveals the following reflection process of planetary waves for the failed forecast: Upward propagation of planetary wave activity from a tropospheric blocking over Alaska is weaker during initial prediction periods; then, the deceleration of the zonal wind in the upper stratosphere becomes weaker over Europe, which produces a preferable condition for the wave reflection; hence, subsequently incoming wave activity from the troposphere over Europe is reflected back over the Siberia inducing the eastward phase tilt of planetary waves, which shuts down the further upward propagation of planetary waves leading to the vortex splitting. Thus, this study shows that the stratospheric condition would be another important control factor for the occurrence of the vortex splitting event, besides anomalous tropospheric circulations enforcing upward propagation of planetary waves.It is also well known that SSWs are classified into the following two types based on the synoptic structure of the stratospheric circulation: "vortex displacement" type characterized by a shift of the polar vortex off the pole and "vortex splitting" type in which the polar vortex breaks up into two pieces [Charlton and Polvani, 2007]. It has been also revealed by several studies based on the composite of reanalysis data sets that the NOGUCHI ET AL.PREDICTABILITY OF VORTEX SPLITTING EVENT 3388

The cloud top height of marine boundary layer clouds (MBLCs) in the mid-latitudes has received less attention than that of subtropical MBLCs and is investigated here using cloud mask data, which were based on observations from the cloud-aerosol lidar and infrared pathfinder satellite observation (CALIPSO) satellite. This study provides a comprehensive overview of the observational characteristics of variations in cloud top height of MBLCs and fog frequency over the mid-latitudes. Seasonal variations in the cloud top height of mid-latitude MBLCs as well as the differences in these seasonal variations between the Northern and Southern hemispheres were determined. For example, over the North Pacific, the cloud top height is high in winter (up to 1800 m) but low in summer (down to 800 m), whereas in the Southern Hemisphere, the seasonal variation is not well defined, with heights ranging from 1300 to 1500 m. While clear seasonal variations in the frequency of fog occurrence are found over the North Pacific and the northwest Atlantic, the fog frequency over the Southern Ocean is almost constant irrespective of the season. High correlations were found between the MBLC top height and stability indexes and between the fog frequency and some surface parameters such as temperature difference between the surface air and the sea surface. The latitudinal variations in the cloud top height of MBLCs in summer and winter over the Southern Ocean were compared with those over the North Pacific. The difference in cloud top heights between nighttime and daytime is also presented.