Search

This chapter presents basic formulation of Multi-Scale Simulator for the Geoenvironment (MSSG) which is a coupled non-hydrostatic AGCM-OGCM developed in Earth Simulator Center. MSSG is characterized by Yin-Yang grid system for both of the components, computational schemes with high accuracy in the dynamical core and high computational performance on the Earth Simulator. In particular some preliminary results from 120-h forecast experiments with MSSG are presented. [ABSTRACT FROM AUTHOR]

We have been developing a global, high-resolution, coupled atmosphereocean general circulation model, named CFES, which was designed to achieve efficient computational performance on the Earth Simulator. A brief description of CFES and some preliminary results obtained from 66-month integration are presented. Although some deficiencies are apparent in the results, realistically simulated smallscale structures such as extratropical cyclones and sea surface temperature fronts in the mid-latitudes, and seasonal variation of tropical sea surface temperature and polar sea-ice extent encourage us to study mechanism and predictability of high-impact phenomena and their relation to the global-scale circulations using CFES. [ABSTRACT FROM AUTHOR]

We have developed a forecast system for the Kuroshio large meander with a high horizontal resolution (approximately 10 km). Using the system, we succeeded in predicting the path transitions of the Kuroshio from the nearshore nonlarge meander path to the offshore nonlarge meander path in 2003, and from the nearshore nonlarge meander path to the typical large meander path in 2004 as well as the occurrence of its triggering small meander south of Kyushu Island. We have also been developing a higher resolution forecast model for coastal oceans and bays south of Japan, where physical and biological states of the ocean are much affected by the path variation of the Kuroshio. This model, although the development is still under way, represents tides and tidal currents in two bays south of Japan in a realistic way. [ABSTRACT FROM AUTHOR]

Using a high-resolution ocean model (1/8 × 1/12, 85 levels) with realistic geometry and forcing, the diffusion coefficient of the isopycnal thickness is evaluated in the Southern Ocean. It is found that the coefficient is large in the Argentine Basin, the Agulhas Retroflection region, and the east of the Kerguelen Pleatue at 2000 m depth. The large coefficient regions coincide with high growth rate regions of the baroclinic instability. [ABSTRACT FROM AUTHOR]

Analysis of the output of high-resolution ocean models reveals the presence of a class of flow structures that is intermediate in scale between the large-scale circulation and the geostrophic eddy field. These structures are coherent over considerably long zonal distances and have a relatively small meridional scale of 300-500km. They take the form of either quasipersistent multiple jets or long-crested Rossby waves with high-meridional and low-zonal wavenumbers. Here we discuss the properties of these features in the Pacific Ocean, possible mechanisms for their formation, and the potential impact on the transport of tracers. [ABSTRACT FROM AUTHOR]

An eddy-resolving hindcast experiment forced by daily mean atmospheric reanalysis data covering the second half of the twentieth century was completed successfully on the Earth Simulator. The domain covers quasiglobal from 75°S to 75°N excluding arctic regions, with horizontal resolution of 0.1° and 54° vertical levels. Encouraged by high performance of the preceding spin-up integration in capturing the time-mean and transient eddy fields of the world oceans, the hindcast run is executed to see how well the observed variations in the low- and midlatitude regions spanning from intraseasonal to decadal timescales are reproduced in the simulation. Our report presented here covers, among others, the El Niño and the Indian Ocean Dipole events, the Pacific and the Pan-Atlantic decadal oscillations, and the intraseasonal variations in the equatorial Pacific and Indian Oceans, which are represented well in the hindcast simulation, comparing with the observations. The simulated variations in not only the surface but also subsurface layers are compared with observations, for example, the decadal subsurface temperature change with narrow structures in the Kuroshio Extension region. Furthermore, we focus on the improved aspects of the hindcast simulation over the spin-up run, possibly brought about by realistic high-frequency daily mean forcing. [ABSTRACT FROM AUTHOR]

High-impact weather systems occasionally cause huge disasters to human society owing to heavy rainfall and/or violent wind. They consist of cumulonimbus clouds and usually have a multiscale structure. High-resolution simulations within a large domain are necessary for quantitatively accurate prediction of the weather systems and prevention/reduction of disasters. For the simulations, we have been developing a cloud-resolving model named the Cloud Resolving Storm Simulator (CReSS). The model is designed for a parallel computer and was optimized for the Earth Simulator in the present study. The purpose of the present research is highresolution simulations of high-impact weather systems in a large calculation domain with resolving individual cumulonimbus clouds using the CReSS model on the Earth Simulator. Characteristic high-impact weather systems in East Asia are the Baiu front, typhoons, and winter snowstorms. The present chapter describes simulations of these significant weather systems. We have chosen for the case study of the Baiu front the Niigata—Fukushima heavy rainfall event on July 13, 2004. Typhoons for simulations are T0418, which caused a huge disaster due to strong wind, and T0423, which caused severe flood over the western Japan in 2004. Snowstorms were studied by an idealized numerical experiment as well as by a simulation of cold air outbreak over the Sea of Japan. These experiments clarified both the overall structures of weather systems and individual clouds. The high-resolution simulations resolving individual clouds permit a more quantitative prediction of precipitation. They contribute to accurate prediction of wind and precipitation and to reduction of disasters caused by high-impact weather systems. [ABSTRACT FROM AUTHOR]

Two applications with a cloud-resolving model are shown utilizing the Earth Simulator. The first application is a case in the winter cold-air outbreak situation observed over the Sea of Japan as a forecast mode. Detailed structures of the convergence zone (JPCZ) and formation of mechanism of transverse convective clouds (T-modes) are discussed. A wide domain in the horizontal (2000 × 2000) was used with a horizontal resolution of 1 km, and could reproduce detailed structures of the JPCZ as well as the cloud streets in the right positions. It is also found that the cloud streets of T-modes are parallel to the vertical wind shears and, thus, similar to the ordinary formation mechanism as longitudinal convective ones. The second application is changes in the Baiu frontal activity in the future warming climate from the present one as a climate mode. At the future warming climate, the Baiu front is more active over southern Japan, and the precipitation amounts increase there. On the other hand, the frequency of occurrence of heavy rainfall greater than 30 mm h-1 increases over the Japan Islands. [ABSTRACT FROM AUTHOR]

A global wanning projection was conducted on the Earth Simulator by using a very high horizontal resolution atmospheric general circulation model with 20-km grid. Tropical cyclones (TCs) and the rain band (Baiu) during the East Asian summer monsoon season are selected as the main targets of this study, because these bring typical extreme events but so far the global climate models have not given reliable simulations and projections due to their insufficient resolutions. The model reproduces TCs and a Baiu rain band reasonably well under the present-day climate conditions. In a warmer climate at the end of this century, the model projects, under the IPCC SRES A1B scenario, that the annual mean occurrence number of TCs decreases by about 30% globally (but increased in the North Atlantic) and TCs with large maximum surface winds increase. The Baiu rain band activities tend to intensify and last longer until August, suggesting more damages due to heavy rainfalls in a warmer climate. [ABSTRACT FROM AUTHOR]

A "global cloud resolving simulation" with horizontal grid interval of 3.5 km is conducted using a nonhydrostatic icosahedral atmospheric model (NICAM). NICAM is a cloud resolving model in the sense that updraft cores of deep cumulus that have a few km in horizontal size are marginally represented using explicit microphysical schemes. Results from the aqua-planet experiment of NICAM are compared with the TRMM PR (Tropical Rainfall Measurement Mission, Precipitation Radar) data that have a horizontal resolution close to the grid interval of NICAM. Precipitation statistics of NICAM are compared to those of the TRMM PR data over oceans. Probability distribution functions of rainfall rate show that relative occurrence of rainfall rate of NICAM is similar to that of the TRMM PR data for strong rains. Spectral representation of rainfall rate shows that the result of NICAM has generally higher rain-top height than that of TRMM PR. NICAM produces stronger rainfall especially for deep convective rains. Both NICAM and TRMM PR have two peaks of rainfall rate in shallow and high rain-top heights for stratiform rains. One of the advantages of using the global cloud resolving model is that the model results provide plenty of information concerning physical quantities, which are not easily retrieved from the observation, such as vertical velocity and ice phase concentrations. In particular, since the precipitation intensity of the TRMM PR data does not take account of updraft velocities of the air, the rainfall speed might be upward relative to the ground in very strong cloud cores of deep cumulus, even when the positive rainfall rate is analyzed by the TRMM PR algorithm. [ABSTRACT FROM AUTHOR]

This chapter describes the updated version of Atmospheric General Circulation Model for the Earth Simulator (AFES 2). Modifications are intended (1) to increase the accuracy and efficiency of the Legendre transform at high resolutions and (2) to improve the physical performance. In particular, the Emanuel scheme replaces a simplified version of the Arakawa-Schubert scheme for the parametrization of cumulus convection. The Emanuel scheme parametrizes O(100m) drafts within subgrid-scale cumuli and does not have explicit dependency upon the grid size. Therefore the cloud model of the Emanuel scheme allows us to use it at high resolutions of O(10km) where the validity of the ensemble cloud model of the Arakawa-Schubert scheme is questionable. Moreover, 10-year test runs indicate that the use of the Emanuel scheme improve the physical performance at a moderate resolution as well. Anomalies of the geopotential height and zonal winds in the middle to upper troposphere are reduced, although the improvements in terms of the distributions of precipitation and sea-level pressure are not significant. Improvements are attributable to a better vertical structure of temperature in the tropics due to more realistic estimation of mixing of the momentum, temperature, and moisture by the Emanuel scheme. [ABSTRACT FROM AUTHOR]

An updated description is presented for the quasi-uniform Conformal-Cubic Atmospheric Model. The model achieves high efficiency as a result of using semi-Lagrangian, semi-implicit time differencing. A reversible staggering treatment for the wind components provides very good dispersion characteristics. An MPI methodology is employed that allows the model to run efficiently on multiple processors. The physical parameterizations for the model are briefly described, and results are shown for the Held-Suarez test, the Aqua-Planet Experiment and an AMIP simulation having 125 km resolution. Antarctic snow accumulation is also shown from a shorter simulation having 50 km resolution. [ABSTRACT FROM AUTHOR]

One major challenge in obtaining useful numerical simulations of weather and climate is addressing the sensitivity of these simulations to the characteristics and distribution of clouds in the model(s). Latent heat release produced in clouds as a consequence of moist convection can dramatically affect the dynamics that govern the development of larger scale weather systems and storm tracks. Also, given the profound effects of cloud distribution on the radiative characteristics of the atmosphere, these interactions critically affect the models' climate and thus our conclusions regarding climate change. A very high resolution global model has recently been run at NOAA's Geophysical Fluid Dynamics Laboratory (GFDL) to investigate the potential value of cloudresolving numerical models to weather forecasts and climate simulations. Dubbed "Project TERRA", this experiment was conceived as an experimental 1-day simulation with GFDL's ZETAC model. [ABSTRACT FROM AUTHOR]

A review of the importance of the cyclone-frontal scale system in climate variability and the ability of present climate models to simulate them has been presented. The analysis of three different Climate models, GISS, the NCAR community climate model CCM3, and the GFDL Finite volume AM2 (M90), have been discussed. The intention here was not to determine which one is better but rather to indicate what deficiency may be common to all of them. Evidence shows that the three models tend to be deficient in the generation of cyclone wave activity with the consequences that heat, momentum, and moisture may be deficient in the extratropical and subpolar regions. This will affect cloudiness, wind stress, and precipitation. Bauer and Del Genio (2005) have shown that the deficiency of moisture and cloudiness over the subpolar regions was due to the lack of cyclone waves to transport moisture and clouds to these regions. A discussion of complementary work done on clustering of cyclone trajectories by Gaffney et al. (2005) was also presented. Consistent with the present analysis, this study also showed that differences in trajectories between reanalysis and model simulation for each cluster of trajectories were here interpreted to be related to the lack of intense high frequency eddies of the GCM. The previous two studies depend on the surface characteristics based on trajectories of the high frequency eddies. The present analysis on the GFDL-GCM is totally eulerian and based on the upper level eddy activities (300mb). However, a similar conclusion has been drawn from the analysis of the band pass frequency of energy and momentum for the GFDL AM2_M90 17 year runs, where it is quite clear that the momentum and energy of the very high frequency is much lower in the model simulation than in the reanalysis. The variance of meridional velocity also shows that the deficiency of the high frequency is in the latitude area where the reanalysis shows it to be positioned in the storm track: the model displaces it south of that. There is also a suggestion that to achieve the correct intensity of the high frequency baroclinic eddies, models should have enough resolution to resolve them, since this intensity depends on the lower level circulation of the frontal circulation system. The mesoscale circulation associated with cyclones could be adequately represented in models with resolution equal or superior to 1/4° resolution. It is clear that to adequately resolve the mesoscale, it is necessary to not only improve the resolution but also to improve the boundary layer and surface fluxes. Clearly, at the present low resolution of climate models, this improvement is probably unattainable. However, if the cloudiness and sea ice over the subpolar regions are important to the overall climate, this should be an attainable goal because no sophistication in the moist convection or sea-ice model could correct those deficiencies due to the unresolved dynamics. [ABSTRACT FROM AUTHOR]

This chapter presents a survey of published literature related to the issue of how the simulation of climate and atmospheric circulation by global models depend on numerical spatial resolution. To begin the basic question of how the zonalmean tropospheric circulation in atmospheric general circulation models (AGCMs) vary with changing horizontal and vertical grid spacing is considered. The appropriate modification of subgrid-scale parameterizations with model resolution is discussed. Advances in available computational power have recently spurred work with quite fine resolution global AGCMs, and the issue of how well such models simulate mesoscale aspects of the atmospheric circulation is considered. Experience has shown that the AGCM simulated circulation is particularly sensitive to resolution in the stratosphere and mesosphere, and so studies related to the middle atmospheric circulation are considered in some detail. Finally, the significance of atmospheric model resolution for coupled global ocean—atmosphere models and the simulated climate sensitivity to large-scale perturbations is discussed. [ABSTRACT FROM AUTHOR]