A thermal model containing time-variable parameters is constructed from an assumed velocity field and density disturbance. The fluid field is similar to Hill's spherical vortex, and the density disturbance is represented by a low-order Hermitian function series. The time variation of the model parameters is estimated by an optimal approximation method for the inviscid and non-diffusive Boussinesq equations. Initially, there is no motion but there is a low-density disturbance in the environment of neutral stratification. The time variation of the model parameters is studied with the aid of numerical integration. The time evolution of this model within a short period seems to be qualitatively comparable with some computer simulation results.
Three types of meso-β scale bands in a precipitation system over southern Kyushu on 7 July 1996 were analyzed using output from numerical models and observational data from both the Japan Meteorological Agency (JMA) and the Torrential Rain Experiment (TREX). Four meso-β scale regions comprised the precipitation region of the system. Region I was characterized by a unidirectional wind environment in which several short convective bands were organized. Convective cells in these short bands displayed characteristics of back-building cells. A hook-shaped band developed over western part of region I where intense outflow from another precipitation region intruded. Another meso-β scale band was simulated in precipitation region III. In this region, mid-level flow advected convective cells downwind as high equivalent potential temperature air entered the band from its flank, maintaining convection along the whole band length. Comparisons of airflow structures and band shapes suggest that the mid-level wind direction plays a lead role in determining the type of meso-β scale band. In addition, cold air outflow from other meso-β scale precipitation regions can influence band shape.
A three-dimensional chemical transport model(CTM) was developed at the Meteorological Research Institute by coupling a chemical module with the MJ98 general circulation model (dynamical module) for the study of stratospheric chemistry.This model, MJ98-CTM, ran for approximately 15 years and the simulated chemical species were investigated, focusing on the time-mean fields. The chemical module was based on the family method and contains major stratospheric species, i.e., 34 long-lived species including 7 families and 15 short-lived species with 79 gas phase reactions and 34 photodissociations. Two types (I and II) of polar stratospheric clouds (PSCs) and sulfate aerosols were included with six heterogeneous reactions on PSCs and three heterogeneous reactions on sulfate aerosols. MJ98-CTMs of T21L45 and T42L45 versions were integrated using climatological monthly mean values of sea-surface temperature and ozone for the dynamical module, and fixed values at the surface for surface-origin species in the chemical module. Radiatively active gases such as ozone, methane, and nitrous oxide were not treated interactively between the two modules. Horizontal resolutions were 5.6°(∼600 km) for the T21 model and 2.8°(∼300 km) for the T42 model. The vertical resolution L45 had a vertical spacing of about 2 km in the stratosphere with a top at 0.01 hPa (80km). Qualitatively, MJ98-CTM suitably reproduced the temporal and spatial features of observed ozone and other chemical species in the middle atmosphere. However, the lower stratosphere held crucial errors for the distributions of chemical species, particularly ozone, yielding positive errors for column ozone. A major cause of these errors can be ascribed to the errors associated with transport: one based on the wind field bias produced by MJ98, and the other from the coarse vertical resolution.