A hollow-fiber membrane chamber (HFMC) was developed as an in situ cultivation device for environmental microorganisms. The HFMC system consists of 48 to 96 pieces of porous hollow-fiber membrane connected with injectors. The system allows rapid exchange of chemical compounds, thereby simulating a natural environment. Comparative analysis through the cultivation of three types of environmental samples was performed using this newly designed device and a conventional agar-based petri dish. The results show that the ratios of novel phylotypes in isolates, species-level diversities, and cultivabilities in HFMC-based cultivation are higher than those in an agar-based petri dish for all three samples, suggesting that the new in situ cultivation device is effective for cultivation of various environmental microorganisms.

A new method is proposed to estimate three-dimensional (3D) material circulation driven by waves based on recently derived formulas by Kinoshita and Sato that are applicable to both Rossby waves and gravity waves. The residual-mean flow is divided into three, that is, balanced flow, unbalanced flow, and Stokes drift. The latter two are wave-induced components estimated from momentum flux divergence and heat flux divergence, respectively. The unbalanced mean flow is equivalent to the zonal-mean flow in the two-dimensional (2D) transformed Eulerian mean (TEM) system. Although these formulas were derived using the “time mean,” the underlying assumption is the separation of spatial or temporal scales between the mean and wave fields. Thus, the formulas can be used for both transient and stationary waves. Considering that the average is inherently needed to remove an oscillatory component of unaveraged quadratic functions, the 3D wave activity flux and wave-induced residual-mean flow are estimated by an extended Hilbert transform. In this case, the scale of mean flow corresponds to the whole scale of the wave packet. Using simulation data from a gravity wave–resolving general circulation model, the 3D structure of the residual-mean circulation in the stratosphere and mesosphere is examined for January and July. The zonal-mean field of the estimated 3D circulation is consistent with the 2D circulation in the TEM system. An important result is that the residual-mean circulation is not zonally uniform in both the stratosphere and mesosphere. This is likely caused by longitudinally dependent wave sources and propagation characteristics. The contribution of planetary waves and gravity waves to these residual-mean flows is discussed.

The transformed Eulerian-mean (TEM) equations are useful in examining how the generation and/or dissipation of atmospheric waves drives the mean meridional circulation. However, the TEM equations do not provide a three-dimensional view of the transport. Several previous studies extended the TEM equation system to three dimensions but usually under the quasi-geostrophic assumption, which excludes small-scale phenomena such as gravity waves. Miyahara recently derived three-dimensional wave activity flux and the corresponding residual circulation applicable to gravity waves. However, his formulation has two flaws. First, the three-dimensional residual mean circulation does not satisfy the continuity equation. Second, the Eulerian-mean flow appears in the advection terms and the residual circulation appears in the Coriolis force term of the zonal momentum equation, unlike in the TEM one. The present study developed theoretical formulae of a three-dimensional residual mean circulation and wave activity flux on the basis of primitive equations that overcome these flaws. It is confirmed that the three-dimensional residual mean circulation accords with the sum of the Eulerian time-mean flow and the Stokes drift and that the three-dimensional wave activity flux accords with the mean tangential forces across material surfaces corrugated by the waves under an assumption similar to the TEM equations. A simple physical meaning is given for the terms including the shear of time-mean flow in the three-dimensional wave activity flux. Moreover, the time mean tracer transport equation is derived using the three-dimensional residual mean circulation. A simple case study using the new formulae was made on the three-dimensional transport of stratospheric ozone in the Southern Hemisphere. It is shown that the product of the Coriolis parameter and the strong poleward/equatorward Stokes drifts also balances the divergence/convergence of the three-dimensional waveactivity flux.

A companion paper formulates the three-dimensional wave activity flux (3D-flux-M) whose divergence corresponds to the wave forcing on the primitive equations. However, unlike the two-dimensional wave activity flux, 3D-flux-M does not accurately describe the magnitude and direction of wave propagation. In this study, the authors formulate a modification of 3D-flux-M (3D-flux-W) to describe this propagation using small-amplitude theory for a slowly varying time-mean flow. A unified dispersion relation for inertia–gravity waves and Rossby waves is also derived and used to relate 3D-flux-W to the group velocity. It is shown that 3D-flux-W and the modified wave activity density agree with those for inertia–gravity waves under the constant Coriolis parameter assumption and those for Rossby waves under the small Rossby number assumption. To compare 3D-flux-M with 3D-flux-W, an analysis of the European Centre for Medium-Range Weather Forecasts (ECMWF) Interim Re-Analysis (ERA-Interim) data is performed focusing on wave disturbances in the storm tracks during April. While the divergence of 3D-flux-M is in good agreement with the meridional component of the 3D residual mean flow associated with disturbances, the 3D-flux-W divergence shows slight differences in the upstream and downstream regions of the storm tracks. Further, the 3D-flux-W magnitude and direction are in good agreement with those derived by R. A. Plumb, who describes Rossby wave propagation. However, 3D-flux-M is different from Plumb’s flux in the vicinity of the storm tracks. These results suggest that different fluxes (both 3D-flux-W and 3D-flux-M) are needed to describe wave propagation and wave–mean flow interaction in the 3D formulation.