An observational analysis of the low-level internal gravity waves over the Kanto Plain in Japan on 15 January 1998 was performed. The data from the Doppler radar for Airport Weather (DRAW), and aircraft soundings of wind and temperature (ACARS), provided unique observations with high spatial and temporal resolutions, and were used to analyze detailed wave structures and environmental conditions, especially wave duct structures. According to the Doppler data analysis, the horizontal wavelength was approximately 6.5km, the ground relative phase speed of the waves was approximately 4m/s, and the direction of propagation was toward 50° (NE). The wave region spread 80km-100km horizontally. Doppler data analysis revealed that the waves maintained their coherent structures, lasted for more than one hour, and propagated horizontally for long distances as an isolated wave packet. The profiles from ACARS revealed the existence of a strong stable layer with vertical wind shear. Above the stable layer, there was a near-neutral layer with a critical level, around which the Richardson number was close to 0.25. These results indicate that the waves were trapped in the duct and propagated horizontally without a large amount of energy loss in the absence of a forcing mechanism. The observed horizontal wavelength, profile of horizontal velocity amplitudes, and surface pressure perturbations are in agreement with our estimations derived from a linear theory on neutral modes. Power spectrum and bandpass filtering analyses conducted on surface pressure show that the frequency and the phase of pressure perturbations are consistent with theoretical relations with internal gravity waves, suggesting that they were caused by the passage of the internal gravity waves.
Through the use of physical initialization, which relates to precipitation and precipitable water initialization we incorporate the observed rainfall and obtained a better description of the diabatic heating from a very high-resolution global model. Interpolating assimilated data sets (basic and derived) onto isentropic surfaces we carried out detailed analysis of the potential vorticity equation for the cases, 1 December 1992 and 4 October 1995. In addition to the horizontal advection of potential vorticity, the potential vorticity Ertel's equation includes contributions from both the vertical and horizontal differentials of diabatic heating and from frictional effects. Often most of these terms are of the order of 10-12kg-1m2s-2K. A significant contribution to the local budget of potential vorticity is provided by (1) the differentials of convective heating in rain areas and (2) the radiative heating related to shallow cloud top cooling in trade wind belts. Viewing this problem over the middle troposphere (near the 325K isentropic surface) and lower troposphere (both slightly above (310K) and below (305K) the tops of undisturbed tropical stratocumulus), we find that the diabatic effects arising from differential heating contribute significantly to the generation of potential vorticity. Tropical short-range forecasts using models designed to conserve potential vorticity exhibit a very large error growth, which can be reduced significantly by including diabatic effects. Three-dimensional trajectories of air parcels in the tropical latitudes show that within 12 to 24 hours, these parcels often undergo physical processes resulting in substantial changes in potential vorticity. Non-conservation of potential vorticity appears to be important over many regions of the disturbed as well as undisturbed tropics.
How and why equatorial convections are activated in certain preferred regions during specific times of the calendar year, are investigated utilizing equatorially symmetric OLR data. In the equatorial African and American continents, semiannual variability is predominant with two peaks of convective activities in boreal spring and fall due to the in-situ radiational heating. Over the oceanic regions, the role of in-situ surface heating becomes insignificant and gives place to remote forcings in excitement of equatorial symmetric convections and associated Rossby-Kelvin wind responses within the equatorial duct between about 15°N and 15°S, that are determined by the Rossby deformation radius at the equator. The active convective phase in the equatorial western Pacific (EWP) lasts about five months from November to March in association with a systematic southward migration of the surface pressure trough. When the trough arrives at the equator in November, the zonal as well as meridional down-pressure gradient winds cause significant low-level convergence to enhance convections in EWP. Here, January is the month of most active convections, since southward down-pressure gradient winds become strongest due to equatorward penetration of the winter-time North Pacific high. During boreal winter, EWP corresponds to the updraft leg of the equatorial E-W overturning with wavenumbers 1 to 2. There exists a gigantic season-fixed clockwise phase rotation of low surface pressure across the Indian Ocean and western Pacific, namely, northward along 75°E in spring to summer, eastward at 10°N from summer to fall, southward along 155°E in fall to winter, and westward at 10°S from winter to spring, thus completing an annual journey. As such, equatorial convections in EWP are activated during the fall-winter phase of southward migration. In the equatorial Indian Ocean (EIO), convections are not really activated before and during the South and Southeast Asian summer monsoon (SEAM), since it persistently induces divergent northward down-pressure gradient winds in EIO. Here, the preferred period of active convections differs significantly with different longitudes. Between about 80° to 100°E (EIO1), October of the post SEAM season is the month of intensified convections due primarily to the convergence via the so-called β-effect. The winter-time Indian Ocean high, which penetrates equatorward along the Kenya coast, is responsible for causing a substantial west-to-east pressure gradient and convergent equatorial westerlies in EIO1. Between about 100° and 120°E (EIO2), December represents the peak convective phase under the influence of the northern hemisphere winter monsoon bursting out of Siberia. The role of this winter monsoon system is two fold, i. e.; first, accelerating equartorward down-pressure gradient winds which meridionally converge into regions of heavy convections near Sumatra and Borneo, and second, enhancing the convergence due to the β-effectin association with an increased west-to-east pressure gradient and intensified westerlies along the equator. Equatorial convections in EIO remain inactive during the northward propagation phase of low surface pressure in spring.
A family method chemical scheme, and a chemistry-radiation coupling scheme for the middle atmosphere, have been developed. These schemes are also applicable to a future atmosphere where concentrations of greenhouse gases and halogen gases will be different from the present atmosphere. A one-dimensional, chemically radiatively coupled model has been constructed to test the scheme. A method of photolysis rate calculation flexible enough for any radiation calculation scheme on plane-parallel atmosphere, a definition of globally averaged solar zenith angles for the 1-D model, and development of a chemical scheme to which heterogeneous reactions on aerosols or Polar Stratospheric Clouds are easy to incorporate are presented. The model includes 163 gas phase chemical reactions of oxygen, hydrogen, nitrogen, hydrocarbon, chlorine, and bromine species for the stratosphere. Although heterogeneous reactions are not considered in this paper, the schemes are flexible for including these reactions. Vertical profiles of concentrations and photolysis rates of chemical constituents at a steady state of the model, and the temperature profile are presented and discussed. Comparison of these profiles with reference profiles in JPL-97, shows that they were successfully calculated by this scheme in the model. Numerical experiments on CO2 doubling with the 1-D model shows that total ozone variation after CO2 doubling is different between the coupled model, and in a fixed photolysis rate model. Mechanism of the slow, extended variation of total ozone in the coupled model is discussed. It is shown that reduction in solar radiation penetration into the atmosphere by O3 increase due to the CO2 doubling, made considerable effects on chemical constituent concentrations in the lower stratosphere.
From in-situ meteorological and ice observation in early February of 1996 and 1997, we estimate the heat budget in the ice cover of the southern Okhotsk Sea. Ice concentration and ice thickness distribution required for its calculation are obtained quantitatively from video analysis. One dimensional-thermodynamical model is used to calculate heat flux. The total heat flux is obtained by summing up the areally weighted heat flux on each ice thickness. As a result, the following characteristics are found in this region. (1) Due to relatively thin ice thickness, turbulent heat flux (sensible+latent) is upward, showing that the sea ice region gives turbulent heat to atmosphere. (2) Thin ice and open area contribute significantly to the total turbulent heat flux over the whole area. (3) Thermodynamical ice growth is limited to below 1cm per day on average. The first and third results are the characteristics of this area located at a relatively low latitude, while the second one is generally observed for polar regions. The results also suggest that surface melting occur at daytime mainly due to solar radiation. The salinity of sea ice samples shows the values lower than that of the first-year ice in the polar regions and rather similar to that of the multi-year ice, which is lowered by surface melting during summer; this may support this suggestion.
Comparison of strato-mesospheric O3 derived from ground-based microwave radiometry at Nagoya University with Umkehr measurements at Tsukuba by JMA (Japan Meteorological Agency) has been carried out for evaluation of microwave measurements. Good agreements between radio and Umkehr measurements have been found from Umkehr layer 6-9, which correspond to the pressure of 15.8-0.99hPa. Layer ozone amounts in layer 10 from microwave measurements show short term enrichment in winter with time scale of about a week, which are not observed by Umkehr measurements. The difference is due to the saturation effect inherent in Umkehr measurements. Fourier coefficients, up to the 2nd harmonic of the seasonal variations of nighttime and daytime mixing ratios, are given in tables for the pressure P=0.013, ..., 10hPa (log P=-1.875, ..., 1.000). QBO type variations of O3 in 1-0.1hPa range described in our earlier paper are due to the superposition of the short term enrichment. The short term enrichment was especially remarkable in 1992-1993 and 1994-1995. The short term enrichment does not correlate with temperature. A plausible cause of the enrichment will be vertical circulation anomalies as is the case of QBO variations in O3 at the lower-stratosphere, and in NO2 at the middle-stratosphere proposed by Randel and Wu (1996).
In this study, in situ observations of wind holes at the Ice Valley in Korea and Nakayama in Fukushima, Japan are carried out. Based on the observational results, we conducted a series of numerical model simulations of the wind hole circulation using a zero dimensional model, one-dimensional model along the flow path, and two dimensional model in the vertical cross section to understand the detail of the wind hole circulation. As a result of the observations and numerical experiments, it is found that: (1) the main driving force of the wind hole circulation is the horizontal pressure gradient force induced by the temperature contrast between outer air and interior talus; (2) the mean wind hole circulation is about 1mm/s and the residence time of air in the tales is about 2 days; (3) cold wind hole circulation of katabatic wind in spring to summer is replace by warm wind hole circulation of anabatic wind in fall to winter; (4) contrasted with the stable and steady katabatic flow in summer, the unstable stratification in winter causes convective overturning of warm and cold air at the Ice Valley with no vegetation along the slope. This seasonal asymmetry of the wind hole circulations acts as a thermal filter which accumulates only the coldness in the talus. It is interesting to note that the hotter the outer air during spring is, the stronger the freezing katabatic wind is. This mechanism, in part, explains the mysterious ice formation during the hottest season at the Ice Valley and Nakayama.
To investigate the maintenance mechanism of the blocking flow, we performed a numerical experiment using a barotropic spectral model on the sphere. The model is linearized about two types of realistic basic flows. One is a blocking flow near Alaska (blocking flow case) and the other is a zonal flow in the North Pacific (zonal flow case). The model is integrated in time with prescribed high-frequency eddies which are generated in the Far East and propagate eastward along the northern flank of the jet stream. After several days of time integration, time mean vorticity flux divergence is calculated. Then, the resulting vorticity flux divergence field is used as the eddy vorticity forcing to compute the second-order flow induced by eddies. It is found that the induced second-order flow intensifies the blocking for the blocking flow case, and that this is not the case for the zonal flow case. To clarify the reason for the difference between the induced second-order flows, we carried out the singular value decomposition (SVD) of the matrix of the barotropic vorticity equation linearized about the basic flows. For the blocking flow case, spatial distributions of the leading singular modes show blocking patterns which are similar to the blocking in the basic flow. It is found that these modes are effectively excited by the eddy vorticity forcing, and enhance the blocking in the basic flow. On the other hand, for the zonal flow case, leading singular modes do not show blocking patterns. These results indicate that blocking flows tend to have easily excited modes that can reinforce the block, and high-frequency eddies could maintain blocking by exciting these modes. The dependency of results of the experiment on the location of high-frequency eddies is also shown. It indicates the limitation of the model to investigate the high-frequency eddies feedback.
Impacts of small scale ocean disturbances induced by atmospheric precipitation on ocean mixed-layer heat and salt budgets are investigated by analyzing the differences in ocean mixed-layers between 1-D and 2-D ocean simulations using a two-dimensional coupled ocean-cloud resolving atmosphere model. The cloud resolving atmosphere model of coupled system is forced by large-scale atmospheric vertical velocities derived from the TOGA COARE observations during a selected seven-day period. Coupled experiments with 1-D and 2-D ocean models show that differences in the horizontal-mean mixed-layer salinity and temperature could be as large as 0.3 PSU and 0.4°C respectively. Shallow mixed layers over the convective areas due to the fresh water flux and deep mixed layers over the convection-free areas due to heat loss result in moderate deepening of the horizontal-mean mixed layers during nighttime in the 2-D ocean simulation while the persistent rainfall maintains shallow mixed layers in the 1-D ocean simulation. Such mixed-layer depth differences are responsible for the differences in saline entrainment and cooling rates, and thus differences in salinity and temperature. Heavy rainfall occurs over convective areas embedded in broad non-convective or clear areas, whereas diurnal signals over whole model areas yield high spatial correlation of surface heat flux. Thus, fresh water flux exhibits larger spatial fluctuations than surface heat flux. As a result, mixed-layer salinities contribute more to the density differences than do mixed-layer temperatures.
Combined tropical precipitation of Global Precipitation Climatology Project (GPCP) version 1 is compared with the assimilated precipitation obtained from the European Center for Medium-Range Weather Forecast (ECMWF), the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR), and the NASA Goddard Earth Observing System (GEOS-1) reanalysis projects. The analysis focuses on the domain within 30°S-30°N, and the period is from July 1987 to December 1993. Annual-mean precipitation shows that the ECMWF overestimates the precipitation over tropical oceans as compared to the GPCP. On the other hand, the NCEP/NCAR and NASA-GEOS-1 underestimate precipitation of the Inter-Tropical Convergence Zone (ITCZ) and South Pacific Convergence Zone (SPCZ). Horizontal distributions of the difference in intra-annual and inter-annual precipitation variations between the GPCP and the three reanalysis datasets are similar to those of the difference in annual-mean precipitation. Overall, the ECMWF exhibits the highest standard deviation of precipitation over tropical oceans, followed by the GPCP, NASA-GEOS-1, and NCEP/NCAR. The correlation features of the area-averaged precipitation with the sea surface temperatures over the eastern equatorial Pacific Ocean are also compared among the four products, and their differences are discussed.
The Japan Meteorological Agency has been monitoring carbon dioxide (CO2) concentrations in the atmosphere continuously with non-dispersive infrared gas analyzers since January 1987 at Ryori, March 1993 at Minamitorishima, and January 1997 at Yonagunijima. At Ryori with our longest record, a large variation was found in CO2 growth rate in association with significant global events such as El Niño and the eruption of Mt. Pinatubo, suggesting that global changes in the carbon cycle are closely connected with the climate. The annual growth rates in CO2 concentration were 3.0ppm at Ryori, 2.8ppm at Minamitorishima, and 3.1ppm at Yonagunijima in 1998. At Ryori and Minamitorishima, they were the largest since the beginning of measurement. It is supposed that this large growth rate is attributed to the strong El Niño event in 1997/98.