In order to evaluate the model performance in simulating the Northeast Asian summer climate, and to investigate the effect of global warming on the summer climate over the Northeast Asian region, the multi-model ensemble of eight atmosphere-ocean coupled general circulation models in the historical (20C3M) and the scenarios (A2, A1B, B1) runs are analyzed, which is participating to the Intergovern mental Panel on Climate Change fourth Assessment Report (IPCC AR4). From comparison of the ob servation and the 20C3M experiment, it is found that the multi-model ensemble quite well simulates the Northeast Asian summer precipitation and circulation, especially in the rst two empirical orthogonal function (EOF) modes and the associated regressed eld. The rst EOF mode represents the decaying phase of El Nino and Southern Oscillation (ENSO), which contributes to the development of the Philip pines Sea anticyclone. The second one is associated with the fast transition of ENSO. In future climate, the increase of the precipitation to 2099 at A2, and A1B simulation reaches 10% com pared with the mean precipitation for 1961-1990 over the Northeast Asian region. After the stabilization of the greenhouse gas concentration in 2100, the precipitation is enhanced during 30 or 50 years more due to the inertia inherent in the climate system. From EOF analysis, it seems that the increased Northeast Asian summer precipitation due to global warming is contributed by the effect of the enhanced monsoon circulation in the decaying phase of El Nino rather than the mean linear increase of global climate or the circulation in the fast transitional period of ENSO.
Using the European Center for Medium-Range Weather Forecasts (ECMWF) reanalysis (ERA-40) data, we examined two summertime teleconnection patterns, the Europe-Japan (EJ) and the Pacic-Japan (PJ) patterns, which prevail over northern Eurasia and the western North Pacic, respectively, and high lighted the effects of the combination of the two on the East Asian summer monsoon variability on a high frequency (HF) component (sub-monthly timescales) and a low-frequency (LF) component (longer than about one month). The combination of positive EJ type 2 (EJ2+), which particularly accounts for the variability of the Ok hotsk high among the two types of EJ pattern (EJ type 1 and EJ type 2), and negative PJ (PJ-) patterns establishes a noticeable tripole structure in the mid-tropospheric geopotential height eld of East Asia that is characterized by two positive anomalies over the Okhotsk Sea and the western North Paci c to the north of the Philippines and a negative anomaly over Japan. It is found that the combination of both pat terns, EJ2+/PJ-, leads to a larger surface temperature decrease in the northeastern part of East Asia than a single teleconnection pattern alone, having a substantial impact on the East Asian summer monsoon circulation. The LF-combination also induces signicant cool sea surface temperature anomalies in the vicinity of Japan. In the HF-combined case, the time evolution of a positive height anomaly over the Ok hotsk Sea considerably differs from that of the LF-combined one. The HF-anomaly is extremely displaced eastward to the Aleutian Islands a few days after the peak day and occupies part of a newly organized wavetrain pattern across the central North Pacic. Further development of the wavetrain pattern may be attributed to anomalous convective heating related to enhanced synoptic-scale disturbances to the east of Japan. Both the LF-EJ2+ and LF-PJ- patterns tend to appear frequently during the post-1980 period, and, as a result, the number of LF-combined cases may have increased in the last two decades. Exceptionally cool summers with signicant amounts of rainfall, particularly in Japan, such as the 1993 and 2003 summers, often arise from such an LF-combination.
This paper describes the interannual and submonthly variability (disturbances with periods of less than 1 month) of the North Pacic subtropical high (NPSH) and the characteristics of its westward extension and eastward retreat in June, July, and August from 1979 to 2001. The study was based on 6-hourly data from the European Center for Medium-Range Weather Forecasts 40-Year Reanalysis with T106 resolution, Climate Prediction Center Merged Analysis of Precipitation data, and typhoon track data provided by the Japan Meteorological Agency. In the western Pacic, the interannual and submonthly variability of the NPSH were smallest in June and largest in August. To examine the characteristics of the westward ex tension and eastward retreat of the NPSH, an NPSH index was dened as the monthly mean anomaly of geopotential height (Z) at 850 hPa averaged over the western edge (125°-150°E, 17°-32°N) of the NPSH. Using this index, ve extreme years of westward extension and ve of eastward retreat were extracted to represent positive and negative years, respectively. Composite differences were calculated by subtracting the values of negative years from those of positive years. Composite analyses based on the NPSH index revealed several characteristics of the NPSH and its surroundings, including large-scale circulation, sta tionary Rossby waves and small-scale disturbances. In June and July, Baiu frontal activity, including meso α-scale disturbances, was stronger when the NPSH extended westward (in positive years). In positive (negative) years, most typhoons occurred in the western Pacic (western to mid-Pacic). The composite difference of submonthly variability of Z at 850 hPa in August indicated a broad distribution of negative anomalies over the western Pacic; the large-scale horizontal structure of these anomalies was similar to that for the composite difference of monthly mean Z at 850 hPa. The NPSH index and submonthly vari ability of Z at 850 hPa in the index area were signicantly negatively correlated.
The distributions of precipitation, radar reectivity, and vertical motion associated with tropical cy clones frequently exhibit a signi cant degree of wavenumber-one asymmetry within the intense inner core. Recent observational and numerical studies attributed the convective asymmetry to the ambient vertical wind shear in which the storms were embedded. However, the precise mechanisms by which the environmental shear controls the convective asymmetry are not yet fully established. Although a previous numerical study by the author suggested that a thermal-wind adjustment to the shear-induced vortex tilt may be a primary mechanism responsible for the asymmetry, the reason why the rainfall maxima were observed predominantly in the downshear-left quadrant (looking downshear) in the Northern Hemisphere rather than right downshear remains to be answered. In the present study, an analytical formula for the amplitude of shear-induced vertical motion asymme try developed in the previous study based on the proposed mechanism was applied to the numerical simu lation data obtained for Typhoon Chaba in 2004. As a result, high correlation coefcients were calculated between simulated and formula-predicted vertical motion asymmetries, especially at low levels where the diabatic enhancement of upward motion was not so signicant, conrming that the proposed mechanism may be a primary triggering mechanism for the vertical motion asymmetry. In the study, to discover the processes that govern the directional preference for the rainfall maxima, a Lagrangian trajectory approach was applied to the simulation data. The results suggested that asym metric water vapor distribution caused by shear-induced vertical motion was a primary factor to locate the rainfall maxima on the downshear-left side rather than right downshear of the storm center. In addition, it was found that the near-collocation in azimuth of upward motion maxima and vortex tilts, which was com monly observed in this and other simulations, was inextricably linked to the formation mechanism for the asymmetric water vapor distribution.
A basic ow, which is symmetric in one of the horizontal direction (say, y direction), is unstable for y-symmetric disturbances when the basic potential vorticity is negative (so called symmetric instability). The temporal dependence of the growing modes is exponential, that is, eσt. In the case of vanishing growth rate, that is, σ → 0, there emerge degenerate modes as shown by Xu (2007). Under the thermal wind-balanced initial condition, Xu (2007) showed that the temporal dependence of the degenerate modes is linear, that is, they grow as a linear function of time. In this note, we show that in general the degener ate modes grow quadratically in time, and that in particular they are reduced to the result of Xu (2007), (i.e., linear growth in time) under the thermal-wind-balanced initial conditions.
We compared features of the precipitation associated with subsynoptic-scale Baiu frontal depression simulated by an atmospheric general circulation model (AGCM; T106L56: a spectral primitive-equation model with 56 σ levels and triangular spectral truncation at zonal wavenumber 106) with the features had been described in observational studies. The 21-year model integration from 1979 to 1999 was constrained by observed sea-surface temperature and sea-ice distribution. As typical examples, this paper examines simulation cases for June 1991. Comparisons with past observational results showed that the AGCM properly simulated the Baiu front and the Baiu frontal precipitation in the averaged elds for a 15-day period in June 1991. However, the area of the largest 1-hour precipitation is simulated at about 500 km northeastward from the area of the largest averaged precipitation. We also found signicant differences between the simulations and the ob servations when we examined the daily precipitation and the maximum 1-hour precipitation for each day simulated by the model. While precipitation associated with subsynoptic-scale Baiu frontal depressions was relatively well simulated, intense precipitation in the trailing portion of these depressions was signi cantly underestimated. The result of the present study indicates that the ability of the AGCM to reproduce such extreme pre cipitation events must be examined in detail by comparing simulated daily and hourly precipitation with observational studies as well as by statistical analyses.