A simplified SST equation for the El Niño-Southern Oscillation (ENSO) is derived based on the budget analysis of the SST anomalies. The analysis is done using both advective and flux forms of the SST anomaly equation and ocean reanalysis data set from the National Centers for Environmental Prediction. The main contributing terms to the SST variation are the zonal advection by anomalous current and the anomalous meridional and vertical heat flux divergences by climatological-mean currents. Note that the anomalous upwelling term associated with the Bjerknes’ coupled instability is relatively small in most of the tropical Pacific. It is also shown that in the equatorial Pacific, the vertical heat flux term can be parameterized in terms of the thermocline depth anomaly and the meridional flux term can be effectively included in the Newtonian cooling term. The meridional flux term plays an important role to drain the heat in the ocean surface layer from the equator into the off-equator and as a result, to shape the meridional structure of SST anomaly. It is suggested that the narrow meridional structure of ENSO SST anomalies simulated by current coupled GCMs can be related to the weak simulated mean meridional current.
Since 1993, CO2 concentration and its flux between the atmosphere and a temperatre forest ecosystem have been examined in the central part of Japan (the Takayama site). Daily variation of CO2 concentration at the Takayama site during three days in July was compared with the results of a numerical simulation, in which both the CO2 flux of anthropogenic origin and that from the ecosystem were considered. The CO2 flux from the ecosystem was estimated from the data at the Takayama site and extended to all the central part of Japan with the aid of a vegetation map. When the CO2 flux at night estimated from the observation at the top of a 27 m tower at Takayama site was used as the grid average in the model, the amplitude of CO2 variation obtained from the calculation was smaller than the observed amplitude at the Takayama site, which is located at the top of a hill. The anthropogenic contamination of CO2 was also investigated. The simulation suggested the possibility of long-range transport of CO2 from Nagoya and Tokyo, two cities located about 150 km south and 300 km southeast of the Takayama site. The effect of these large cities was greater than that of nearby Takayama City.
Investigating the variability of tropical cyclone (TC) activity in the western North Pacific (WNP) during the 49-yr period 1951-99, interdecadal variation is confirmed. Two kinds of periods are identified in this study: High Frequency Period (HFP) when TC activity is enhanced, and Low Frequency Period (LFP) when TC activity is reduced. Significant differences in the number of TCs between HFP and LFP are found in the typhoon season from July through October. The differences in TC activity between HFP and LFP are also found in the areal extent of TC genesis. The area of TC genesis in HFPs extends more to the east than that in LFPs. Seasonal conditions of sea surface temperature (SST), relative vorticity at 850hPa, divergence at 200 hPa and outgoing longwave radiation (OLR) are analyzed to discuss the differences in TC activity between the two kinds of periods. Both oceanic and atmospheric conditions differ between HFP and LFP with statistical significance. In HFP, the SST in the eastern WNP (east of 150°E) is higher, and the convective activity at the area from 10°N to 20°N is stronger, than in LFP. These results show that the atmospheric and oceanic circumstances for HFPs in the tropical and subtropical WNP enhance the genesis of TCs in comparison with those for LFPs. It is suggested that the variability of SST in the Pacific, and that of large-scale atmospheric circulation are the main causes of the interdecadal variability of TC activity.
An observational study of snow-cover/atmosphere interactions that occur over Eurasia during the spring snow-disappearance season was conducted. Snow-disappearance timing was determined using five-day averaged snow depth data for 1966 to 1990 in the former Soviet Union (FSU). The climatological snow disappeared in the FSU area during early March in the southwest, and later during early June in the northeast. An empirical orthogonal function (EOF) analysis was applied to interannual anomalies of the snow-disappearance timings. The first EOF is the continental mode with widespread loadings of the same sign, the second mode (EOF2) is characterized by a loading pattern with an east-west dipole structure. The regional focus is placed on central Eurasia in the western pole of the EOF2 loadings, since the spring snow-cover for this region is found to exhibit a significant correlation with the Indian summer monsoon. A composite analysis for the opposite phases of EOF2 shows that, regardless of the leading snow depth during the late winter, the enhanced (reduced) large-scale advection of southerly warm upper-air towards central Eurasia during March and April was most likely to cause the earlier (later) snow disappearance. During mid-April, surface air temperature anomalies are largest over a wide area of central Eurasia, which has high variability in snow-cover extent, and south of the area. These temperature anomalies are larger than those at the 500 hPa level, likely due to the effect of the snow-cover, whereas the temperature anomalies disappeared by early May. This suggests that the snow-cover anomalies over central Eurasia may not have a direct effect on the Indian summer monsoon through the land-surface/atmosphere interaction.
During the winter seasons from 1993-94 to 1998-99 at Eureka (80.0°N, 86.0°W) in the Canadian high arctic, we have observed tropospheric aerosol layers by using a Mie-scattering-polarization lidar system and we have investigated lidar parameters (Mie backscattering coefficients, aerosol depolarization ratios, and Ångström exponents) of the layers. We calculated isentropic back trajectories in order to investigate the source regions of these layers. We also estimated the size of the particles in the layers by calculating modeled lidar parameters and comparing them with the observed lidar parameters. Daily sampling of aerosol particles was carried out in the late winter of 1998-99. Isentropic back trajectories of the aerosol layer observed over Eureka at different altitudes suggested that the source regions of the aerosol were Eurasia, the northern Pacific and Atlantic Oceans, the Arctic Ocean, and regions near the coast of the Arctic Ocean. The lidar parameters calculated by using the Mie theory and the modeled particle size distribution suggested that, although the Mie backscattering coefficient was determined largely by the number concentration of the accumulation mode particles of diameter 0.2 μm to 0.3 μm, the aerosol depolarization ratio and Ångström exponent of the aerosol layers were influenced strongly by the number concentration and the geometric mean diameter of the coarse particles. The trajectories, characteristics, and chemical compositions of aerosols suggest that the arctic layers observed by lidar were composed of sulfate aerosol or sea salt.
The purpose of this study is to further our understanding of LAWS occurring in the lee of mountains through an observational investigation of the formation processes. The location of the study is Hanamaki Airport, situated in a basin in the Tohoku Region, Japan, where there is many reports of LAWS. A statistical analysis, based on LAWS reports submitted by flight pilots, and surface wind data observed at the airport, revealed the importance of changes in wind direction, as well as wind speed when an aircraft encounters LAWS. Thus, the standard deviation of the horizontal wind vector is defined as a WSI (Wind Shear Index) to evaluate the occurrence of severe LAWS. A case study using upper wind data in the boundary layer, recorded by commercial aircraft flying through severe LAWS systems, pointed out the importance of both a sudden change in wind direction and a descending current, especially close to the ground. To clarify the fine structure and formation process of these wind characteristics, a six-month observation of the boundary layer winds, using a Doppler sodar installed at the airport, was conducted from December, 1996. A WSI analysis was applied to Doppler sodar observations and revealed the following three significant factors when a severe LAWS system is foremd: 1) A strong westerly wind in the middle or upper boundary layer, causing a high WSI current, is observed; 2) The process in which a high WSI peak reaches from the boundary layer down to the ground; 3) and, A descending current is present close to the ground after the high WSI peak touches down.
Regional climate in East Asia under 1CO2 and 2CO2 conditions, was simulated for continuous 10-year periods by the RegCM2.5 developed by NCAR, using the output of a CO2 transient run from NCAR-CSM as lateral and surface boundary conditions in order to evaluate the performance of the nested system for the use of climate change simulation caused by global warming for that region. In this study, January and June climates were analyzed. Through the validation of the simulated present climate, it was clarified that the typical precipitation phenomenon which occurs on the northwestern side of Japan during the winter monsoon is relatively well reproduced in the RegCM, but weakly in the CSM. It indicates that the RegCM is essential for the prediction of regional climate change for the East Asia region. Although the present climate reproduced by the RegCM has some marked biases, e.g. the large cold bias in the higher latitude in winter and the missing of the Bai-u front in mainland China, they are mainly due to the overstimation of sea ice area, and the northward shift of the NPH (North Pacific High) in the CSM, respectively. The SST bias in the CSM significantly contributes to the surface air temperature bias on the coast. In the climate change simulations, the large-scale distributions of SLP and temperature in the RegCM bear a resemblance to those of the CSM in both months. On the other hand, the regional scale precipitation change patterns are different between the RegCM and the CSM in June, because the precipitation band near Japan is well reproduced in the RegCM both in the 1CO2 and the 2CO2 climate. In this simulation, some notable climate change features are found, such as the temperature increase at higher latitudes in January, or intensification of the NPH extending to the southwest in June. Although these changes are statistically significant, they are mainly influenced by the bias in the CSM because the changes occur over the bias region, and their magnitudes do not necessarily exceed the bias of the simulated present climate. From these results, it should be stressed that it is of utmost importance that the AOGCM information is of good quality in the prediction of regional climate change.
The impacts of ground hydrology on the high-latitude oceans, such as melting glaciers and discharge from rivers into the oceans, can affect global climate by mediating the flow of low-density, fresh water inflow that strengthens the ocean’s stratification. This suppresses the thermohaline circulation, and also promotes sea-ice formation. Our time series analysis based on sea-ice and river-discharge data indicates that the effect of this fresh water on the sea-ice in the Okhotsk Sea, into which the second largest Siberian River, the Amur, discharges, is relatively unimportant. Interannual variations in the ice extent are negatively correlated with the amount of discharge. We find circumstantial evidence that the inflow of warmer river water tends to raise the sea surface temperature, and that it suppresses ice formation in the following winter. This potential explanation for the negative correlation implies that sensible heat transported by large rivers in high latitudes should be reconsidered in studying global climate change.
In this study, we analyzed the vertical precipitation profiles in the tropics derived from radar onboard the Tropical Rainfall Measuring Mission satellite. For a given rain type, principal component analysis was done to all the profiles that have similar rainfall rate at 2 km altitude. It is found that the first principal component can explain more than 80% of the variations in the profiles, and the reconstructed first principal components closely resemble the ensemble mean. It is therefore concluded that the mean profiles are representative of the typical pattern of the vertical precipitation structure. The mean profiles for deep convections and stratiform rains are then derived and analyzed for the entire year of 1998 in the tropical belt of 15°S to 15°N. The results show that on a logarithm scale the slopes of the profiles are approximately constants if the profiles are divided vertically into 3 or 4 layers. The difference in slopes among layers may reflect the difference of dominant microphysical processes by which the precipitating hydrometeors either grow or evaporate. The difference between rain profiles between over ocean and over land, and the variation of rain profiles with latitudes are also investigated in this study. As one of the applications of the typical profiles to satellite remote sensing, we calculated the brightness temperature — rainfall rate relations using profiles derived in this study and those provided by previous investigators. It is demonstrated that the brightness temperatures resulted from vertical profiles of previous investigators generally agree, to some extent, with those resulting from the profiles of this study. However, some significant discrepancies exist over some rainfall rate ranges and for high microwave frequencies. It is expected that using the profiles of this study will lead to the improvement of satellite rain retrievals because they are derived from a much larger set of observational data, and are more representative of precipitations in the tropics.
The lagged correlations between monthly Asian summer monsoon indices and El Niño-Southern Oscillation (ENSO) index change prominently in the middle of the summer season, based upon data from the late 1970’s to late 1990’s. Following this change in correlations, the traditional summer monsoon season (June-July-August-September) could be divided into two sub-periods in terms of the interannual variability. One is early summer (June), in which the variability of the Asian monsoon is strongly influenced by the anomalous state of ENSO in the previous winter. Another is mid-late summer (July-August-September), in which the Asian monsoon is related to the anomalous state of ENSO in the following winter rather than the previous winter. The precursory signals of the anomalous Asian summer monsoon which are associated with the anomalous state of ENSO in previous seasons, are valid only for the variability of monsoon in the early summer, but not for that in the whole summer season. Therefore, the drastic change of persistence of the ENSO/monsoon system occurs after the early summer, and a new anomalous state tends to start from the middle or late summer. In this coupled system of ENSO and monsoon, the role of the western Pacific seems to be much different from the eastern Pacific. The anomalous state of sea surface temperature (SST) over the western Pacific warm pool area tends to persist from winter until the following late spring (May), and the related abnormal convective activity over that region can be maintained until the following early summer (June). Such kind of characteristics of persistence over the western Pacific is likely to have a memory effect of the anomalous ENSO state, and plays an active role in influencing the variability of the Asian monsoon in the following early summer.
In this study, predictability for the barotropic component of the atmosphere is examined based on analog weather maps in the historical data. The limit of predictability P is defined as the time taken for the initial error to reach the climate noise level which is defined by one standard deviation from the long term mean of the fluctuation in the observed atmosphere. According to the quadratic error growth model by Lorenz (1982), the predictability P is expected to obey a logarithmic function rather than a linear function of the initial error. Although we searched 15, 667, 760 combinations of weather maps, there are no good analog pairs to investigate the error growth for a sufficiently small initial error. For this reason, model experiments were conducted to demonstrate that the quadratic error growth model is applicable to infer the behavior of a small error from the distribution of a large error. From the results of the model experiments, and the best analog pairs in the historical data, we estimated that the predictability for the real atmosphere increases about 6.3 days when the initial error energy is reduced to 1/10. Hence, we may extend the predictability for the barotropic component of the atmosphere if we can reduce the initial error in the vertical mean of the atmosphere.
Tropical cyclone formation was markedly suppressed in the western North Pacific during 1998. To explore this, we contrast monthly mean fields for January, May, and July 1998 with climatologies, and with what is known about the preferred regions and mechanisms for tropical cyclogenesis by season. The atmospheric variables of interest include, surface wind speed, rainfall, outgoing longwave radiation (OLR), total precipitable water (TPW), sea surface temperature (SST), and upper level velocity potential. Prior to June 1998, the El Niño/Southern Oscillation (ENSO) warm event of 1997-98 dominated the atmospheric circulation in the tropical Indian and Pacific Ocean basins. A rising motion anomaly was evident just south of the equator in the eastern Pacific. Consistent with this anomaly were increased convective activity, wetter than normal air, and higher than normal SST. A center of anomalous subsidence was located over the western Pacific for the same time period. Consistent with that anomaly were suppressed convective activity, and drier than normal air. The rising and subsiding anomaly pair defines a shift in the Walker circulation that is a signature of an ENSO warm event. This paper argues that suppressed tropical cyclone formation over the western North Pacific is another effect of this signature. By July 1998 the ENSO warm event had terminated and the center of rising motion had shifted westward to the Bay of Bengal in contrast to the climatological position in the western Pacific, east of the Philippines. This resulted in a region in the southeastern Indian Ocean of anomalous rising motion, increased convective activity, wetter than normal air, and higher SST. Compensating anomalous subsidence was again located over the western Pacific in a region that is normally favorable for tropical cyclogenesis. Thus, even though the effects of the ENSO warm event had abated, and even though the SST in the western Pacific were above the climatological SST, the anomalies in the vertical motion of the large-scale circulation were sufficient to suppress tropical cyclone formation.
The Radio Science Center for Space and Atmosphere (RASC) and the Indonesian National Institute of Aeronautics and Space (LAPAN) conducted daily radiosonde soundings continuously for about 30 months during October 19, 1993 and March 11, 1996, at Bandung, Indonesia (6.9°S, 107.6°E). We obtained a total of 785 profiles of winds and temperatures at 0-38 km, and the same number of humidity profiles below 10 km. In this paper, we first discuss the climatological characteristics of winds, precipitable water vapor content (PWC) and Brunt-Väisälä frequency squared, N2, in the troposphere and lower stratosphere. An annual PWC cycle was clearly observed with sharp contrasts between dry and rainy seasons, centered in July-October and December-March, respectively. The N2 variations were particularly enhanced in the upper troposphere (11-14 km), and the seasonal cycle of N2 is anti-phase to that of PWC. We have further analyzed the seasonal variations in wind velocity and temperature perturbations with a vertical scale of 2 km, which are used as indices of gravity wave activity. The annual cycle of the gravity wave activity is clearly recognized in the entire troposphere, which is anti-phase to the seasonal variations of PWC, that is, in-phase with the N2 variation in the upper troposphere. The gravity wave energy seems to be proportional to the background N2 value, as suggested by a saturated gravity wave model. In the lower stratosphere the annual cycle of the gravity wave activity disappears, but the long-term variations are related to the quasi-biennial oscillation (QBO) structure.
To further our understanding of the large-scale convection organization in the tropics, including super cloud clusters and the Madden-Julian oscillation (MJO), a 120-day long integration is performed using a 2D (zonal-height) cumulus-scale-resolving model. The periodic horizontal domain spans 40, 000 km and mimics the equatorial circumference. A zonal inhomogeneity of the sea surface temperature (SST) is taken into account by allowing the SST to vary between 302 K over the central area of 5, 000 km width and lower value outside with its minimum at 299 K. Such a computational setup provides an analogue of the tropical western Pacific warm pool. Due to computational limitations, we employ a cloud-resolving horizontal grid of 1 km only in the warm-pool area of 10, 000 km scale, and a stretched coarser grid across the rest of the domain. The initial surface flow is taken to be an easterly at 5 m s-1 which makes zonal asymmetry allowing the Wind-Induced Surface Heat Exchange (WISHE) set in. The numerical simulation features a large-scale gravity wave of zonal wavenumber 1, referred to as the MJO-like wave, that propagates eastward with the phase velocity in the range of 8 to 15 m s-1 and travels across the computational domain in 40 to 60 days. The overall evolution of convection exhibits two organization patterns: i) an eastward-propagating convection (EPC), and ii) a quasi-stationary convection (QSC). EPC is identified as an envelope of westward-moving cloud clusters of O (100 km) scale which develop successively to the east of the existing clusters in accord with the propagation of the upward motion of the MJO-like wave. QSC has a horizontal scale in excess of several thousands of kilometers and it is associated with the existence of the warm pool. Propagation of the MJO-like wave modulates QSC quasi-periodically. The convection in the active regime of QSC has an extent of O (1, 000 km), propagates west-to-east, and involves a hierarchy of organization similar to that in Oouchi (1999). It is identified as a super cloud cluster. It is argued that QSC can provide a basic framework for explaining the slow phase velocity (3 to 10 m s-1) of the super cloud cluster. The governing mechanism in the active phase of the super cloud cluster and MJO-like wave can be viewed as wave-CISK, with WISHE playing an important role in driving the eastward movement of the super cloud cluster. Similarities and differences in the wave-CISK mechanism between this study and previous investigations using parameterized convection are discussed.
The response of the Asian continent to transient increases in future anthropogenic radiative forcings using the data generated in a set of numerical experiments performed with four coupled atmosphere-ocean global climate models (A-O GCMs) is examined here. These A-O GCMs have demonstrated reasonable skill as regards their ability to simulate the broad features of present-day observed climatological features over the Asian region. A plausible scenario of climate change over Asia and its six sub-regions as inferred from these A-O GCMs due to the future emissions of greenhouse gases and/or sulfate aerosols is presented. In general, the projected warming over Asia is higher during NH winter than during summer for both the decades 2050s (2040-2069) and 2080s (2070-2099). The rise in surface air temperature is likely to be most pronounced over the North Asia region in all the seasons. Each of the four A-O GCMs considered here simulates an enhanced hydrological cycle and an increase in annual precipitation over most of Asia. The inter-model differences in projections of precipitation are quite large even when averaged for the entire Asian continent suggesting rather limited confidence in the future projections of regional scale precipitation in currently available A-O GCM simulations.