An oscillation similar to the quasi-biennial oscillation (QBO) is obtained in a two-dimensional (the zonal-height domain along the equator) model. The two-dimensional model is taken from a preliminary version of the Center for Climate System Research/National Institute for Environmental Studies (CCSR/NIES) general circulation model (GCM). The maximum zonal wavenumber of the model is 10, and it has a vertical resolution of approximately 1km in the stratosphere, with 45 model layers in σ-coordinates. The Kuo scheme is used for the convective parameterization. The sea surface temperature is set to 300K as the surface condition. The model has no rotation, so the only stratospheric normal mode is a two-dimensional gravity wave. The period of the oscillation found in the model results is too short, being less than 100 days. The amplitude is about twice as large as that of the observed QBO. The amplitude of the gravity wave in the zonal wind is approximately 50ms-1, with a period of about 6 days in the lower stratosphere. The wave disturbance in the zonal wind has a dominant zonal wavenumber 1, resulting in the phase speed being about twice as fast as that of the Wallace-Kousky wave. The vertical wave momentum flux is approximately 2×6.25 times greater than that found in a three-dimensional mechanistic model of the QBO. Considering the amplitude of the oscillation, the 6.25 amplitude squared factor may explain the fast period oscillation relative to that of the QBO.
Three major teleconnection patterns prevailing at 500hPa height, e. g., North Atlantic Oscillation (NAO), Pacific-North American (PNA) and Western Pacific (WP) patterns, are investigated in the northern Winter season associated with cyclone tracks and precipitation patterns. Cyclone position and tracks, objectively analyzed using NCAR Sea Level Pressure Grids (1964-1990) and ECMWF 1000hPa height grids (1980-1990), are composited in reference to the extreme appearance of positive or negative teleconnection patterns around the Atlantic and Pacific. The correlations between the teleconnections and the inter-annual variability of precipitation are also investigated. Significant changes of major cyclone routes are commonly and clearly found in the two data sets as follows. In case of ± NAO patterns, major zonal cyclone tracks fluctuate meridionally, and cyclones tend to move eastward in the middle parts of North Atlantic, or they direct northeastward along the northwestern Atlantic to reach the Norwegian Sea. Significant variations of precipitation associated with these changes of cyclone tracks are found in southwestern and northwestern parts of the west coast of Europe. In the case of dominant ± PNA patterns, cyclones tend to direct eastward within 40°-50°N over the North Pacific and turn north to migrate to the Gulf of Alaska to Bristol Bay, or they move northeastward along the northwestern edge of Pacific to migrate to the Bering Sea or strike on the east coast of Canada. The former pattern corresponds to a strengthening of the eastern part of the Aleutian low, and the latter pattern provides the positive correlation with precipitation on the east coast of Japan and a part of the west coast of Canada. Dominant ± WP patterns are responsible for the meridional fluctuations of apparent zonal distribution of cyclone tracks with its strike of a SSW-NNE direction over the Northern Pacific. The northward shift of this distribution provides the increase of the precipitation amount in the southwestern islands and central Pacific regions of Japan.
A statistical setup based on the mixed distribution of rain rate provides the quadratic relationship between the rain rate variance and the probability that rain rate exceeds a fixed threshold level. The paper proposes single- and double-threshold methods for the quadratic estimation of the area-average rain rate variance. An argument leads to a statistical explanation for choosing optimal thresholds. Empirically determined single optimal thresholds are compared with those for lognormal, gamma and inverse Gaussian distributions. Empirical results for the GATE I data set show an optimal threshold of 31mm/hr for the single-threshold method, while the lognormal, gamma and inverse Gaussian distributions respectively give optimal thresholds of 42.4, 24.8 and 27.4mm/hr, derived from the proposed criterion. Illustration indicates that the single-threshold method which uses a single optimal threshold is as effective as the double-threshold method which uses a pair of optimal thresholds for linearly estimating the area-average rain rate first and second moments.
The latent heat release is a consequence of phase change (vapor, liquid and solid) of water. An algorithm has been developed (Tao et al., 1990) to estimate the latent heating of cloud systems as a function of their vertical hydrometeor profiles (termed a hydrometeor/heating algorithm). The derivation as well as the validation of the algorithm was based on results of a non-hydrostatic cloud model. The improvement of the hydrometeor/heating algorithm and its performance when tested on cloud systems which occurred in various geographic locations are discussed in this paper. Also the application of the hydrometeor/heating algorithm to incorporate surface precipitation as well as vertical hydrometeor profiles derived from multi-channel passive microwave rain retrieval algorithms is presented. The hydrometeor/heating algorithm requires information associated with the vertical profiles of various hydrometeors (ice and water). Although these vertical hydrometeor profiles can be derived from the TRMM radar directly and microwave sensors indirectly, the heating/hydrometeor algorithm needs to differentiate between large and small water/ice particles. Thus, a second retrieval algorithm (termed a convective-stratiform heating algorithm) has been developed. The inputs for this new algorithm are surface rainfall rates and amount of stratiform rain. This convective-stratiform heating algorithm also needs a look-up table which consists of latent heating profiles associated with various types of cloud systems at different geographic locations. The procedure as well as the performace of the convective-stratiform heating algorithm is discussed in this paper.
The feedback effects of cirrus clouds in climate change are investigated, taking account of the temperature dependence of both ice water content and size distribution. The sensitivity parameter which is defined as a sum of partial derivatives is calculated with an accurate radiation code applying the 4-stream discrete-ordinates method and the method of exponential-sum fitting for gaseous transmission (ESFT). Microphysical properties of cirrus clouds are parameterized in terms of the cloud temperature from the data of aircraft observations. The value of ∂H/∂[IWC]×∂[IWC]/∂T, i. e., a product of partial derivatives associated with ice water content feedback, is estimated to be 1.3-2.3Wm-2°C-1, meaning a positive feedback effect, and the value of ∂H/∂[SD]×∂[SD]/∂T which is associated with size distribution feedback is estimated to be -0.1--1.8Wm-2°C-1, meaning a negative feedback effect. The positive ice water content feedback is significantly cancelled by the negative size distribution feedback especially at cloud temperature in the vicinity of -45°C. It is thus shown that the temperature dependence not only of ice water content but also of size distribution is important to estimate the microphysical feedback of cirrus cloud.
To obtain the relationship between the structure and reflectance of winter maritime stratocumulus cloud tops, observations using an aerial stereophotographic method were carried out from January 1989 to January 1991 in the area off Wakasa Bay in the Japan Sea, and around the Amami Islands in the Pacific Ocean. The relationship between cloud top heights and the reflectance of the cloud was first investigated, and a positive correlation was found to exist. An especially high correlation was obtained for the clouds over the Amami Island data. Next, the relationship between the cloud area ratio and the reflectance of the cloud top surface was examined. Results indicated that in areas of low reflectance, the cloud top heights were relatively low and widely distributed. The cloud tops were sharp and jagged in shape. In contrast, in areas of high reflectance, high cloud top heights were found, being uniform on the average. Their shapes were flat and trapezoidal in comparison with the jagged nature of the other clouds. For the clouds north of the Amami Islands, the cloud top heights were lower than those found in other cases. The difference between the maximum and minimum cloud top heights was more than 400m and the values of reflectance were greater than those in the other observations. The difference in the reflectance of the clouds over the Amami Islands may result from the difference between the liquid-water paths.
A case study was made of long-lasting mesoscale cloud clusters observed by the Geostationary Meteorological Satellite (GMS) from 13 to 15 July 1986, which formed on the continent and traveled over the Baiu-frontal zone. The cloud clusters separated into southern- and northern-cloud groups around 130°E. The southern-cloud group decreased its traveling speed. New cumulonimbus-cloud groups formed 40 to 80km to the west of the pre-existing group in the southern-cloud group. Further, new cumulonimbus clouds appeared about 15km to the west of the pre-existing one in each group. Simultaneous formation of two scales of convective clouds occurred in the slow-moving cloud group. The area of traveling clouds, which originated from deep convective clouds in the northern cloud group, lasted for more than 10 hours after the deep convection decayed. Bright bands and streaks were recognized in their radar echo around 135°E. They were located in the layer of convective instability at middle to upper levels, with a large gradient of temperature at 850mb and a synoptic-scale updraft at the middle level around 140°E. It is suggested that the cloud area originating from “convective clouds” was changed to “stratiform clouds.”
Stratospheric aerosols originating from Mt. Pinatubo (15.14°N, 120.35°E), which erupted violently on June 15, 1991, were observed by lidar at Naha (26.2°N, 127.7°E) on Okinawa Island. Aerosols had already reached Okinawa by mid September 1991, the start point of this observation series. The altitude range of the aerosol layer is from just above the tropopause to about 30km. The total amount of aerosols began to increase in mid November corresponding to the appearance of layers at higher altitude, in the 30 to 33km region, and maximum backscattering ratio 23.1 was observed at 25.8km altitude on December 4, 1991. Depolarization ratio observations demonstrated that there were not only volcanic ash particles present but also spherical particles, probably sulphate droplets, from the start of this observation series. The highly depolarized region was observed on lower side of the layer and lower depolarization on the higher side. The depolarization ratio reached a peak value of 0.22 at about 16.4km on November 18, 1991, and has gradually decreased since.