The GPS/MET (Global Positioning System/Meteorology) experiment was carried out between April 1995 and February 1997. Height profiles of the atmospheric refractive index were obatined at 1-60km from the active limb sounding of occulted radio signals from the GPS satellites. By assuming the hydrostatic relation for a dry atmosphere, a new global data set of high resolution temperature profiles has become available. This paper deals with the determination of the thermal structure near the tropopause in the equatorial region using the GPS/MET data. First, we have compared the GPS/MET profiles with nearby radiosonde results at two balloon launch sites in Indonesia; Bandung (6.9°S, 107.6°E)and Pontianak (0.03°N, 109.3°E), and determined that the rms deviation was approximately 1K in the upper troposphere when effects of humidity are small, and about 2K in the lower stratosphere. The GPS/MET profiles are found to be very unique in revealing detailed temperature structure, including sharp inversions and step-wise increase of temperature gradient near the tropical tropopause, which has not been achieved by a conventional satellite measurement. The monthly mean value of the minimum temperature (Tmin) near the tropopause agreed well between radiosonde profiles at Bandung and the GPS/MET results, showing an annual variation with warm enhancements in August-September, and broader cold periods from January to April. The altitude corresponding to Tmin became lower/higher for warmer/colder Tmin. Taking advantage of the global coverage of the GPS/MET data, we have investigated the longitude distribution of Tmin and Hmin, which generally agreed well with earlier studies.
The sensitivity of the Indian summer monsoon to cumulus convection parameterization is investigated using the Center for Ocean-Land-Atmosphere studies (COLA) General Circulation Model (GCM). Two simulations for a period of nine years are performed with observed sea surface temperature, one parameterizing convection using the Kuo scheme and the other using a type of Arakawa Schubert scheme (the Relaxed Arakawa-Schubert [RAS] scheme). Dynamic fields such as winds, velocity potential, etc., are compared with the corresponding fields of National Center for Environmental Prediction (NCEP) reanalysis data sets. The model simulated precipitation is compared with the corresponding observed precipitation data. The seasonal climatology of nine years of model integration shows that the differences between the two cumulus convection schemes have significant impact on the Indian summer monsoon circulation and its associated precipitation. The large-scale features are better simulated with the RAS scheme than with the Kuo scheme. The RAS scheme simulates stronger low level westerly over the Arabian sea, which leads to more moisture transport from the Arabian sea to the central India, than the Kuo scheme. The convective heating over the Bay of Bengal and the Arabian sea regions is larger with the RAS scheme than with the Kuo scheme, except near the surface and in the upper troposphere. With the RAS scheme, the convective heating maximum in the upper troposphere is more intense in the Bay of Bengal region than in the Arabian sea. Compared to the RAS scheme, the Tibetan anticyclone becomes weaker and shifts eastward, accompanying weaker tropical easterly jet, with the Kuo scheme. Similarly the divergent outflow in the upper troposphere associated with Asian monsoon is reduced significantly with the Kuo scheme than with the RAS scheme. Although both schemes are unable to simulate the Indian monsoon rainfall close to observation, the RAS scheme simulates and resolves better the positions of rainfall maxima over the Bay of Bengal and west coast regions. These differences are attributed to the closures and feedbacks of the two cumulus parameterization schemes. The GCM simulated time series of rainfall with both schemes shows that the intra-seasonal variation (monsoon cycle) is phase-locked to the seasonal migration.
Analyses were made of wind speed and specific humidity profiles under neutral conditions obtained by radiosoundings over the gently rolling terrain of the Little Washita Basin, Oklahoma. Inspection of the wind speed profiles showed that the regional (scale of 1 to 10km) roughness of this basin was z0=0.45±0.21m, while the displacement height, d0, was found to be 8.9m. The logarithmic layer of wind speed was observed to occupy the range, (38±43)z0≤(z-d0)≤(297±147)z0 or (0.15±0.10)hi where hi is the height of the bottom of the capping inversion. This result is comparable with those of previous investigations for the lower limit of the range; however, this result also suggests that over this terrain hi is the relevant length scale for the upper limit. Evaporation values at the regional scale were computed by combining the wind speed profiles with the specific humidity profiles. Comparison with surface flux measurements at Bowen ratio ground stations supports the idea that profiles in the neutral atmospheric surface layer aloft at around 100m above the ground are conditioned by upwind surface characteristics at the meso-γ scale.
Sea surface wind (SSW) fields north of 20°N over the North Pacific are reconstructed using sea level pressure (SLP) fields (Trenberth's Corrected Yearly-Monthly Northern Hemisphere SLP, version 010.1) for the period of 97 years from 1899 to 1995. The reconstruction method is purely empirical as follows. First, since the mean height of anemometers installed on ships in the 1980s is approximately 35m, values of monthly mean SSW in the 1980s computed using winds of the Comprehensive Ocean-Atmosphere Data Set (COADS) are adjusted to those at the standard height of 10m. Then, monthly climatologies of SSW and of SLP fields averaged for the ten years from 1981 through 1990 are computed. Next, monthly anomalies of both SSW and SLP fields during the ten years are calculated. Under the assumption of a geostrophic wind balance, geostrophic wind (GW) anomalies are computed using monthly SLP anomalies. Based on comparison between GW and COADS SSW anomalies, two adjustment factors of the reduction coefficient and correction angle are simultaneously estimated for each of four seasons using a neutral regression analysis. Adopting these adjustment factors for GW anomaly fields, corrected wind anomalies are obtained. Finally, the reconstructed SSW fields are obtained as the sum of corrected wind anomalies and ten-year monthly climatologies. When we regard height-corrected COADS SSW fields as the reference winds, root-mean-square differences of the reconstructed SSW fields can be estimated to be less than 1m s-1 in magnitude, and less than 15 degrees in wind direction for the winds greater than 2m s-1. Comparison of the reconstructed SSWs with NCEP reanalysis SSWs is also made and shows that NCEP reanalysis SSWs are systematically 8% greater in magnitude than the reconstructed SSWs throughout the comparison period from 1950 to 1995. Time series of magnitudes of the reconstructed SSWs averaged in the wintertime mid-latitude westerlies region shows interannual and decadal/multidecadal time scale variations: from mid-1930s to mid-1940s, westerlies was remarkably strengthened compared with the rest of the period.
The spatial distribution of sensible heat flux over rice paddies is estimated on a several tens kilometer scale, together with possible error estimation using satellite infrared data. It is found that three types of correction of satellite infrared temperature (correction for atmospheric absorption, satellite viewing angle, and the surface emissivity) over rice paddy yield estimation errors of 25Wm-2 for sensible heat flux on a half-hourly basis as root mean squares errors (RMSEs). These errors are sufficiently small compared with general observational errors. In particular, the viewing angle correction is so significant that the RMSE is 46Wm-2 without the correction. Atmospheric and emissivity corrections using the LOWTRAN7 with radio sounding data yield a good estimation of surface infrared temperature, with RMSE being 1.0°C. An empirical and linear parameterization proposed by Troufleau et al. (1997) is applied for the viewing angle correction. In the approach, parameters are refitted based on the surface infrared temperature data. The Troufleau et al. (1997) parameterization is applied over a wide range of leaf area index (LAI) from 0.01 through 5. This parameterization efficiently corrects the temperature deficit by the viewing angle effect caused by canopy geometry. However, this parameterization is insufficient when the sensible heat flux is large because the temperature deficit is essentially nonlinear with regards to the difference between surface infrared temperature and air temperature.
Homogeneous upper air data for 50 years (1949-1998) from the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalyses project, sea surface temperatures and sea level pressure are used to bring out the three dimensional structure of two dominant decadal/multi-decadal variations in the tropics. The global three dimensional modes represent generalized forms of inter-decadal modes studied earlier only with surface data. In the vertical, both modes show approximate first baroclinic structures over the tropics. The Walker circulation associated with the multi-decadal mode has a wavenumber two structure in the zonal direction. It is shown that the magnitude of major ascending and descending motions associated with the multi-decadal Hadley and Walker circulations, are comparable to those associated with the dominant inter-annual mode. Implications of these large scale global circulations associated with the low frequency oscillations in modulating regional climate on a inter-annual time scale are discussed.
A hurricane model developed at GFDL, NOAA, was combined with each of AVN and NOGAPS global analyses to construct typhoon prediction systems GFDS and GFDN, respectively. The GFDS system performed 125 (178) forecast experiments for 16 (24) storms in the western North Pacific basin during 1995 (1996). It exhibited considerable skill in the forecast of tropical cyclone tracks. The average forecast position errors at 12, 24, 36, 48 and 72h in 1995 (1996) were 95 (108), 146 (178), 193 (227), 249 (280), and 465 (480) km. The improvement with GFDS in the typhoon position forecast over CLIPER was roughly 30%. The reduction of position errors in both average and standard deviations indicates superior forecast accuracy and consistency of GFDS, although there existed systematic northward bias in the forecast motion at low latitudes. On the other hand, intensity forecast was not satisfactory, showing a tendency to overpredict weak storms and underpredict strong storms, similar to the tendency in the Atlantic. Two sets of forecasts performed in the 1996 season, the one by GFDS and the other by GFDN, were compared with each other. Forecast skills of the storm position with the two systems were comparable. However, the two forecast positions tended to be systematically biased toward different directions. As a result, when the two forecasts were averaged, the mean error was 10% smaller than that of each forecast. Also, overall improvement in track forecast was obtained in supplemental experiments in which individual forecasts were corrected for systematic biases. Though systematic bias is not steady, there may be ways to utilize it for improvement of tropical cyclone forecasts.
The gravity wave response of a quiescent atmosphere to a growing convective cell is investigated, using a non-hydrostatic compressible model. A series of experiments are carried out, categorized as: (i) “MOIST” experiments that employ a parameterization of Kessler warm cloud microphysics, and (ii) “DRY” experiments performed with a prescribed heating function. The heating function used in the DRY experiments is based on the latent heat released and absorbed in the convective cell of the MOIST experiment. A shallow mode disturbance having a strong updraft at low levels develops near the convective cell when the cell has reached the later growing stage. This disturbance is interpreted as the gravity wave response to higher vertical-mode forcings that increase during the earlier growing stage of the convective cell, and decrease in the later growing stage. For this shallow mode disturbance, not only the increase in the height of heating, but also the narrowness of the width of heating are important. It is found that this shallow mode disturbance is a particular response to a growing convective cell. The updraft in the low levels of the shallow mode disturbance provides a net vertical displacement. This results from a top-heavy heating profile which forms at the maximum stage of the convective cell. The newly developed cell near the original cell is probably triggered by this net vertical displacement.
Anomalous east-west asymmetric anomalies were seen in sea surface temperature (SST), and convective activity over the equatorial Indian Ocean during October-December 1997. Using NCEP/NCAR reanalysis and sea surface height data obtained from TOPEX/POSEIDON satellite altimeter, its triggering process and strengthening mechanism are identified. The climatological wind over the equatorial Indian Ocean exhibits different seasonal cycle between its western and eastern region. During the summer monsoon season, westerly wind prevails over the western Indian Ocean, on the other hand, easterly wind is dominant over the eastern Indian Ocean at almost the same time. During the 1997 summer, divergent easterly wind anomalies were obvious over the equatorial Indian Ocean due to a warm episode of the El Niño, which weakened (accelerated) the climatological westerly (easterly) wind over the western (eastern) Indian Ocean. As a result, the east-west SST contrast was produced in the succeeding autumn through changing evaporative cooling and upwelling. Corresponding to these SST changes, the convective activities were enhanced (suppressed) over the western (eastern) Indian Ocean and actual wind became easterly in place of climatological westerly wind during October-December 1997. The above easterly anomalies induced westward-moving downwelling Rossby waves, and led to the maximum SST in January 1998 in the western Indian Ocean. On the other hand, eastward-moving downwelling Kelvin waves were generated after the termination of easterly wind anomalies, which were consistent with the SST warming in the eastern Indian Ocean for the period February-June 1998. In this manner, a coupling process between the modulated Walker Circulation associated with the El Niño event and the monsoon circulation from summer to autumn is a crucial factor for inducing the above asymmetric anomalies. Moreover, the oceanic waves are found to be closely related with enhancement of these asymmetric structures.
The impact of two cumulus parameterizations on intensity of a simulated tropical cyclone was examined using a limited area numerical model. A moist convective adjustment scheme and Arakawa-Schubert scheme were used as a cumulus parameterization for each numerical experiment. Typhoon Flo (1990) was selected as a case study. Difference in the cumulus parameterizations had drastic influence on a difference in sea-level pressure averaged over the core region (within the 250km radius from the storm center) of the simulated storm from an early stage of the integration. That is, compared with the experiment with moist convective adjustment, the averaged sea-level pressure with Arakawa-Schubert scheme was relatively low. Central pressure in the latter was also lower than the former, which showed that the latter produced a more intense tropical cyclone consistent with the observation. The difference in central pressure, however, appeared at a later stage, compared with the difference in the averaged sea-level pressure. Temperature tendency due to a moist convective adjustment was considerably negative in the lower troposphere and so was actual temperature tendency, which resulted in a rapid drop of temperature. Consequently, the total air mass within the inner-core region became greater in the experiment with moist convective adjustment than with the Arakawa-Schubert scheme. On the other hand, the difference in the low-level temperature field between the two experiments results in the difference in radial profile of rainfall, and the radius where the eyewall is located. The difference probably brings about the delayed appearance of the central pressure difference. Another numerical experiment was conducted with modified moist convective adjustment not to be cool in the lower troposphere. The result confirmed that cooling in the lower troposphere within the inner core-region greatly inhibits intensification of the simulated storm.
On 26 June 1998 during a field experiment called X-BAIU-98, an orographic rainband, called the Nagasaki line in this study, was observed extending northeastward from the Nagasaki Peninsula in western Kyushu, Japan. The convective cells in this rainband, which were about 5km in horizontal scale and 40min in duration, propagated northeastward at a speed of about 10ms-1. They were deep in height in a northeastward direction. Around the Nagasaki line, a moist convectively unstable atmosphere was observed in the lower layer together with environmental winds that included southerly winds near the surface and southwesterly jet at 3-4km in height. Numerical simulations of the Nagasaki line were conducted using an operational Regional Spectral Model (RSM) of the Japan Meteorological Agency and a Nonhydrostatic Cloud Model (NHM) of the Meteorological Research Institute. While the RSM simulated only a weak precipitation area in western Kyushu, the NHM reproduced many characteristics of the observed Nagasaki line. Sensitivity experiments for topography, humidity and wind profiles showed that a moist convectively unstable atmosphere, mesoscale convergence, and winds having both a strong southwesterly jet at 3-4km and a strong vertical wind shear in the lower troposphere are essential for the formation of the Nagasaki line. Although small and low, mountains on the Nagasaki Peninsula are capable of forming an organized precipitation band under such environmental fields.
Using the back-propagation neural network, a model for predicting tropical cyclone intensity changes in the western North Pacific at 12, 24, 36, 48, 60, and 72h is developed. The data used include the storm positions and intensities, the NCEP/NCAR reanalysis fields, and the sea surface temperature fields for western North Pacific storms occurred during a 14-year period of 1983-1996. The predictors of the neural network model are selected based upon those of the multiple linear regression model. A regression analysis shows that the vertical wind shear predictor is consistently important over the prediction intervals. The average intensity prediction errors from the neural network model with climatology, persistence, and synoptic predictors are 7-16% smaller than those from the multiple linear regression model with the same predictors. Even the performance of the neural network model with only climatology and persistence predictors is slightly superior to that of the multiple regression model that includes synoptic predictors as well. It is revealed that the neural network model does not always improve upon the regression model for every year during the 14 years. However, the number of years that the neural network model is superior to the regression model is (much) larger than the number of years in the reversed situation, and appears to increase with decreasing prediction interval. Sensitivity experiments show that the average intensity prediction errors from the neural network model seem to be insensitive to the number of hidden layers or the number of units in hidden layer. However, there is some room for further improvement of the neural network model upon the regression model with a better hidden-layer structure for tropical cyclone intensity prediction. This study suggests that the neural network model that includes climatology, persistence, and synoptic predictors can be used as an effective tool in tropical cyclone intensity forecasts.
This paper summarizes the findings of an analysis undertaken to evaluate the performance of a range of A-O GCMs in terms of the simulation of broad scale patterns of present-day climatology over Asia, and its six sub-regions. The outputs generated in transient control and anomaly experiments with the seven A-O GCMs compiled at the Data Distribution Centres in Hamburg and Norwich (set up by the Intergovernmental Panel on Climate Change) have been used in this study. The study suggests that the HadCM2, the ECHAM4, the CSIRO and the CCSR models exhibit some skill in reproducing the observed present-day surface climatic elements over the Asian continent. However, most A-O GCMs considered here have limitations in the simulation of present-day regional climatology.