Journal of the Meteorological Society of Japan. Ser. II
Online ISSN : 2186-9057
Print ISSN : 0026-1165
ISSN-L : 0026-1165
Volume 64A
Displaying 1-50 of 67 articles from this issue
  • F. H. Bushby
    1986 Volume 64A Pages 1-10
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     This paper covers the principal milestones in Numerical Weather Prediction from the “beginning” to the present day and is divided into three main sections. First the beginning which, in my view, was L. F. Richardson’s (1922) book on Weather Prediction by Numerical Processes following some earlier work by Bjerknes. The second section is devoted to the theoretical and practical studies which were started by Charney and his colleagues at Princeton University at the end of World War II with the advent of electronic computers and which have continued in many centres up to the present day. The final section deals with the introduction of NWP into operational forecasting and some personal reminiscences of those involved in the early days.
    Download PDF (17492K)
  • A. Hollingsworth
    1986 Volume 64A Pages 11-59
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The field of objective analysis for numerical weather prediction is enjoying rapid development because of theoretical advances in analysis methods coupled with progress in observing and computing technology. This review considers current and projected work on objective analysis under three main headings: theory, verification and practice.
     Recent developments in the theory suggest that there are close links between all analysis methods based on optimality principles. Moreover, simpler analysis methods such as successive correction can be modified to converge to an optimal method in simple circumstances. The basic job of any analysis algorithm is to filter random observation error from the data and then interpolate the filtered data to a regular grid. A new viewpoint on Optimum Interpolation shows the algorithm in a simple light with an intimate connection between the filtering and interpolation capabilites.
     Verifications of observations, forecasts, and analyses are reviewed, using statistics on one-point and two-point height and velocity correlations. The two-point correlations are inter-related through the kinematic equations of two-dimensional turbulence. In the limit of vanishing spatial lag they degenrate to the one-point correlations. These diagnostics offer a means to compare the error characteristics of different observing systems. When used to verify short-range forecasts, they provide a complete description of the structure of the forecast errors. Phillips’ simple theory of forecast errors can reproduce many of the features of the empirically determined structures. New results indicate that there may be a need for anisotropic and non-separable correlation functions for analysis.
     The correlation diagnostics offer new and rigorous ways to estimate analysis error at observation points. They provide a quick and easy measure of the efficiency of an analysis system. They should prove useful in identifying the reasons for system dependence in the results of observing system experiments.
     In discussing current and future research we consider the special problems of analysis in data rich areas over land and near mountains; the problems of analysis over the mid-latitude oceans; and then the problems of tropical analysis. Extensions of O/I offer new approaches to these problems which will be explored in the near future. In the slightly longer term, new developments in the theory of 4-dimensional analysis may well supersede current methods. Finally we consider developments in data quality control, an area of great importance for all other developments. New developments here, using Bayesian methods, offer promise.
    Download PDF (85426K)
  • Sigbjørn Gronås, Knut Helge Midtbø
    1986 Volume 64A Pages 61-74
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     Bratseth (1986) has proposed a successive correction scheme for numerical analysis where the solution converges toward the solution obtained by optimal interpolation. This paper describes how this method is used operationally in two limited area prediction systems, with a mesh of 150 km and 50 km respectively. Data quality control is first performed in a seperate run, and the rate of convergence is used to identify errors. In the following analysis the scale of the horizontal correlation function is decreased with the iteration number. The principle of optimal interpolation is then abandoned, however, the procedure might be looked upon as a simple way of adapting the error statistics to the actual situa. tion. The convergence of the method is investigated, and 5-10 iterations are found to be sufficient for the operational analyses. The method is simple and flexible, and is believed to have advantages compared to traditional optimal interpolation.
    Download PDF (20546K)
  • M. J. P. Cullen, J. Norbury
    1986 Volume 64A Pages 75-82
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     An alternative approach to the definition of balance using an energy principle is presented. In middle latitudes it reduces to Lagrangian semi-geostrophic theory. Numerical calculations are shown which illustrate how such a balanced state can be found by iterative methods.
    Download PDF (9786K)
  • Kenji Takano, Wayman E. Baker, Eugenia Kalnay, David J. Lamich, Joan E ...
    1986 Volume 64A Pages 83-89
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     For the first time, a four-dimensional stratospheric/tropospheric data assimilation system with a top analysis level at 0.4 mb has been developed and used to produce physically consistent gridded analyses for the stratosphere as well as the troposphere for a period during the First GARP Global Experiment (FGGE) and Limb Infrared Monitor of the Stratosphere (LIMS) (November 1978-May 1979).
     The system consists of a two-dimensional optimum interpolation analysis with 18 mandatory pressure levels and a 19 level fourth order stratospheric/tropospheric general circulation model with a horizontal resolution of 4° (latitude) by 5°(longitude) and a top at 0.3 mb. The system allows the utilization of stratospheric data including LIMS, Tiros-N retrievals, rocketsondes and vertical temperature profile radiometer (VTPR) soundings in addition to the other FGGE level 2b data. These data are analyzed every six hours.
     In order to examine the quality of the analyzed data forecast experiments starting from different analyses are performed for the period of the stratospheric sudden warming of late February 1979. The results indicate that by employing the present four-dimensional assimilation approach, the medium-range forecast skill for this event is improved.
    Download PDF (12210K)
  • Per Kållberg, Frédéric Delsol
    1986 Volume 64A Pages 91-106
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     Cloud-track-wind data, also often called SATOB data, provide an invaluable source of observational information from otherwise sparsely covered parts of the lower and upper troposphere, particularly in the tropics. SATOB data were, for the first time, collected on a global basis during FGGE. During the analysis of these data for the ECMWF Main level III-b production, some questions arose regarding apparent systematic biases in the data in and near jetstreams. In preparation for a reassimilation of parts of the FGGE at ECMWF these questions were adressed again, benefitting from the accumulated experience of global data assimilation at the Centre. Several types of diagnostics, based both on the data assimilation. scheme and independent, collocated observations were extracted and evaluated, based on the Final FGGE level II-b data from December 1978.
     The biases were found to be negative, underestimating the windspeed, and quite systematic. All datasets from different producers that could be reliably assessed were found to suffer. The problem was most pronounced in the subtropical jetstreams, where individual wind analyses were found to be seriously deformed by too weak SATOB windspeeds, especially in areas void of other sources of data. Impact was also seen in monthly averages. In a simple calibration experiment, an attempt was made to reduce the biases. The result showed a great impact of the calibration, indicating both the importance of SATOB data and the risks associated with their systematic biases.
    Download PDF (27800K)
  • Akio Arakawa, Jeng-Ming Chen
    1986 Volume 64A Pages 107-131
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The problem of cumulus parameterization is a closure problem, in which we seek a limited number of equations that govern the statistics of a system with huge dimensions. A great deal of uncertainty still exists in the choice of appropriate closure assumptions for this problem.
     While concentrating on the thermodynamical aspects of parameterizing deep cumulus clouds, some of the existing schemes are reviewed with an emphasis on their closure assumptions. It is shown that the closure assumptions in these schemes are some combination of four types.
     It is emphasized that observations should be used more extensively than in the past to directly verify and improve closure assumptions and to assess the limit of parameterizability. As an example, results from our analysis of the macroscopic behavior of moist convection are presented. The results suggest that cumulus clouds are basically parameterizable, in spite of the existence of mesoscale organization.
    Download PDF (40634K)
  • Akihide Segami
    1986 Volume 64A Pages 133-140
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The impacts of second-order turbulence closure scheme on the JMA’s operational regional model forecasts are discussed, using the Mellor-Yamada’s Level 2.5 model and Level 2 model. The forecasts are performed under the winter monsoon case over Japan Island. A large difference is appeared in the stable region over the land. Level 2 model predicts considerably thin stable layer compared with the observation. This structure is improved by Level 2.5 model because there remains small downward flux in the stable layer for Level 2.5 model. Over the sea, observed mixed layer is fairly well predicted by both model, although tops of the neutral layer become higher and bases of the stable layer become lower than the observed ones. The difference between two models is not so large. Forecast differences in the free atmosphere are very little. There are some differences between two models in the heat and momentum supply from the surface, but these are not sufficient to change the forecast of the free atmosphere in the short forecast period. This may be due to the large grid distance of the model.
    Download PDF (12113K)
  • JF. Geleyn
    1986 Volume 64A Pages 141-149
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     In the framework of a if-exchange parameterization for PBL turbulent fluxes we propose a simple way to include the effects of shallow convective clouds. The method is dimensionally consistent with cloud buoyancy production, is cheap and avoids the introduction of additional arbitrary parameters by simply extending the dependency of the exchange coefficient on the Richardson number from the dry to the moist case.
     In the case of the french large scale operational model it is very efficient in suppressing the defects associated with the PBL collapse that is usually observed when one uses only “dry” K-coefficients.
    Download PDF (14131K)
  • V. Ramanathan
    1986 Volume 64A Pages 151-175
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     Analyses of satellite radiation budget measurements, atmospheric response times calculations, and general circulation model results are invoked to elucidate the radiatively driven components of the atmospheric general circulation. The clear-sky and cloud radiative processes are shown to exert significant vertical, latitudinal and longitudinal gradients in the diabatic heating within the troposphere and the stratosphere. The meridional heating gradient, which drives the general circulation, is altered significantly by clouds. In the tropics, clouds introduce a dipole heating pattern of convergence of IR radiation in the upper troposphere and a reduction of solar radiation absorbed by the ocean-surface. The magnitudes of the convergence and the reductionn of radiation fluxes can become as large as 100 W·m-2.
     For the column as a whole, more than 75% of the zonal mean absorbed solar energy in low latitudes is balanced by the IR emission. Likewise, the net diabatic heating in the low-latitude troposphere results from a delicate balance between IR cooling and convective heating. These estimates reveal the strong control of radiative-convective processes in governing the observed thermal structure. The time scales for radiative-convective adjustments are in the range of 10-20 days within the troposphere and are in the range of 20 to 100 days in the lower stratosphere. Hence, adjustments to radiative perturbations occur on time scales of importance to the law frequency variance problem. Furthermore, the atmospheric response to zonal mean radiative heating perturbations normally involves a radiative-convective adjustment. Examples are given to illustrate the mechanisms by which meridional gradients in radiative heating due to cirrus and other processes alter the zonal winds and the propagation of planetary waves.
     The major conclusion of this study is that the observed negative anomalies in the outgoing IR radiation following intense warm episodes of tropical sea-surface temperatures, the so-called El Nino, are indeed anomalies in the cloud-radiative forcing. During the 1982/1983 El Nino phenomenon, over the tropical eastern Pacific, the mid- to upper- tropospheric radiative heating must have increased by 50 to 100 W·m-2 or more and the solar energy reaching the ocean surface must have decreased by at least 25 to 50 W·m-2. These changes, which have been ignored in the earlier studies, can account for a substantial fraction of the observed atmospheric changes. The inferred decrease in the solar energy to the ocean surface should play a substantial role in dissipating the warm anomalies in the eastern Pacific ocean.
    Download PDF (42392K)
  • Richard C. J. Somerville, Sam Iacobellis
    1986 Volume 64A Pages 177-185
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     Cloud-radiation interactions are crucial to climate modeling and will become increasingly important in numerical weather prediction as forecast range increases and as models are developed to predict secondary variables, such as surface temperature. This study explores the role of cirrus cloud optical thickness feedbacks in contributing to climate changes due to increasing atmospheric concentrations of carbon dioxide and other trace gases. For low and middle clouds, which are approximately black bodies for infrared radiative transfer, an increase in cloud optical thickness serves primarily to increase the cloud albedo. Thus, if a climate warming is accompanied by an increase in average atmospheric absolute humidity and hence in average cloud liquid water content, low and middle cloud albedo may increase and serve as a negative feedback which tends to reduce the surface temperature increase. Cirrus clouds, however, are not generally black in the infrared, and so an increase in cloud optical thickness may increase the cirrus greenhouse effect as well as the cirrus albedo. We test this hypothesis using a radiative-convective equilibrium model. Our major result is that cirrus cloud optical thickness feedbacks may indeed tend to increase the surface warming due to trace gas increases. However, this positive feedback from cirrus appears to be generally weaker than the negative feedbacks due to lower clouds. Our results thus confirm those of earlier research indicating that the net effect of cloud optical thickness feedbacks may be a negative feedback which might reduce by a factor of about two the surface warming expected from doubling atmospheric concentrations of carbon dioxide. All of these results are obtained under severely idealized assumptions, however, and should be regarded as tentative. Much further research will be required to determine the effect of cloud optical thickness feedbacks definitively.
    Download PDF (16314K)
  • Ulla Hammarstrand
    1986 Volume 64A Pages 187-197
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     Parameterization schemes for stratiform and convective condensation with cloud water content and cloud cover as dependent variables are implemented in a limited area model for weather prediction. The scheme for stratiform clouds is based on the approach according to Sundqvist (1978, 1981). The scheme for the convective clouds is based on Kuo’s (1974) approach, which is modified to include the prediction of cloud water. A description of the latter scheme is given in the present paper. Aspects regarding the necessary combination of the schemes in connection with transition between the two stability regimes are also discussed. A specific 36 hour forecast is chosen for detailed examination of the predicted cloud and precipitation patterns over Scandinavia. Verifications of the forecast fields are performed utilizing mostly synoptic data. The predicted cloud covered area coincides well with the observed and the amounts of accumulated precipitation is also in agreament with observations. There are no observations of cloud water content availible and the realism of this field can therefore only be judged indirectly by comparision to the precipitation rate. In connection with the attempts to verify the predicted fields of cloud water, cloud cover and precipitation, the urgent need for relevant observational data stood out prominantly.
    Download PDF (18464K)
  • Louis Berkofsky
    1986 Volume 64A Pages 199-209
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     In dealing with the prediction of atmospheric dust concentration by means of mathematical models, a knowledge of the boundary condition—the dust concentration in a thin layer near the ground—is needed. This concentration must be related to the erosion rate.
     The erosion process consists of three phases: detachment, transportation, and deposition. In this paper, we formulate a simple erosion equation, on the basis of reasonable assumptions about the three phases of the erosion process. Solutions of simplified versions of this equation yield reasonable results. In practice, the nonlinear from of this equation would have to be solved interactively with the atmospheric model to yield ground concentration as a function of time and space.
    Download PDF (16707K)
  • André Robert
    1986 Volume 64A Pages 211-214
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The semi-implicit formulation of a model is described in the context of the shallow water equations. A formulation that is both semi-Lagrangian and semi-implicit is also described and its potential advantages are discussed briefly. This last formulation is used in a complete model in three dimensions and some results are presented.
    Download PDF (5207K)
  • Fedor Mesinger, Zaviša I. Janjić
    1986 Volume 64A Pages 215-222
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     A review is presented of spatial aspects of the numerical methods used in weather prediction models. Considerations that have and are being made in distinguishing among various finite-difference techniques are summarized. Properties that have been examined in order to arrive at an estimate and/or at an understanding of the skill numerical schemes have in simulating atmospheric dynamical processes are listed and discussed. A number of points are illustrated by brief presentations of some as yet unpublished results of the authors. One of these concerns the difference in spectral properties of the fields simulated by two horizontal advection schemes. Another is addressed at error calculations done for a recent “θ-conserving” sigma system pressure gradient force scheme. Finally, issues presently attracting interest in horizontal as well as in vertical differencing are reviewed, and some recent results are summarized.
    Download PDF (13392K)
  • J. R. Bates, A McDonald
    1986 Volume 64A Pages 223-232
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     A semi-Lagrangian and alternating direction implicit method for integrating a multilevel primitive equation model is presented. The method derives from an earlier scheme developed by Bates for integrating the shallow water equations, though splitting is not used in the present case.
     A linear analysis assuming an isothermal basic state shows that the scheme is unconditionally stable for advection and has the same lenient stability criterion for gravity-inertia waves as in the shallow water case.
     Integrations are carried out using real atmospheric data and the model’s performance is compared with that of an explicit semi-Lagrangian model and of a semi-implicit semi-Lagrangian model.
    Download PDF (12564K)
  • John L. McGregor
    1986 Volume 64A Pages 233-246
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     A two-time-level semi-Lagrangian primitive equations model has been constructed using the efficient alternating direction implicit (ADI) procedure of Bates (1984), together with split horizontal advection and geostrophic adjustment processes. If the simplest splitting method is used, the solution suffers from large neutral-amplitude oscillations. However a more accurate splitting arrangement may be used in which the adjustment processes are staggered in time. In this case a smoothly developing solution is obtained, even without explicit diffusion.
     Further improvements are obtained if the accuracy of the ADI procedure is increased by a repetition ensuring spatial symmetry. Time traces are used to show the similarity of the solutions for various timestep sizes. Only for the largest timesteps are there noticeable differences, mainly in the form of extra gravity wave activity. Comparisons using a variety of horizontal advection schemes indicate that residual truncation errors from the splitting procedure are also made with a limited-area semi-implicit model. The vertical mode initialization scheme of Bourke and McGregor (1983) was successfully applied to the semi-Lagrangian model; convergence was facilitated by the use of centered advection formulae during the initialization iterations.
    Download PDF (21552K)
  • J. Van Isacker, W. Struylaert
    1986 Volume 64A Pages 247-253
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The Laplace transform technique applied as an initialisation and a time-integration scheme of a baroclinic primitive equations model is outlined. The numerical stability of this procedure and the selective damping of the high-frequency waves are improved by the introduction of an elliptic domain in the complex projection field. The time-step is doubled compared to a split-explicit integration scheme.
    Download PDF (10426K)
  • Ji Zhongzhen, Zeng Qingcun
    1986 Volume 64A Pages 255-261
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     In last years, several problems about nonlinear computational stability have been studied. Many nonlinear computational unstable examples of Richtmyer-type and Fornberg-type are obtained. It is shown that the instability may occur for leap-frog Lilly’s scheme and leapfrog Arakawa’s scheme which are used in the meteorology and keep energy-conservation instantaneously. In Galerkin finite element model and spectral model with leap-frog time-difference, similar examples of nonlinear instability also exist. The general mechanisms of nonlinear computational instability are discussed, particularly the effects of initial conditions on the computational stability are analysed. Furthermore, some theorems on the sufficient conditions for computational stability have been proved and some schemes with non-negativity of numerical operator which can guarantee the nonlinear computational stability are given.
    Download PDF (9863K)
  • O. P. Sharma, H. Upadhyaya, Th. Braine-Bonnaire, R. Sadourny
    1986 Volume 64A Pages 263-271
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The feasibility of regional forecasting using a stretched-coordinate global general circulation model is tested on a series of simulations of the 1979 summer monsoon onset, using enhanced resolution (or “zoom”) over the Indian Ocean area via a coordinate transform. It is found that the medium-to-long-range simulations are sensitive to the formulation of lateral diffusion and its dependency on grid size. The present experiments, performed at relatively coarse resolution, show definite improvement of the simulated circulation and precipitation features over the monsoon area in the medium range, du to the use of coordinate stretching.
    Download PDF (16040K)
  • Akio Kitoh, Tatsushi Tokioka
    1986 Volume 64A Pages 273-280
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The impact of vertical resolution on medium-range weather forecasts is studied with the 12- and 24-Iayer versions of the Meteorological Research Institute general circulation model (MRI·GCM). Two cases are selected for the Northern Hemisphere winter.
     Systematic changes are not clear in the forecast skill with the increase of vertical resolution, while several systematic differences are found in the time mean fields.In the model stratosphere of the Northern Hemisphere, more northward heat flux is achieved in the 24-layer version by an ultralong wave than that in the 12-layer version. In one case, wavenumber 2 contributed to the increase of the poleward heat transport in the 24-layer version, while in the other case wavenumber 1 did. Polar temperature is less cold and thus the polar night jet is less exaggerated in the 24-layer version. Origin of the difference seems to come from the vicinity of the tropopause.
    Download PDF (13010K)
  • D. K. Purnell
    1986 Volume 64A Pages 281-290
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The time-tendency operator for a shallow-water model is split into ‘advection’ and ‘adaption’ operators. Rossby eigen-modes and a gravity-wave subspace are defined for perturbations on any state where, over the whole domain, either the divergence is uniform, or else the absolute vorticity has a uniform sign. It is then shown that
     a) the adaption operator yields zero tendency for Rossby modes,
     b) advection does not transform gravity-wave modes into Rossby modes,
     c) that (a) and (b) allow an ‘exact splitting for Rossby modes’
     This splitting allows selective approximation of gravity wave modes without introducing error in the Rossby-modes, so that economical long timesteps can be used without degrading the accuracy of the part of the solution which is significant for weather prediction.
     A particular computational method using this splitting is constructed and tested experimentally. This has less time-differencing error for Rossby-modes than the ‘leapfrog’ scheme, and has no computational mode. This particular scheme strongly attenuates high-frequency gravity wave modes.
    Download PDF (14810K)
  • Yuk-kwan Chan
    1986 Volume 64A Pages 291-296
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     A new finite difference scheme is constructed for the divergence term with a view to decreasing the errors in its computation.
     A divergence scheme is designed such that it will give zero divergence for any non-divergent wind field.In other words, the computational error in applying this scheme to any non-divergent wind field is zero. Since any wind field can be separated into a non-divergent part and a divergent part, the computational error in applying this scheme to any wind field is equal to the error involved with the divergent part only. At the same time, the divergent part wind is an order of magnitude smaller than the non-divergent part. Hence the error is reduced.
     The new scheme is simple, especially in latitude-longitude or Mercator map projections.
      The scheme has been applied to a 3-Ievel limited area primitive equation model for Hong Kong and is found to give much improvement over the usual divergence schemes.
    Download PDF (6966K)
  • Ralph Masenge
    1986 Volume 64A Pages 297-302
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     Mathematical formulation of weather prediction problems often gives rise to sets of coupled nonlinear partial differential equations which describe the interplay between the dynamic processes of advection, udjustment, and diffusion.In general, such systems of nonlinear equations are solved using numerical techniques. In this paper we consider a nonlinear singularly perturbed advection diffusion problem. Employing the usual Newton iteration method we derive an approximate linear advection diffusion problem and discuss some practical numerical methods for solving it.
    Download PDF (8059K)
  • H. M. Helfand, J. Pfaendtner, R. Atlas
    1986 Volume 64A Pages 303-315
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     A benchmark series of ten-day weather forecasts has been run with the GLA Fourth Order GCM with both a 4° latitude by 5°longitude resolution and a 2° latitude by 2.5° longitude resolution. Ensemble statistics of forecast skill and maps of systematic error fields have been generated for both resolutions. The enhanced resolution added 24 hours of useful predictive skill to the sea level pressure forecasts and 6 hours to the 500 mb height forecasts, but 5 to 6 days into the forecasts the advantage of the finer resolution was lost. The systematic error fields showed that by 8 days the "climate drift" of the 2 ° × 2.5° forecasts had become pronounced and had caused the loss of predictive skill relative to the 4° × 5° forecasts. Additional resultes indicate that a gravity wave drag parameterization scheme might alleviate the climate drift problem.
     Two cases have been studied in greater detail. A significantly improved short-range prediction of the Queen Elizabeth II storm by the 2° × 2.5° model is attributed to reduced truncation error and to more accurately simulated small-scale forcing. The useful skill of a ten-day forecast which was dominated by a quasi-stationary blocking-high pattern over Eurasia was extended by 48 h to day 8 by the 2° × 2.5°model with a realistic simulation of the blocking high and a successful prediction of a cyclone that developed rapidly off the east coast of Asia between day 6 and day 8 of the prediction.
    Download PDF (23643K)
  • Yoshinobu Masuda, Haruo Ohnishi
    1986 Volume 64A Pages 317-326
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     A numerical scheme for integrating the primitive equation over the spherical earth with a quasi-uniform grid system, which is an icosahedral-hexagonal grid, is presented. Making use of the stream function and the velocity potential, all the components of the primitive equations are expressed by Jacobian, Laplacian and flux divergence terms. The finite difference forms for these terms are represented by the line-integral which is easy to form mass a conserving scheme. The scheme presented here conserves exactly the totaland almost exactly the total energy and absolute potential enstrophy of a simple divergent flow. 100 days forecast was performed by this grid system with the Rossby-Haurwitz wave as the initial condition. The total mass was conserved satisfactorily, and the conservation of total energy was also achieved within 0.001% error throughout 100 days integration. The error in the conservation of total absolute potential enstrophy is 0.01-0.1%, and the results of 100 days forecast are satisfactory.
    Download PDF (15257K)
  • Gary L. Russell, Kenji Takano, Frank Abramopoulos
    1986 Volume 64A Pages 327-336
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The accuracy of horizontal difference schemes used in the hydrodynamics parts of General Circulation Models are compared by means of numerical experiments for the shallow water equations on a sphere. As expected, the phase lag of moving waves decreases as the order of accuracy of a scheme increases or as the grid resolution increases. Overall, Takano and Wurtele's partial fourth order energy and potential enstrophy conserving sdheme on the C grid is most accurate. It is clearly superior to the other schemes for the Rossby-Huarwitz wave number 6 initial conditions for coarse grid resolution.
    Download PDF (14290K)
  • J. Coiffier, Y. Ernie, JF. Geleyn, J. Clochard, J. Hoffman, F. Dupont
    1986 Volume 64A Pages 337-345
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The general features of the new french operational spectral model are presented with emphasis on the initialization of the surface variables. By using zonal mean diagnostics, several shortcomings of the original formulations were detected. The use of corrected fields near the equator, of the increment method for vertical interpolation as well as improvements in the physics contributed to the production of better forecasts. These progresses are also documented by the informative method of the zonal mean diagnostics. Results of the screen level data assimilation in an operational framework are presented.
    Download PDF (12853K)
  • T. Kitade, K. Yanagino, N. Sato, M. Oozeki
    1986 Volume 64A Pages 347-357
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The operational hemispheric spectral forecast model was changed in several aspects at Japan Meteorological Agency (JMA). The revised forecast model was used to obtain the initial guess for the analysis in the forecast-analysis cycle as well as the new forecast. The performance of the new model was examined by comparing the forecast results with those by the old model. The new analysis obtained through the new forecast-analysis cycle was used as the initial condition for the new forecast model, while the old routine analysis was used for the old forecast model. Improvements on the performance of the new forecast system were clearly found in the objective skill scores. Subjective examinations of the forecast maps also indicated an improvement of the new model forecasts.
     An experiment was carried out in order to identify which component of the model change contributes to the improvement of the forecast skill. All components of the change in the forecast model have small positive impacts on the improvement of the forecast on average. It was found that the improvement of the analysis through the forecast-analysis cycle most contributes to the improvement of the forecast skill. The reason for the improvement of the analysis is also discussed.
    Download PDF (15286K)
  • W. Stern, R. Pierrehumbert, J. Sirutis, J. Ploshay, K. Miyakoda
    1986 Volume 64A Pages 359-363
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     An assessment is made of the areas of focus for improving extended-range forecasting. Two topics currently being researched involve the reduction of systematic error by improving a GCM’s accuracy and the refinement of the transition between the data assimilation phase and the forecasting phase.
     Subgrid-scale orographic parameterizations have been the subject of recent model improvement activities. Results are shown for an envelope orography with an N48L9 gridpoint model and using a mountain gravity wave drag scheme with an R42L18 spectral model. In both cases there is an encouraging reduction in the systematic errors.
     Proper initialization of tropical features, i. e. 40-50 day waves, may be crucial for extended-range predictions in the extra-tropics as well as the tropics. Using a continuous data assimilation scheme the 40-50 day oscillations in the tropics appear to be well maintained from the assimilation to the forecast phase. However, the assimilation system underestimates precipitation and evaporation rates.
    Download PDF (7796K)
  • Tetsu Hiraki
    1986 Volume 64A Pages 365-385
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
      The intercomparisons of global 3-day forecast with the JMA model and the BMRC model are studied. Precise comparisons of verifications, diagnoses in the models and impact studies reveal that the main differences between the models are concerning with the strength of the diabatic heating and the Hadley circulation in the tropics. It is found in this study that the combination of the physical processes is important as well as each physical process to maintain the Hadley circulation and subtropical jets for the models.
    Download PDF (35462K)
  • Carlos R. Mechoso, Akio Arakawa
    1986 Volume 64A Pages 387-395
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The sensitivity of stratospheric forecasts to tropospheric forecasts is investigated using the UCLA GCM. For this purpose, forecasts for the major stratospheric sudden warming of February 1979 are compared to “forecasts” from identical initial conditions but with selected components of the tropospheric fields periodically replaced by the observed. The comparison shows that the accuracy of stratospheric forecasts can be drastically improved when errors in the tropospheric forecast, particularly in the zonal mean flow, are reduced.
    Download PDF (14860K)
  • Akimasa Sumi
    1986 Volume 64A Pages 397-407
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The performance of the enhanced PBL parameterization scheme proposed by Sumi (1985) have been examined by using an ensemble of 31 forecasts for 3 days conducted in January, 1984. The following results are obtained;
     (i) The warming bias in the lower troposphere in the polar region is much reduced. This is due to the suppression of overprediction of the mature cyclone in the polar region.
     (ii) The zonal wind field is slightly improved due to the improvement of the temperature fields in the troposphere.
     (iii) The improvement of the surface pressure field is slight.
     It is shown that the inclusion of the effects due to the sub-grid scale undulations represented in this scheme is effective in the improvement of the predicted fields. This suggests the effect of processes relating to the land (and the sub-grid scale undulations) should be included in our models. However, it has not yet been clarified how to formulate these processes. The details of parameterization of these processes should be further investigated.
    Download PDF (19065K)
  • Isamu Yagai, Tatsushi Tokioka
    1986 Volume 64A Pages 409-419
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     Effects of sub-grid-scale topography are examined with a version of the MRI general circulation model (GCM-I; Tokioka et al., 1984) by increasing the surface drag coefficient over the continents. Three experiments are performed in perpetual January condition : 1) the control experiment (DO) ; 2) the experiment with the drag coefficient increased as a function of the standard deviation of sub-grid-scale topography (Dl), and 3) the experiment with the drag coefficient over land increased as a function of the bulk Richardson of the planetary boundary layer (D2).
     In Dl, the impact of the increased surface drag coefficient on the time mean fields appears mainly in the level below 700 mb. The intensity of the horizontal wind around high mountains is reduced. Thus, the planetary waves forced by the large scale mountains are reduced, especially around the Tibetan Plateau. There are precipitation changes in correspondence with those in low-level divergence which is roughly in balance with the change in the planetary vorticity advection term divided by the Coriolis factor. Short-comings of the present model (Tokioka et al., 1985), are refined in Dl in the following respects: weakening of the too strong anticyclonic flow around the Tibetan Plateau at low levels, reduction of excessive evaporation over the Bay of Bengal, reduction of precipitation over the western part of Asia and over the equatorial Indian Ocean and increase of precipitation over the maritime continent.
     In D2 almost the same results as those in Dl are obtained. However, non-weakened transient disturbances tend to intrude from west into the area from the Mediterranean Sea to the northern periphery of the Tibetan Plateau. This indicates the important role of sub-grid-scale topography around the Alps and the Iranian Peninsula in preventing intrusion of them.
     Finally, the impacts on the medium-range forecasting are assessed. Very small impact on the 500 mb geopotential eddy is seen up to 10 days. However, in the temperature and wind field at 900 mb, large differences are found locally even within the time scale of the short-range forecasting.
    Download PDF (18880K)
  • Zeng Qing-cun, Yuan Chong-guang, Zhang Xue-hong, Liang Xin-zhong, Bao ...
    1986 Volume 64A Pages 421-430
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     A grid-point global general circulation model (IAP GCM) suitable for medium and long-range developed and systematically weather predictions and climate simulations has been tested by a series of numerical simulations. The main characteristics of the model are as following:
     (1) Introduction of departures of temperature T′, geopotential φ′ and surface pressure ps′ from their “standards” to cancel automatically the large trunctional errors in the mountain regions.
     (2) Introduction of new coordinates and new variables. This leads the energy equation to a very compact form and the grid to a more flexible arrangement.
     (3) The computational scheme exactly keeps the “available” energy conservation (if the dissipation is omitted), and is free from computational modes.
     (4) The formulation and calculation of boundary terms or boundary conditions at the interfaces of atmosphere-ocean and atmosphere-land are physically consistent. Hence there is no false source.
     The results of several basic numerical experiments show that the model has good computational stability and accuracy, and is able to simulate the most large-scale circulations and their variabilities. The simulated January and July climatic circulations and other elements are of reasonable accuracy. Furthermore, the model can successfully simulate the abrupt seasonal changes in the atmospheric general circulations over eastern Asia, the southern oscillation, the northern oscillation and the semi-annual oscillation. In addition, the simulated quasi-stationary oceanographic circulation driven by the simulated wind strees is also satisfactary.
    Download PDF (15252K)
  • Chen Jiabin, Ji Liren, Wu Wanli
    1986 Volume 64A Pages 431-440
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     An improved scheme for a spectral model has been investigated in this paper. Through the introduction of a reference atmosphere, prognostic variables become smoother on the tilted sigma-surface over the mountains or a frontal zone, and hence truncation errors are reduced. As a result, the problems of the Gibbs waves and of negative values of model topography over the sea, which are encountered in current spectral models, can be solved by the improved scheme. For comparison we have performed numerical integrations of barotropic and baroclinic Rossby-Haurwitz waves, and the predictions of real cases both with the improved scheme and the current scheme in a spectral model. The results show that the scheme presented in this paper is better than the current scheme.
    Download PDF (14439K)
  • Hajime Nakamura, Naoyuki Hasegawa
    1986 Volume 64A Pages 441-453
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     Using the JMA global spectral model a forecast experiment was performed to investigate the mechanisms to maintain the Baiu (Meiyu) front and the air masses around it. The initial condition was June 15, 1984 and the model was integrated 15 days until June 30.
     Over the China continent the Meiyu front, which is characterized by a stationary band of large humidity contrast, was well predicted by the model. When the Tibetan Plateau was Tibet and propagated removed from the model, a baroclinic wave developed in the eastward. It disturbed the stationary Meiyu front. This suggests one of the roles of the Tibetan Plateau is to prevent from mixing the warm, moist air to the south and the cold, dry air to the north of it and to suppress development of baroclinic disturbances. This situation is a favorable condition for the Meiyu front to be stationary.
     In the model the Baiu front shifted northward around Japan and the north Pacific Ocean and the subtropical high extended westward and northward compared with the observation. The polar air mass north of the front became warm and the Aleutian low weakened substantially.This error was found in whole troposphere below 300 mb. The cooling effect due to the low sea surface temperature over the north Pacific Ocean was investigated by integrating a model with artificial this strong cooling in the lower atmosphere. However,only improved the temperature in the lower troposphere. The temperature field and circulation did not change in the middle and upper troposphere. Heat budget analysis in the north Pacific Ocean using observed data showed that the cold air advection associated with the lows which develop around the Aleutian Islands is equally or more important than the cooling from the sea surface to maintain the cold polar air mass. Especially in the middle and upper troposphere the southward cold air advection is the major contributing factor.
     There was a positive temperature error in the Siberia and the north China in the model, which was due to the excessive sensible heat flux. This was related to a small value of the efficiency of the surface evaporability used in the model. Increase of the parameter can reduce the temperature error over the land. However, the air in the PBL became wetter than the observation. It suggests that an appropriate treatment of the process at the ground surface is important to predict correctly the characteristics of the air mass north of the Meiyu front.
    Download PDF (24917K)
  • M. Imbard, R. Juvanon du Vachat, A. Joly, Y. Durand, A. Craplet J. F. ...
    1986 Volume 64A Pages 455-465
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     In the French Weather Service the “PERIDOT” System produces high resolution short range numerical forecasts over France and its immediate surroundings. The system includes: a fine-mesh analysis making direct use of NOAA satellite radiances, a nonlinear normal mode initialization and a numerical weather prediction model with a complete physical package It has been in operational use since January 1985.
     A statistical comparison of the model diagnosed screen level parameters with the corresponding surface weather observations has been done. It has been used to improve the physical package and can give an estimate of the skill of the operational forecast. Finally, this skill has been compared with the one of the large scale “EMERAUDE” forecasting system.
     The relative humidity and total precipitation forecasts shown below as examples are providing a useful information for the forecaster.
     The fine mesh model is also used for research experiments. It has been tested on an area covering most of the United States for a situation studied in the NCAR project OSCAR. For this purpose it was integrated with different domains and different meshes with particular attention paid to total precipitation forecasts.
    Download PDF (17658K)
  • C.D. Hall
    1986 Volume 64A Pages 467-472
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     Precipitation forecasts from the UK operational limited-area model have been verified and some results are presented here. Total forecast precipitation over 6-month summer and winter seasons is compared with high-resolution data obtained from the UK raingauge network. In general good agreement is found between the model and observations though some systematic errors are evident. In particular precipitation over high ground in the west and north of the UK is underforecast. Forecasts of 6-hour accumulations are assessed by comparing with observations from 369 stations over the European area covered by the model. Threat scores are given for threshold values of 0.05 and 2.0 mm and for various forecast periods. It is found that frontal rain is usually well forecast. A separate assessment was made of the representation of fronts in the model; 72 percent of 24-hour forecasts of frontal position over the UK are found to be within 100 km of the verifying position.
    Download PDF (8574K)
  • Yasuo Tatsumi
    1986 Volume 64A Pages 473-483
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     Tatsumi (1986) proposed a spectral limited area model with time depenbent lateral boundary condition and successfully applied the spectral method to the Japan Meteorological Agency (JMA) operational 12-level fine-mesh limited area Asia model (FLM). We applied the proposed spectral method to the JMA operational 13-level very-fine-mesh Japan model (VFM). The VFM has twice as fine a resolution as the FLM, and it includes more sophisticated physical processes than the FLM.
     The forecast comparison of the grid and the spectral VFMs shows that the spectral-VFM is superior to the grid-VFM for the forecast of a rapidly developing cyclone near the Japan Islands by its forecast position and surface pressure of the cyclone center. We conclude that the developed cyclone characterized by the concentrated strong vortex near the center might be better resolved by the spectral-VFM than by the grid-VFM. Except near the cyclone center the two VFMs produced substantially the same forecast. The advantage of the spectral-VFM is presumably due to the fact which Tatsumi (1986) pointed out that the spectral method:
     a) has no computational dispersion,
     b) is free from systematic retardation of phase speed, and
     c) completely excludes aliasing error.
    Download PDF (17285K)
  • S.-L. Tung, I. Halberstam, C. Johnson
    1986 Volume 64A Pages 485-493
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
      A description of the current status of the Relocatable Limited Area Model (RLAM) is presented. The purpose of the model is to allow flexibility in studying the effects of various parameterizations and numerical techniques on regional forecasts. To date, RLAM has been developed to the point where it can be located anywhere on the globe with appropriate mappings (polar stereographic, Lambert conformal, mercator, or latitude-longitude) and offers a choice of differencing schemes and lateral boundaries. Physical parameterizations of boundary layer fluxes and dry and moist convection have been incorporated from the National Meteorological Center’s Quasi-Lagrangian Nested-Grid Model. Results are presented in the form of forecasts from various combinations of differencing schemes and boundary formulations compared with forecasts from a global, spectral model in use at the Air Force Geophysics Laboratory. These results indicate that besides concerns for stability, selections of different formulations can lead to significantly different forecasts.
    Download PDF (15204K)
  • Hiroshi Yokoyama, Akihide Segami
    1986 Volume 64A Pages 495-502
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The “increment” method in the p-sigma conversion to obtain the initial fields is applied to a regional model in order to get more realistic initial planetary boundary layer (PBL). This method vertically interpolates the increment, i. e , the difference between the analysis and the forecast of previous run, while the operational method interpolates the analysis only. Forecast experiments are carried out for two cases in winter, in order to investigate the impact of the method on the initial and forecast fields. Results are compared with those by the operational method.
     Initial fields by the increment method can maintain fairly well the mixed layer over the warm sea and the stable layer above the cold continent. The patterns are much close to the observations. In another case, the initial moisture and wind fields around a cyclone become more realistic by the increment method. The improvements on the initial fields affect not only the spin-up time of precipitation but also the development of a cyclone.
    Download PDF (12102K)
  • Kazuo Kurihara, Mitsuru Ueno, Nobuo Miura
    1986 Volume 64A Pages 503-516
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     In Baiu season intense rainstorms occur frequently in western Japan along the Baiu front. On 22-23 July 1983, an intense rainfall event continued more than ten hours in the western part of the Chugoku district and the maximum 1-day precipitation amount reached 300 mm.
     We made a forecast experiment using limited area models with grid sizes of 63.5 km and 25.4 km and evaluated the effect of increased grid resolution and representation of orography in the models on this intense rainfallevent.
     The two models predicted less amouts of precipitation than the observation. The locations of maximum precipitation tend to move eastward in this models, while it stayed almost at the same position in the observation.
     The features of sea level pressure, wind and water vapor do not differ significantly between the 63.5 km and 25.4 km models. However, there are large differences in the intensities of upward motion and positive vorticity in the Baiu frontal zone between the two models. They are much stronger due to the more active condensation heating in the 25.4 km model. This fact results in some improvement of the intensity and horizontal distribution of precipitation.
     The inclusion of a steep orography in the 25.4 km model show a tendency that the location of maximum precipitation become stationary and close to that of the observation and its amount increased.
    Download PDF (20345K)
  • K. Ninomiya
    1986 Volume 64A Pages 517-531
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     Some problems about the numerical prediction of mesoscale severe phenomena are discussed. First, the outline and defects of the meso model, which has been operated since 1983 in JMA, are described. Then some results of recent prediction experiments made to improve the mesoscale numerical prediction are presented. The results indicate that the high resolution primitive model with proper dynamical initialization procedure has the capability of predicting local severe phenomena.
    Download PDF (23454K)
  • E. Müller, D. Frühwald, I. Jacobsen, A. Link, D. Majewski, J ...
    1986 Volume 64A Pages 533-546
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     The regional-scale “Europa-Modell“ (EUM) is presented as part of a new NWP system designed at the Deutscher Wetterdienst. The model considers all relevant subgrid-scale processes placing special emphasis on turbulent fluxes, soil processes, and the hydrological cycle. In particular the prediction of cloud water is used to parameterize the formation of rain and snow. Examples of forecast experiments demonstrate the model’s capability to simulate regional-scale meteorological processes and structures (e.g. fronts, smog situation). Using EUM as a research tool the development of mesoscale structures, the long-term behavior, and the contribution by different cloud physical processes to precipitation are studied. Finally, the application of the model output as a data source for follow-up models is discussed and a method to generate quasi-diagnostic data sets covering longer periods presented.
    Download PDF (21303K)
  • G. A. Mills, L. M. Leslie
    1986 Volume 64A Pages 547-561
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     On two successive days in November 1979, lines of convection developed over South Australia, with widespread damage occurring on the second day. In this paper the quality of the numerical weather prediction guidance for these two events is assessed, using a mesoscale (75 km grid spacing) NWP model. It is shown that model forecasts of divergence, low level flux convergence of moist static energy (FCE), stability indices, and vertical motion provide high quality forecast guidance. It is shown that the position of the low level moist tongue, and maxima in the fields of moist convective instability and low level convergence were common to both events. However the greater magnitude of these indicators, a greater 1000-500 mb shear, and a stronger frontogenetic forcing near the convective line the second day do suggest the potential for more intense convection on that afternoon.
    Download PDF (23654K)
  • B. W. Golding
    1986 Volume 64A Pages 563-572
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     A non-hydrostatic numerical weather prediction model has been implemented on a 15 km grid covering the United Kingdom to provide guidance to local weather forecasters in the period up to a day ahead. The model uses a turbulent kinetic energy formulation for vertical diffusion and carries cloud water as an additional variable. Account of the scales represented has been taken in the convection scheme. Radiation budgets are computed for the surface and the highest cloud top. Surface synoptic reports are incorporated to add mesoscale detail to the initial specification of boundary layer and cloud variables. Trials of the system have shown considerable skill in surface temperature and wind forecasts. The precipitation forecasts are superior to previous numerical models and show a realistic orographic enhancement. Some encouraging cloud and fog forecasts have been obtained.
    Download PDF (16383K)
  • Mitsuru Ueno, Ryusuke Taira, Tatsuya Kudo
    1986 Volume 64A Pages 573-584
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     A dynamic assimilation method using observed rainfall rates is proposed, aiming at improving short range forecast of precipitation. In this method, a mesoscale model is integrated for an earlier period of prediction with the inclusion of latent heat estimated from the observed rainfall rates. The result of preliminary tests is encouraging.
    Download PDF (18166K)
  • Huw C. Davies
    1986 Volume 64A Pages 585-594
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     Some issues that arise in relation to the assimilation of cloud condensational heating in mesoscale prediction models are considered from a theoretical standpoint. A study is undertaken of the response of a simple atmospheric flow model to various forms of externally prescribed, mesoscale, steady, zero horizontally or area-averaged diabatic heating distributions. It is shown that the amplitude and/or phase of the response exhibits strong sensitivity to the depth of the “cloud” layer and to the shape of the vertical diabatic profile within the layer. On the premise that sustenance of the cloud system requires a phase match between the low-level convergence and the diabatic heating rate a search is made for forms of vertical diabatic profiles that satisfy this criterion. The results gleaned from this inverse approach are consistent with contentions that the contribution of evaporative cooling in a sub-cloud layer plays an important role in determining the response of “shallow” systems, and that the relative disposition of condensational heating between the upper and lower regions of the cloud layer is an important factor for “deep” systems. The general applicability of these inferences is limited by the idealised nature of the theoretical model employed in the study. At the very least, they serve as tentative, cautionary remarks on the issue of the assimilation of condensational heating effects in NWP models.
    Download PDF (15588K)
  • Chi-Sann Liou, Russell L. Elsberry
    1986 Volume 64A Pages 595-604
    Published: 1986
    Released on J-STAGE: February 18, 2015
    JOURNAL FREE ACCESS
     On 28-30 March 1984, a complicated case of cyclone development occurred in the southeastern region of the United States. A synoptic-scale cyclone moved from east Texas to North Carolina and several meso-alpha scale cyclone centers developed within the larger scale cyclone. One of the mesoscale centers rapidly grew into a 966 mb low along the east coast of the U. S. A. at 18 GMT 29 March 1984.
     An updated version of the Navy Operational Regional Atmospheric Prediction System (NORAPS) is used to study the impact of cumulus convection, parameterized by a Kuotype scheme, on numerical prediction of this meso-alpha scale cyclone development. Two 36 h predictions are made starting from 00 GMT 28 March 1984. The Control prediction is a simulated operational run with full model physics. The Experimental prediction is similar to the Control except the Kuo-type cumulus parameterization is removed, so that the non-convective precipitation is the only process for latent heat release. The impacts of parameterized cumulus convection on meso-alpha scale cyclone prediction are examined by comparing the predicted cyclone development and the heating and moisture distributions around the cyclone center in the Control and Experimental predictions. At 00 GMT 29 March, a newly generated smaller scale (meso-alpha) low appears to be imbedded in the larger scale cyclone in both predictions. The smaller scale cyclone did not grow in the Control run. However, the cyclone deepened 13 mb within 12 hours in the Experimental run. Delayed and more intense latent heat release is clearly observed in the Experimental prediction, which reflects the result of removing the cumulus parametepization and may be responsible for the development of meso-alpha scale cyclone in the later periods.
    Download PDF (17703K)
feedback
Top