Some experiments in dynamic initialization for a simple primitive equation model

Temperton, Clive

doi:10.1002/qj.49709942009

Cited by 24 publications

(4 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ten minutes constituted one initialization time step. This initialization scheme, unlike those of say, Mesinger [21 ] and Temperton [22], is not restorative and hence none of the variables (u, v, z) was restored during the initialization process. The streamline associated with the wind field from the initialization experiment are given in Fig.…”

Section: Initialization Experimentsmentioning

confidence: 96%

Some aspects of dynamic initialization using real tropical wind data

Kiangi

1977

Arch. Met. Geoph. Biokl. A.

View full text Add to dashboard Cite

SummaryIn this study an iterative initialization scheme suitable for tropical regions is designed, tested and used as an initialization vehicle in an experiment involving real tropical data. The model equations for the initialization scheme are the so called shallow water equations bounded by rigid walls at the north and south boundaries. The advective terms of the model equations are evaluated using the quasi-lagrangian scheme. This advective scheme was modified following Kanamitsu [18] so as to prevent the occurrence of the so called two grid distance waves. Time differencing scheme in the time integration (10 time steps forward ans 10 time steps backward) of the model equations is the Euler-backward type. The initialization scheme is found to conserve mean total energy, mean potential vorticity and mean square potential vorticity following an air parcel. The designed initialization scheme was used in an initialization experiment involving real tropical wind data. In this experiment, we started with observed wind field (rotational plus irrotational component) and the geopotential computed from non linear balance equation. The results from this experiment show that the initially imbalanced motion and mass fields tend to settle down to some form of balance during the initialization process. It is further noted that the "initialized" motion fields have more organized streamline pattern than the original wind field. Also, the streamline patterns shown that the difference between the initialized and original motion fields is small. This is desirable for any initialization scheme. Zusammenfassung

show abstract

Section: Initialization Experimentsmentioning

confidence: 96%

Some aspects of dynamic initialization using real tropical wind data

Kiangi

1977

Arch. Met. Geoph. Biokl. A.

View full text Add to dashboard Cite

show abstract

“…Following Temperton [7], the model chosen represents a two dimensional, incompressible flow, in the presence of nonlinear interactions. The partial differential equations, in the momentum form, are as follows:…”

Section: The Modelmentioning

confidence: 99%

“…Repeated application of the above process finally filters out the high frequency waves and brings about a certain degree of balance between the mass and the wind fields. Temperton [7] made a detailed analysis of these schemes but found them wanting in certain respects. He proposed a new scheme, which is known as the "averaging scheme".…”

Section: Introductionmentioning

confidence: 99%

Application of the dynamic initialization technique to a primitive equation model

Sinha

Kulkarni

1982

Arch. Met. Geoph. Biocl. A.

View full text Add to dashboard Cite

SummaryAn attempt is made to study certain characteristic features of the "dynamic initialization" (DI) technique, incorporating the averaging scheme. The limited area barotropic version of the primitive equation model is used. Numerical experiments are performed by varying the number of time-steps during the iterations, N, from 3 to 7. It is found that forN = 5, the least ageostrophic wind component (Area-mean value) is obtained. However, the least "noise" is generated during the time-integration, for N = 6, which also gives the best predicted fields. These features are discussed. Zusammenfassung Anwendung der dynamisehen Initialisationstechnik auf ein primitives Gleichungs-ModellEs werden bestimmte charakteristische Merkmale der dynamischen Initialisationstechnik (DI) mit Einbeziehung des Mittelungsschemas untersucht. Die auf ein begrenztes Gebiet bezogene barotrope Version des primitiven GleichungsmodeUs wird angewendet. Numerische Experimente werden mit Variierung der Zahl der Zeitschritte der Iteration N, yon 3 his 7 durchgefLihrt. Es wurde gefunden, daft fdrN = 5 die ldeinste ageostrophische Windkomponente (Gebietsmittelwert)erhalten wird. Der kleinste ,,L~rm" wird jedoch bei der Zeitintegration fiir N = 6 erzeugt, was demnach die besten vorhergesagten Felder ergibt. Diese Merkmale werden besprochen.

show abstract

“…Sugi (1986) also proved that this method is equivalent to NMI with a frequency dependent underrelaxation factor, and named it "dynamical normal mode initialization" (hereafter, we abbreviate it to I)NI following Sugi). The old version of dynamic initialization was also capable of eliminating high-frequency gravity waves from the initial field of a hemispheric model (Temperton, 1973(Temperton, , 1976. It used, however, much more computational time than DNI because the nonlinear terms were updated at each forward 'backward integration.…”

Section: Introductionmentioning

confidence: 99%

Dynamic Normal Mode Initialization for a Limited-area Model

Satomura

1988

Journal of the Meteorological Society of Japan

View full text Add to dashboard Cite

The dynamic normal mode initialization scheme (DNI) proposed by Sugi is applied to a limited-area model. Performance of two integration schemes for DNI is examined, and it is shown that Okamura's time integration scheme is inefficient and a backward implicit scheme is appropriate for DNI applied to the limited-area model. All physical processes can be included in the integration cycle of DNI.Horizontal wind and precipitation rate are compared using a control forecast, a forecast with the nonlinear normal mode initialization (NMI), and with DNI. The deformation of horizontal wind near the mountains is reconstructed by DNI better than by NMI. The area-averaged precipitation is not largely improved by DNI, but the rms error of the precipitation from the control forecast is smaller in DNI than in NMI in the first three hours.

show abstract

Some experiments in dynamic initialization for a simple primitive equation model

Cited by 24 publications

References 13 publications

Some aspects of dynamic initialization using real tropical wind data

Some aspects of dynamic initialization using real tropical wind data

Application of the dynamic initialization technique to a primitive equation model

Dynamic Normal Mode Initialization for a Limited-area Model

Contact Info

Product

Resources

About