The realizable Markovian closure and realizable test-field model. II. Application to anisotropic drift-wave dynamics

Physics of Plasmas

Volumn 4, Issue 11, 1997, Pages 3895-3909

  Bowman, John C ^a,c   Krommes, John A ^b  

^a MAX PLANCK INSTITUT FÜR PLASMAPHYSIK   (Germany)

^b PRINCETON PLASMA PHYSICS LABORATORY   (United States)

^c UNIVERSITY OF ALBERTA   (Canada)

Author keywords

[No Author keywords available]

Indexed keywords

APPROXIMATION THEORY; COMPUTER SIMULATION; DAMPING; FOURIER TRANSFORMS; MARKOV PROCESSES; MATHEMATICAL MODELS; NAVIER STOKES EQUATIONS; STATISTICAL METHODS; WAVE PLASMA INTERACTIONS;

FREQUENCY SHIFTS; HASEGAWA-MIMA EQUATION; LINEAR WAVES; REALIZABLE TEST FIELD MODEL;

PLASMA TURBULENCE;

EID: 0031268489     PISSN: 1070664X     EISSN: None     Source Type: Journal    
DOI: 10.1063/1.872510     Document Type: Article

References (56)

1. R. H. Kraichnan, Phys. Rev. 109, 1407 (1958).
2. R. H. Kraichnan, J. Fluid Mech. 5, 497 (1959).
3. R. H. Kraichnan, J. Math. Phys. 2, 124 (1961).
4. D. C. Leslie, Developments in the Theory of Turbulence (Clarendon, Oxford, 1973).
5. J. A. Krommes, in Handbook of Plasma Physics, edited by M. N. Rosenbluth and R. Z. Sagdeev (North-Holland, Amsterdam, 1984), Vol. 2;
6. Basic Plasma Physics II, edited by A. A. Galeev and R. N. Sudan, Chap. 5.5, pp. 183-268.
7. J. C. Bowman, J. A. Krommes, and M. Ottaviani, Phys. Fluids B 5, 3558 (1993).
8. A closure is said to be realizable if there exists an underlying probability density function for the statistics it predicts.
9. R. H. Kraichnan, in Theoretical Approaches to Turbulence, Vol. 58 of Applied Mathematical Sciences Series, edited by D. L. Dwoyer, M. Y. Hussaini, and R. G. Voigt (Springer, New York, 1985), Chap. V, p. 91.
10. R. H. Kraichnan, J. Fluid Mech. 47, 513 (1971).
11. R. H. Kraichnan, J. Fluid Mech. 56, 287 (1972).
12. R. H. Kraichnan, Phys. Fluids 8, 575 (1965).
13. Y. Kaneda, J. Fluid Mech. 107, 131 (1981).
14. J. Qian, Phys. Fluids 26, 2098 (1983).
15. V. Belinicher and V. L'vov, Sov. Phys. JETP 66, 303 (1987).
16. R. H. Kraichnan, Phys. Fluids 9, 1728 (1966).
17. J. R. Herring and R. H. Kraichnan, J. Fluid Mech. 91, 581 (1979).
18. S. V. Bazdenkov and N. N. Kukharkin, Phys. Fluids A 5, 2248 (1993).
19. A. Hasegawa and K. Mima, Phys. Rev. Lett. 39, 205 (1977).
20. A. Hasegawa and K. Mima, Phys. Fluids 21, 87 (1978).
21. J. G. Charney, Geof. Publ. XVII, 3 (1948).
22. C. E. Leith and R. H. Kraichnan, J. Atmos. Sci. 29, 1041 (1972).
23. J. C. Bowman, J. Sci. Comput. 11, 343 (1996).
24. A. Pouquet, M. Lesieur, J. C. André, and C. Basdevant, J. Fluid Mech. 72, 305 (1975).
25. J. C. Bowman, J. Fluid Mech. 306, 167 (1996).
26. J. C. Bowman, Ph.D. thesis, Princeton University, Princeton, NJ, 1992.
27. G. Hu, J. A. Krommes, and J. C. Bowman, Phys. Lett. A 202, 117 (1995).
28. G. Hu, J. A. Krommes, and J. C. Bowman, Phys. Plasmas 4, 2116 (1997).
29. J. A. Krommes and J. C. Bowman, Bull. Am. Phys. Soc. 33, 2022 (1988).
30. J. C. Bowman and J. A. Krommes, Bull. Am. PHys. Soc. 33, 2022 (1988).
31. G. Holloway and M. C. Hendershott, J. Fluid Mech. 82, 747 (1977).
32. This is apparent even from a heuristic renormalization (resonance broadening) of the δ function in the wave kinetic equation to a Lorentzian.
33. The quasistationary formulation used in Ref. 30 does not ensure that the transient energies are non-negative since η can achieve negative values during the evolution.
34. R. H. Kraichnan, Phys. Fluids 6, 1603 (1963).
35. There is a typographical error in the triad assignments given in Ref. 33.
36. J. R. Herring, J. Atmos. Sci. 32, 2254 (1975).
37. L. Turner, "EDQNM model of evolution of homogeneous anisotropic turbulence in a helicity representation," submitted to J. Fluid Mech. (1997).
38. G. R. Newman and J. R. Herring, J. Fluid Mech. 94, 163 (1979).
39. M. Larcheveque, J. P. Chollet, J. R. Herring, M. Lesieur, G. R. Newman, and D. Schertzer, in Turbulent Shear Flows, edited by L. J. S. Bradbury, F. Durst, B. E. Launder, F. W. Schmidt, and J. H. Whitelaw (Springer, New York, 1980), Vol 2, pp. 50-65.
40. J. R. Herring and R. M. Kerr, J. Fluid Mech. 118, 205 (1982).
41. J. R. Herring, D. Schertzer, M. Lesieur, G. R. Newman, J. P. Chollet, and M. Larcheveque, J. Fluid Mech. 124, 411 (1982).
42. A. Hasegawa, C. G. Maclennan, and Y. Kodama, Phys. Fluids 22, 2122 (1979).
43. M. Ottaviani, J. C. Bowman, and J. A. Krommes, Phys. Fluids B 3, 2186 (1991).
44. A more sophisticated technique for extending the computational power of the DIA, which partially accounts for the non-Gaussian statistics predicted during a previous evolution, has been developed in H. A. Rose, Physica D 14, 216 (1985).
45. The substantial memeory requirements, roughly 200 Megabytes, make it difficult to continue the evolution without a dramatic decrease in the speed of the calculation because of the need to swap data to disk.
46. W. P. Dannevik (private communication, 1990).
47. R. E. Waltz, Phys. Fluids 26, 169 (1983).
48. Waltz describes the equations he solves as the result of applying a "resonance approximation" to "the DIA weak coupling theory." He also sometimes refers to these equations as the DIA itself; however they are really just a quasistationary version of the EDQNM closure.
49. T. H. Dupree, Phys. Fluids 10, 1049 (1967).
50. 2 appears because the fundamental nonlinear interaction arises from the polarization drift.
51. J. A. Krommes, J. C. Bowman, and M. Ottaviani, Bull. Am. Phys. Soc. 35, 1957 (1990).
52. The value of Θ at the singular points (0,y) is arbitrary since this corresponds to points in the integration that precisely satisfy k=p cos r, which may be removed since the integrand is bounded.
53. R. H. Kraichnan, Phys. Fluids 7, 1030 (1964).
54. The difference between this algorithm and the one used for the energy equation appears only above second order.
55. If the error was computed solely by the ratio of the predictor-corrector difference to the current value of a corrected variable that happened to be passing through zero, this relative error could become arbitrarily large. The error is therefore computed with respect to both current and previous values; our formula removes the artifact just described by calculating a combination of relative and absolute errors.
56. These tolerance parameters are chosen to yield an efficient evolution that does not become unstable. Of course, even if the evolution is stable, it may still be inaccurate; in this case the maximum error tolerance should be decreased. In practice, our convergence tests indicate that whenever the numerical scheme is stable, the results are sufficiently accurate, bearing in mind the inherent inaccuracies of statistical closures.