Basis Dispersionlet Bispectral Analysis*

Ritz-type bispectral analysis¹ is a class of techniques that use statistical correlations in turbulence fluctuation data to infer the underlying model equations.  This offers a novel approach to theory-experiment comparison, and more recent algorithms can additionally infer profile functions or topological boundaries from fluctuation data.  Out of this class, basis dispersionlet analysis is a technique in which model equations are represented using wavelet-like operators, or dispersionlets.  This offers a compromise between accurate modeling of dispersion relations, as with the basis function method, and accurate modeling of profile functions, as with the basis operator method. 

The present work will demonstrate this technique on SOLT simulation data.  It will also compare different dispersionlet sets by a variety of metrics, such as residual error, accuracy of dispersion curve, and ability to resolve spatial features in the model. 

 

References:

1.  Ch. P. Ritz and E. J. Powers, Physica D 20, 320 (1986).