next previous
Up: Performance analysis for

A. Appendix

As demonstrated by Sasiela (Sasiela 1994), the properties of Mellin transforms (see for instance (Colombo 1959; Dautray & Lions 1987)) and Gamma functions are useful in problems dealing with wave propagation in turbulence. This technique is applied here to solve the integral in Eq. (12 (click here)), Eq. (24 (click here)), Eq. (34 (click here)) and Eq. (35 (click here)).
The Mellin transform pair is given by:
displaymath3198

equation1124
The general form of the integral to solve has been presented by Chassat (Chassat 1989) and can be written (tex2html_wrap_inline3212):
 equation1135
From the properties of the Mellin transforms (Colombo 1959; Dautray & Lions 1987) and using tex2html_wrap_inline3214, the integral becomes:
 eqnarray1151
Tables of Mellin transforms (Sasiela 1994; Colombo 1959; Dautray & Lions 1987) are helpful to solve Eq. (A3 (click here)). Moreover, the Mellin transform of functions can usually be expressed as the ratio of Gamma functions. We use the notation:
equation1170
We have for Jn the tex2html_wrap_inline3218 order Bessel function of the first kind:
displaymath3199

 equation1193
It leads to the Mellin-Barnes integral:
displaymath3200

 equation1207
This integration can be performed using the method of pole residues. The value of the integral, as given by Cauchy's formula, is just 2itex2html_wrap_inline3220 times the sum of the residues at the enclosed poles. The result can be expressed in terms of generalised hypergeometric functions (Abramovitz & Stegun 1965). For instance, in the case tex2html_wrap_inline3222 1, one of the interesting cases, the integral is given by:
displaymath3201

displaymath3202

equation1252
where:
displaymath3203

 equation1272
When the function is expressed in terms of generalised hypergeometric functions, the integral leads to the more restrictive convergence condition which can be expressed as: tex2html_wrap_inline3224. This condition determines the maximum field of view which can be reached by this method.
The limited convergence domain encountered in the evaluation of integrals involving the product of three Bessel functions has been first remarked by Tyler (Tyler 1990). Chassat proposed a method using the Bessel recurrency law (Chassat 1992):
 equation1282
Two easily evaluable Mellin integrals appear, tex2html_wrap_inline3226 and tex2html_wrap_inline3228
equation1289

equation1296
With the notation tex2html_wrap_inline3230 and tex2html_wrap_inline3232, Eq. (A2 (click here)) becomes:
eqnarray1306
The Bessel reccurency law becomes reccurency law between Mellin integrals which can be written as:
eqnarray1314
Now, we can use the Mellin transform tables to solve tex2html_wrap_inline3234 and tex2html_wrap_inline3234 which can be expressed as Mellin-Barnes integrals of the following type:
equation1316
where G and F are the Mellin transform of g and f respectively. g and f are defined as:
equation1324

 equation1333
We have the Mellin transforms:
displaymath3204

equation1358

eqnarray1362

equation1379
The Mellin transform of f(Kx) is given by Eq. (A2 (click here)). Finally, using the Cauchy's formula, we find:
tex2html_wrap_inline3252
eqnarray1385

eqnarray1433
tex2html_wrap_inline3254
eqnarray1478

eqnarray1498


next previous
Up: Performance analysis for

Copyright by the European Southern Observatory (ESO)