1: \documentclass[aps,twocolumn]{revtex4}%
2: \pdfoutput=1
3: \usepackage{amsfonts}
4: \usepackage{amsmath}
5: \usepackage{amssymb}
6: \usepackage{graphicx}%
7: \setcounter{MaxMatrixCols}{30}
8: %TCIDATA{OutputFilter=latex2.dll}
9: %TCIDATA{Version=5.00.0.2606}
10: %TCIDATA{CSTFile=revtex4.cst}
11: %TCIDATA{Created=Tuesday, February 19, 2008 16:42:21}
12: %TCIDATA{LastRevised=Sunday, July 20, 2008 12:44:20}
13: %TCIDATA{<META NAME="GraphicsSave" CONTENT="32">}
14: %TCIDATA{<META NAME="SaveForMode" CONTENT="1">}
15: %TCIDATA{BibliographyScheme=Manual}
16: %TCIDATA{<META NAME="DocumentShell" CONTENT="Articles\SW\REVTeX 4">}
17: %TCIDATA{Language=American English}
18: \newtheorem{theorem}{Theorem}
19: \newtheorem{acknowledgement}[theorem]{Acknowledgement}
20: \newtheorem{algorithm}[theorem]{Algorithm}
21: \newtheorem{axiom}[theorem]{Axiom}
22: \newtheorem{claim}[theorem]{Claim}
23: \newtheorem{conclusion}[theorem]{Conclusion}
24: \newtheorem{condition}[theorem]{Condition}
25: \newtheorem{conjecture}[theorem]{Conjecture}
26: \newtheorem{corollary}[theorem]{Corollary}
27: \newtheorem{criterion}[theorem]{Criterion}
28: \newtheorem{definition}[theorem]{Definition}
29: \newtheorem{example}[theorem]{Example}
30: \newtheorem{exercise}[theorem]{Exercise}
31: \newtheorem{lemma}[theorem]{Lemma}
32: \newtheorem{notation}[theorem]{Notation}
33: \newtheorem{problem}[theorem]{Problem}
34: \newtheorem{proposition}[theorem]{Proposition}
35: \newtheorem{remark}[theorem]{Remark}
36: \newtheorem{solution}[theorem]{Solution}
37: \newtheorem{summary}[theorem]{Summary}
38: \newenvironment{proof}[1][Proof]{\noindent\textbf{#1.} }{\ \rule{0.5em}{0.5em}}
39: \begin{document}
40: \title{Singularities in Speckled Speckle: Screening }
41: \author{David A. Kessler}
42: \affiliation{Physics Department, Bar-Ilan University, Ramat-Gan IL52900, Israel}
43: \author{Isaac Freund}
44: \affiliation{Physics Department, Bar-Ilan University, Ramat-Gan IL52900, Israel}
45: \date{14 July 2008}
46:
47: \begin{abstract}
48: We study screening of optical singularities in random optical fields with two
49: widely different length scales. \ We call the speckle patterns generated by
50: such fields \emph{speckled speckle}, because the major speckle spots in the
51: pattern are themselves highly speckled. \ We study combinations of fields
52: whose components exhibit short- and long-range correlations, and find unusual
53: forms of screening.\ \
54:
55: \end{abstract}
56: \maketitle
57:
58:
59: \section{INTRODUCTION}
60:
61: Screening of charged topological singularities - vortices [1, 2 (Chap. 5),
62: $3$ (Sect. $4.8$)] in scalar fields, C points [$2,$ (Chaps. $12$ and $13$)] in
63: vector fields - has been extensively studied in random fields with a single
64: correlation length [$4-19$]; here we study screening of these singularities in
65: random fields with two widely different correlation lengths. \ We call such
66: fields \textquotedblleft speckled speckle\textquotedblright\ because, as
67: illustrated in Fig. \ref{Fig1}(a), the major speckle spots of the field are
68: themselves highly speckled. \ Speckled speckle fields can be generated by
69: illuminating a random diffuser with two concentric, overlapping beams: one,
70: the $a$ beam, is tightly focused and intense, the other, the $b$ beam, is
71: weak and diffuse.
72:
73: The statistical properties of speckled speckle can be highly anomalous, with
74: relative number densities of critical points (vortices, C points, extrema, and
75: umbilic points) differing from normal \ speckle values by orders of magnitude
76: [$20,21$]. The spatial arrangement of vortices and C points is also anomalous,
77: with these singularities forming dense clusters of a kind not found in normal
78: speckle fields, Fig. \ref{Fig1}(a) [$20,21$].
79:
80: Screening can be either short- or long-ranged. \ Nonsingular random sources
81: produce random fields that exhibit short range screening [$4-19$]. \ In such
82: systems positive (negative) topological charges are surrounded by a local net
83: excess of negative (positive) charge, leading to charge neutrality within a
84: characteristic distance, the screening length, that can be less than the
85: average separation between charges [$19$].
86:
87: Singular sources, such as a ring of finite radius but zero width [$8$],
88: produce random fields that exhibits long-range screening [$8,16,19$]. \ The
89: singularities in the field produced by a ring form a quasi-lattice in which
90: positive/negative singularities occupy alternate corners of a square cell,
91: Fig. \ref{Fig1}(b). \ Local defects in the lattice destroy the local charge
92: neutrality that would produce short range screening, and screening sets in
93: only asymptotically.
94:
95: Thus, both short- and long-range screening depend upon the spatial
96: arrangements of the charges. \ These arrangements are anomalous in speckled
97: speckle, which can therefore be expected to exhibit unusual forms of screening.
98:
99: A major observable consequence of screening is strong damping of fluctuations
100: of the topological charge $Q$. \ These fluctuations are characterized
101: by their variance $\left\langle Q^{2}\right\rangle $; the behavior of
102: $\left\langle Q^{2}\right\rangle $ for speckled speckle is our main concern here.
103:
104: The plan of this paper is as follows. \ In Section II we discuss the charge
105: variance in a bounded region, in Section III we review $\left\langle
106: Q^{2}\right\rangle $ in normal speckle for fields with short- and with
107: long-range correlations, and in Sections IV-VII we present results for
108: $\left\langle Q^{2}\right\rangle $ for four qualitatively different forms of
109: speckled speckle. \ We briefly summarize our main findings in the concluding
110: Section VIII.%
111:
112: %TCIMACRO{\FRAME{fbpFU}{1.727in}{2.9023in}{0pt}{\Qcb{Vortex structures.
113: %\ Positive (negative) vortices are shown by white (black) filled circles.
114: %\ (a) Speckled speckle. \ A random diffuser is illuminated by two concentric
115: %disks of light, $a$ and $b$. \ The diameter of disk $b$ is ten times the
116: %diameter of $a$; the total optical power in $a$, however, is $10$ times that
117: %in $b$. \ Major (minor) speckle spots in the speckled speckle field are due
118: %primarily to beam $a$ ($b$). \ Vortices of the combined beam cluster in the
119: %dark regions between $a$ field speckle spots because they require perfect
120: %destructive interference between the strong $a$ and weak $b$ fields. \ In
121: %normal speckle produced by a single disk vortices tend to be uniformly
122: %distributed with only minor clustering. \ (b) Normal speckle phase map
123: %produced by a random diffuser illuminated with a single ring; the vortices
124: %tend to form a lattice with a square unit cell. }}{\Qlb{Fig1}}%
125: %{sssscreening_fig1.eps}{\special{ language "Scientific Word";
126: %type "GRAPHIC"; maintain-aspect-ratio TRUE; display "USEDEF";
127: %valid_file "F"; width 1.727in; height 2.9023in; depth 0pt;
128: %original-width 3.8623in; original-height 6.5302in; cropleft "0";
129: %croptop "1"; cropright "1"; cropbottom "0";
130: %filename '../../../Current/SSSScreening/SSSScreening ScientificWord/SSSScreening Figs/SSSScreening_Fig1.eps';file-properties "XNPEU";}%
131: %}}%
132: %BeginExpansion
133: \begin{figure}
134: [pb]
135: \begin{center}
136: \includegraphics[
137: natheight=6.530200in,
138: natwidth=3.862300in,
139: height=2.9023in,
140: width=1.727in
141: ]%
142: {SSSScreening_Fig1.pdf}%
143: \caption{Vortex structures. \ Positive (negative) vortices are shown by white
144: (black) filled circles. \ (a) Speckled speckle. \ A random diffuser is
145: illuminated by two concentric disks of light, $a$ and $b$. \ The diameter of
146: disk $b$ is ten times the diameter of $a$; the total optical power in $a$,
147: however, is $10$ times that in $b$. \ Major (minor) speckle spots in the
148: speckled speckle field are due primarily to beam $a$ ($b$). \ Vortices of the
149: combined beam cluster in the dark regions between $a$ field speckle spots
150: because they require perfect destructive interference between the strong $a$
151: and weak $b$ fields. \ In normal speckle produced by a single disk vortices
152: tend to be uniformly distributed with only minor clustering. \ (b) Normal
153: speckle phase map produced by a random diffuser illuminated with a single
154: ring; the vortices tend to form a lattice with a square unit cell. }%
155: \label{Fig1}%
156: \end{center}
157: \end{figure}
158: %EndExpansion
159:
160:
161: \section{CHARGE VARIANCE IN SHORT- AND LONG-RANGE SCREENING}
162:
163: We assume isotropy, circular Gaussian statistics [$3$ (Chap 2), $22$ (Chap.
164: $2$)], and a circular region of radius $R$. \ The charge variance $\left\langle
165: Q^{2}\right\rangle $ in this region is related to the autocorrelation function
166: of the field $W(r)$ by [$19$ (Eq. 39)],%
167: \begin{equation}
168: \langle Q^{2}\rangle=\frac{1}{2\pi}\int_{0}^{2R}\sqrt{4R^{2}-r^{2}}%
169: \frac{(W^{\prime}(r))^{2}}{1-W^{2}(r)}dr. \label{Q2David}%
170: \end{equation}
171: where $W^{\prime}(r)=dW(r)/dr$. \ The number density of charges $\eta$ is
172: [$23$]%
173: \begin{equation}
174: \eta=-\frac{W^{\prime\prime}(0)}{2\pi}. \label{eta}%
175: \end{equation}
176:
177:
178: For the short range correlations produced by extended sources, $W^{\prime
179: }(r) $ decays rapidly with $r$, and in the limit of large $R$,%
180: \begin{equation}
181: \langle Q^{2}\rangle\approx\frac{R}{\pi}\int_{0}^{\infty}\frac{(W^{\prime
182: }(r))^{2}}{1-W^{2}(r)}dr. \label{Q2short}%
183: \end{equation}
184: Thus, for short-range screening $\left\langle Q^{2}\right\rangle $ grows with
185: the perimeter, i.e. $\left\langle Q^{2}\right\rangle \sim R\sim\sqrt{N}$, in
186: contrast to the case of no screening, where $\left\langle Q^{2}\right\rangle $
187: grows with the area, $\left\langle Q^{2}\right\rangle \sim R^{2}\sim N$
188: [$8,19$].
189:
190: For the long-range screening produced by a singular ring source of radius $p$
191: and zero width, the large $R$ limit is [$19$ (Eq. 48)],
192: \begin{subequations}
193: \label{LongRangeQ2}%
194: \begin{align}
195: \left\langle Q^{2}\right\rangle & \approx\dfrac{pR}{\pi^{2}}\left[
196: {\cal{K}}+\ln\left( \rho R\right) \right] ,\label{LRQ2}\\
197: {\cal{K}} & =\pi{\cal{D}}+\gamma+5\ln2-3\approx2.81182,\label{DavidsK}\\
198: {\cal{D}} & =\int_{0}^{\infty}dx\,\frac{J_{0}^{2}(x)J_{1}^{2}(x)}%
199: {1-J_{0}^{2}(x)}\approx0.563047, \label{DavidsD}%
200: \end{align}
201: where $\gamma\approx0.577216$ is Euler's constant. \ Thus, long-range
202: screening yields a charge variance that grows asymptotically as $R\ln R$; this
203: growth rate is significantly faster than the short-range growth rate
204: proportional to $R$, but is very much slower than the unscreened rate
205: proportional to $R^{2}$. \ For, say, a large region that contains $10^{4}$
206: charges, short-range (long-range) screening damps out charge fluctuations
207: relative to no screening by a factor of $\sim100$ ($\sim22$).
208:
209: \section{CHARGE VARIANCE IN NORMAL SPECKLE}
210:
211: We review here the charge variance in normal speckle produced by sources with
212: a single characteristic length scale. \ In later sections we build our
213: composite sources with their two different length scales from binary
214: combinations of these single sources, and compare composite-source charge
215: variances with single-source variances.
216:
217: \subsection{Source Distributions and Autocorrelation Functions}
218:
219: Listed below are the source distributions $S(u) $, where $u$
220: measures radial displacements in the source plane, the total optical power in each
221: source, $P$, and the autocorrelation functions $W(r) $ of
222: the speckle field. \ $S(u)$ and $W(r)$ are
223: related by the VanCittert-Zernike theorem [$3$ (Sect. 4), $22$ (Sect. 5.6)].
224: \ We study four fields with autocorrelation functions that decay at different rates.
225:
226: (\emph{i}) A Gaussian, superscript (G), of $1/e$ width $p$, and intensity
227: (optical power/unit area) $I^{\left( \text{G}\right) }$ at the peak;
228: henceforth the \textquotedblleft Gaussian\textquotedblright,
229: \end{subequations}
230: \begin{subequations}
231: \label{Gauss}%
232: \begin{align}
233: S^{\left( \text{G}\right) }(u) & =I^{\left( \text{G}\right) }\exp\left(
234: -\left[ u/\left( 2p\right) \right] ^{2}\right) ,\label{Gauss_S}\\
235: P^{\left( \text{G}\right) } & =4\pi p^{2}I^{\left( \text{G}\right)
236: },\label{Gauss_P}\\
237: W^{(\text{G})}(r) & =\exp\left( - p^2r^2\right) . \label{Gauss_omega}%
238: \end{align}
239:
240:
241: (\emph{ii}) A uniform disk, superscript (D), of radius $p$, and uniform
242: intensity $I^{\left( \text{D}\right) }$; henceforth the \textquotedblleft
243: Disk\textquotedblright,
244: \end{subequations}
245: \begin{subequations}
246: \label{Disk}%
247: \begin{align}
248: S^{\left( \text{D}\right) }(u) & =I^{\left( \text{D}\right) }%
249: \Theta\left( u-p\right) ,\label{Disk_S}\\
250: P^{\left( \text{D}\right) } & =\pi p^{2}I^{\left( \text{D}\right)
251: },\label{Disk_P}\\
252: W^{\left( \text{D}\right) }(r) & =2J_{1}\left( pr\right)
253: /\left(pr\right) . \label{Disk_omega}%
254: \end{align}
255: $\Theta\left( x\right) $ is the Heaviside step function defined by
256: $\Theta\left( x\leq0\right) =0$, $\Theta\left( x>0\right) =1$.
257:
258: (\emph{iii}) A nonuniform (inverse square root) disk, superscript (S), of
259: radius $p$, with intensity $I^{\left( \text{S}\right) }$ at the disk center,%
260: \end{subequations}
261: \begin{subequations}
262: \begin{align}
263: S^{\left( \text{S}\right) }\left( u\right) & =\frac{I^{\left(
264: \text{S}\right) }}{\sqrt{1-\left( u/p\right) ^{2}}}\Theta\left(
265: u-p\right) ,\label{Sinc_S}\\
266: P^{\left( \text{S}\right) } & =2\pi p^{2}I^{\left( \text{S}\right)
267: },\label{Sinc_P}\\
268: W^{\left( \text{S}\right) }(r) & =I^{\left( \text{S}%
269: \right) }\text{sinc}\left( pr\right) , \label{Sinc_omega}%
270: \end{align}
271: where, sinc$\left( x\right) \equiv\sin\left( x\right) /x$. \ In what
272: follows we refer to this source as the \textquotedblleft
273: Sinc\textquotedblright.
274:
275: (\emph{iv}) A singular ring, superscript (R), of radius $p$, which we write
276: as
277: \end{subequations}
278: \begin{subequations}
279: \label{Ring}%
280: \begin{align}
281: S^{\left( \text{R}\right) }(u) & =I^{\left( \text{R}\right) }%
282: \varepsilon\delta\left( u-p\right) ,\label{Ring_S}\\
283: P^{\left( \text{R}\right) } & =2\pi p\varepsilon I^{\left( \text{R}%
284: \right) },\label{Ring_P}\\
285: W^{\left( \text{R}\right) }(r) & =J_{0}(pr) ,
286: \label{Ring_Omega}%
287: \end{align}
288: where $\delta\left( x\right) $ is the Dirac delta function. \ In what
289: follows we refer to this source as the \textquotedblleft
290: Ring\textquotedblright.
291:
292: $S^{\left( \text{R}\right) }(u)$ is the limit of a finite width annulus
293: $s^{\left( \text{R}\right) }\left( u\right) $ of mean radius $p$ and width
294: $\varepsilon$,
295: \end{subequations}
296: \begin{subequations}
297: \label{Annulus}%
298: \begin{gather}
299: s\left( u\right) =\Theta\left( u-p-\varepsilon/2\right) -\Theta\left(
300: u-p+\varepsilon/2\right) ,\label{Annulus_s}\\
301: \lim_{\varepsilon\rightarrow0}\left[ s\left( u\right) /\varepsilon\right]
302: =\delta\left( u-p\right) . \label{Annulus_S}%
303: \end{gather}
304: $I^{\left( \text{R}\right) }$ is therefore the uniform intensity in the
305: annulus. \ In the limit $\varepsilon\rightarrow0$, $I^{\left( \text{R}%
306: \right) }$ diverges, $P^{\left( \text{R}\right) }$ in Eq. (\ref{Ring_P}),
307: however, is assumed to remain finite.
308:
309: $W$ and $\left\langle Q^{2}\right\rangle $ for the finite width annulus
310: $\left( \varepsilon>0\right) $ is discussed in [$19$ (Sect. 5)], where it is
311: shown that over the region $\varepsilon r<1$ the experimentally attainable
312: annulus is an excellent approximation to the theoretical singular ring.
313:
314: \subsection{Charge Variance}
315:
316: In Fig. \ref{Fig2} we plot $\left\langle Q^{2}\right\rangle /\left(
317: pR\right) $ vs. $pR$ for the above four sources. \ For the Gaussian (G), Disk
318: (D), and Sinc (S), screening is short-ranged, and the large $R$ limit of
319: $\left\langle Q^{2}\right\rangle $ is given in Eq. (\ref{Q2short}).
320:
321: For the Gaussian, Eq. (\ref{Q2short}) can be evaluated analytically, yielding%
322: \end{subequations}
323: \begin{equation}
324: \left\langle Q^{2}\right\rangle _{Gauss}\approx \tfrac{1}{4}\sqrt{2/\pi}\zeta\left(
325: 3/2\right) pR=0.521093pR, \label{LambdaScrnGauss}%
326: \end{equation}
327: with $\zeta\left( x\right) $ is the Riemann zeta function, whereas for the
328: Disk (D) and Sinc (S), Eq. (\ref{Q2short}) is evaluated numerically, yielding%
329: \begin{equation}
330: \left\langle Q^{2}\right\rangle _{Disk}\approx0.227210pR, \label{Q2pRDisk}%
331: \end{equation}
332: and%
333: \begin{equation}
334: \left\langle Q^{2}\right\rangle _{\operatorname{Si}nc}\approx0.305898pR.
335: \label{Q2pRSinc}%
336: \end{equation}
337:
338:
339: For the Ring (R) screening is long-ranged, and the large $R$ limit of
340: $\left\langle Q^{2}\right\rangle /\left( pR\right) $ is given in Eq.
341: (\ref{LongRangeQ2}).
342:
343: For very small $R$ for all sources [$19$ (Eq. 82)],%
344: \begin{equation}
345: \left\langle Q^{2}\right\rangle _{R\rightarrow0}\approx\eta\pi R^{2}=N,
346: \label{SmallRlimit}%
347: \end{equation}
348: i.e. there is no screening; the reason is that for a sufficiently small area
349: the probability of finding the required screening charges within the area is
350: vanishingly small. \ This result is illustrated in Fig. \ref{Fig2}(c) for all four sources.%
351:
352: %TCIMACRO{\FRAME{ftpFU}{2.4587in}{5.1301in}{0pt}{\Qcb{Charge variance
353: %$\left\langle Q^{2}\right\rangle $ for normal speckle obtained from numerical
354: %integration of Eq. (\ref{Q2David}). \ (a) Short-range screening. \ (b)
355: %Long-range screening. \ The dependence on the parameter $p$ in Eqs.
356: %(\ref{Gauss})-(\ref{Ring}) is here scaled out by plotting $\left\langle
357: %Q^{2}\right\rangle /\left( pR\right) $ vs. $pR$. \ As can be seen, for
358: %$pR>1$, the results in (a) quickly asymptote to the theoretical values in Eqs.
359: %(\ref{LambdaScrnGauss})-(\ref{Q2pRSinc}), whereas the result in (b) asymptotes
360: %to the theoretical form (thin straight line) in Eq. (\ref{LongRangeQ2}). \ (c)
361: %$\left\langle Q^{2}\right\rangle $ for small $R$. \ In all four cases (G -
362: %Gauss, D - Disk, S - Sinc, and R - Ring), the curves approach the
363: %$R\rightarrow0$ limit given in Eq. (\ref{SmallRlimit}). }}{\Qlb{Fig2}%
364: %}{sssscreening_fig2.tif}{\special{ language "Scientific Word";
365: %type "GRAPHIC"; maintain-aspect-ratio TRUE; display "USEDEF";
366: %valid_file "F"; width 2.4587in; height 5.1301in; depth 0pt;
367: %original-width 3.9868in; original-height 8.3688in; cropleft "0";
368: %croptop "1"; cropright "1"; cropbottom "0";
369: %filename '../../../Current/SSSScreening/SSSScreening ScientificWord/SSSScreening Figs/SSSScreening_Fig2.tif';file-properties "XNPEU";}%
370: %}}%
371: %BeginExpansion
372: \begin{figure}
373: [pt]
374: \begin{center}
375: \includegraphics[
376: natheight=8.368800in,
377: natwidth=3.986800in,
378: height=5.1301in,
379: width=2.4587in
380: ]%
381: {SSSScreening_Fig2.pdf}%
382: \caption{Charge variance $\left\langle Q^{2}\right\rangle $ for normal speckle
383: obtained from numerical integration of Eq. (\ref{Q2David}). \ (a) Short-range
384: screening. \ (b) Long-range screening. \ The dependence on the parameter $p$
385: in Eqs. (\ref{Gauss})-(\ref{Ring}) is here scaled out by plotting
386: $\left\langle Q^{2}\right\rangle /\left( pR\right) $ vs. $pR$. \ As can be
387: seen, for $pR>1$, the results in (a) quickly asymptote to the theoretical
388: values in Eqs. (\ref{LambdaScrnGauss})-(\ref{Q2pRSinc}), whereas the result in
389: (b) asymptotes to the theoretical form (thin straight line) in Eq.
390: (\ref{LongRangeQ2}). \ (c) $\left\langle Q^{2}\right\rangle $ for small $R$.
391: \ In all four cases (G - Gauss, D - Disk, S - Sinc, and R - Ring), the curves
392: approach the $R\rightarrow0$ limit given in Eq. (\ref{SmallRlimit}). }%
393: \label{Fig2}%
394: \end{center}
395: \end{figure}
396: %EndExpansion
397:
398:
399: \section{COMPOSITE SOURCES AND THEIR AUTOCORRELATION FUNCTIONS}
400:
401: For scalar (single component) fields the $a$ and $b$ beams have the same, say,
402: linear polarization, and the singularities whose screening is of interest here
403: are the phase vortices [$1-19$]. \ For vector (two component) fields the $a$
404: and $b$ beams have orthogonal linear polarizations, and the relevant
405: singularities that screen each other are either right- or left-handed C points
406: [$2$]: right-handed C points do not screen left-handed ones, and vice versa.
407:
408: We write our composite source as%
409: \begin{equation}
410: S^{\left( \text{T}_{a}\text{T}_{b}\right) }\left( u\right) =S_{a}^{\left(
411: \text{T}\right) }\left( u\right) +S_{b}^{\left( \text{T}%
412: \right) }\left( u\right) , \label{CompositeS}%
413: \end{equation}
414: where the source type specifier $\left( \text{T}_{a}\text{T}_{b}\right) $ is
415: a binary combination of T$_{a,b}=$ G, D ,S, R. \ From the VanCittert-Zernike
416: theorem, the corresponding autocorrelation function is%
417: \begin{equation}
418: W^{\left( \text{T}_{a}\text{T}_{b}\right) }(r) =\frac
419: {W_a^{\left( \text{T}\right) }\left(r\right) +K^{\left( \text{T}_a%
420: \text{T}_b\right) }W_b^{\left( \text{T}\right) }\left(r\right)
421: }{1+K^{\left( \text{T}_{a}\text{T}_{b}\right) }}, \label{CompositeW}%
422: \end{equation}
423: where the dimensionless constant
424: \begin{equation}
425: K^{\left( \text{T}_{a}\text{T}_{b}\right) }=P_{b}^{\left( \text{T}%
426: \right) }/P_{a}^{\left( \text{T}\right) }. \label{K}%
427: \end{equation}
428: $S_{a,b}^{\left( \text{T}\right) }(u) $, $P_{a,b}^{\left(
429: \text{T}\right) }$, and $W_{a,b}^{\left( \text{T}\right) }\left(
430: r\right) $, are listed in Eqs. (\ref{Gauss})-(\ref{Ring}), with $p=a,b$ as appropriate.
431:
432: Similarly, the number density of singularities is for composite scalar fields,%
433: \begin{equation}
434: \eta^{\left( \text{T}_{a}\text{T}_{b}\right) }=-\frac{1}{2\pi}%
435: \frac{W_a^{\prime\prime\left( \text{T}\right) }(0)
436: +K^{\left( \text{T}_{a}\text{T}_{b}\right) }W_b^{\prime\prime\left(
437: \text{T}\right) }(0) }{1+K^{\left( \text{T}_{a}%
438: \text{T}_{b}\right) }}. \label{CompositeEta}%
439: \end{equation}
440: This is also the number density of right- and of left-handed C points.
441:
442: As a specific example, for the intense, tightly focused $a$ beam a Gaussian
443: (G), and the diffuse weak $b$ beam a Sinc (S),
444: \begin{subequations}
445: \label{CompositeExample}%
446: \begin{align}
447: S_{a}^{\left( \text{G}\right) }(u) & =I_{a}^{\left( \text{G}\right) }%
448: \exp\left( -\left[ u/\left( 2a\right) \right] ^{2}\right)
449: ,\label{CompositeExample_Sa}\\
450: S_{b}^{\left( \text{S}\right) }\left( u\right) & =\frac{I_{b}^{\left(
451: \text{S}\right) }}{\sqrt{1-\left( u/b\right) ^{2}}}\Theta\left(
452: u-b\right) ,\label{CompositeExample_Sb}\\
453: W^{\left( \text{GS}\right) }\left( r\right) & =\frac{\exp\left(
454: -ar\right) +K^{\left( \text{GS}\right) }\text{sinc}\left( br\right)
455: }{1+K^{\left( \text{GS}\right) }},\label{CompositeExample_W}\\
456: K^{\left( \text{GS}\right) } & =\left[ 2b^{2}I_{b}^{\left( \text{S}%
457: \right) }\right] /\left[ a^{2}I_{a}^{\left( \text{G}\right) }\right]
458: ,\label{CompositeExample_K}\\
459: \eta^{\left( \text{GS}\right) } & =\frac{2a^{2}+K^{\left( \text{GS}%
460: \right) }b^{2}/3}{1+K^{\left( \text{GS}\right) }}%
461: ,\label{CompositeExample_Eta}%
462: \end{align}
463:
464:
465: In the composite-source examples that follow we usually take $K\sim0.01$ and
466: $b/a=100$. \ There are two reasons for these choices: (\emph{i}) small $K$
467: together with large $b/a$ produces results that vividly illustrate the unusual
468: screening properties of speckled speckle, and (\emph{ii}) this combination of
469: parameters permits a significant degree of analysis. \ We consider that these
470: parameters, which are convenient for the theoretician, to be experimentally
471: possible; admittedly, they may be difficult to achieve in practice.
472:
473: With the above choice of parameters, the $R$ dependence of $\left\langle
474: Q^{2}\right\rangle $ separates into three distinct regions:
475:
476: \bigskip
477:
478: I. $R<1/b$.
479:
480: In this region Eq. (\ref{SmallRlimit}) holds for all composite sources with
481: $\eta$ equal to the number density of $b$ field charges%
482: \end{subequations}
483: \begin{equation}
484: \eta\approx\eta_{b}\approx-\frac{K}{2\pi}W_b^{\prime\prime\left( \text{T}\right)
485: }(0) .\label{Omegab}%
486: \end{equation}
487: The reason is that for sufficiently small $R$ the probability of finding an
488: $a$ beam charge in the area is negligible; only $b$ beam charges are present,
489: so only these charges contribute to $\left\langle Q^{2}\right\rangle $. \ This
490: result is verified by direct comparison (not shown) with the exact result in
491: Eq. (\ref{Q2David}).
492:
493: \bigskip
494:
495: II. $1/b<R<1/a$.
496:
497: In this region, $a$ in Eq. (\ref{CompositeW}) may be set equal to zero,
498: because for $ar<1$, $W_a(r) \approx W_a(0) =1$, and
499: $W_a^{\prime}(r) \approx W_a^{\prime}(0) =0$,
500: independent of the $a$ beam type - G, D, S, or R. \ As will become apparent, these
501: approximations yield good agreement with the exact result in Eq.
502: (\ref{Q2David}).
503:
504: \bigskip
505:
506: III. $R>1/a.$
507:
508: In this region there is no accurate approximation that is applicable, however,
509: as discussed below, approximations good to $\sim10\%$ are available.
510:
511: Below we discuss screening in composite beams for the four qualitatively
512: different combinations of $a$ and $b$ beams in which the individual beams
513: exhibit either short- or long-range correlations.
514:
515: \section{BOTH BEAM $a$ AND BEAM $b$ EXHIBIT SHORT-RANGE SCREENING}
516:
517: We start with region II, $1/b<R<1/a$. \ Here the area contains many $b$
518: charges but practically no $a$ charges. \ We denote the $b$ charge
519: contribution to $\left\langle Q^{2}\right\rangle $ by $\left\langle
520: Q^{2}\right\rangle _{b}$, where%
521: \begin{equation}
522: \left\langle Q^{2}\right\rangle _{b}\approx\frac{R}{2\pi}K^{\left(
523: \text{T}_{a}\text{T}_{b}\right) }\int_{0}^{\infty}\frac{\left[ dW_b^{\left(
524: \text{T}\right) }(r) /dr\right] ^{2}}{1-W_b^{\left(
525: \text{T}\right) }(r) }dr. \label{SSQ2b}%
526: \end{equation}
527: In obtaining this result we make use of the fact that $K\ll1$ and $\left[
528: KW_b^{\left( \text{T}\right) }\right] ^{2}\ll W_b^{\left( \text{T}%
529: \right) }$.
530:
531: For $b$ a Gaussian (G), Eq. (\ref{SSQ2b}) can be evaluated analytically, and
532: we have,%
533: \begin{align}
534: \left\langle Q^{2}\right\rangle _{b}^{\left( G\right) } & \approx\frac
535: {K^{\left( \text{T}_{a}\text{G}\right) }b\left[ \zeta\left( \frac{3}%
536: {2}\right) -1\right] }{2\sqrt{\pi}}R,\label{SSQ2bG}\\
537: & =0.454843K^{\left( \text{T}_{a}\text{G}\right) }bR.\nonumber
538: \end{align}
539:
540:
541: For $b$ a Disk or a Sinc, Eq. (\ref{SSQ2b}) is evaluated numerically, and we
542: have for the Disk (D),%
543: \begin{equation}
544: \left\langle Q^{2}\right\rangle _{b}^{\left( D\right) }\approx0.187153K^{\left(
545: \text{T}_{a}\text{D}\right) }bR, \label{SSQ2bD}%
546: \end{equation}
547: and for the Sinc (S),%
548: \begin{equation}
549: \left\langle Q^{2}\right\rangle _{b}^{\left( S\right) }\approx0.238531K^{\left(
550: \text{T}_{a}\text{S}\right) }bR. \label{SSQ2bS}%
551: \end{equation}
552:
553:
554: In region III, $R>1/a$, we assume that the $b$ charges continue to contribute
555: $\left\langle Q^{2}\right\rangle _{b}$ to $\left\langle Q^{2}\right\rangle $.
556: \ In addition, the area now includes many $a$ charges. \ We label the
557: contribution of these charges $\left\langle Q^{2}\right\rangle _{a}$, and
558: write%
559: \begin{equation}
560: \left\langle Q^{2}\right\rangle \sim\left\langle Q^{2}\right\rangle
561: _{a}+\left\langle Q^{2}\right\rangle _{b}, \label{SSQaPlusQb}%
562: \end{equation}
563: i.e., we neglect interactions between the $a$ and $b$ beams, and the
564: possibility that $a$ charges can screen $b$ charges, and vice versa. \ Within
565: the framework of this approximation we write,%
566: \begin{equation}
567: \left\langle Q^{2}\right\rangle _{a}^{\left( G\right) }=\left\langle
568: Q^{2}\right\rangle _{Gauss},\text{ }\left[ \text{Eq.(\ref{LambdaScrnGauss}%
569: )}\right] \label{SSQ2aG}%
570: \end{equation}%
571: \begin{equation}
572: \left\langle Q^{2}\right\rangle _{a}^{\left( D\right) }=\left\langle
573: Q^{2}\right\rangle _{Disk},\text{\ }\left[ \text{Eq.(\ref{Q2pRDisk})}\right]
574: \label{SSQ2aD}%
575: \end{equation}%
576: \begin{equation}
577: \left\langle Q^{2}\right\rangle _{a}^{\left( S\right) }=\left\langle
578: Q^{2}\right\rangle _{\operatorname{Sinc}},\text{ }\left[
579: \text{\ Eq.(\ref{Q2pRSinc})}\right] \label{SSQ2aSinc}%
580: \end{equation}
581: with $p$ replaced by $a$.%
582:
583: %TCIMACRO{\FRAME{ftpFU}{2.4102in}{4.9986in}{0pt}{\Qcb{$\left\langle
584: %Q^{2}\right\rangle /R$ vs. $R$ for composite sources in which both the $a$ and
585: %$b$ beams have short range correlations. \ The thick curves are the exact
586: %result in Eq. (\ref{Q2David}). \ The solid lines labeled B are the theory in
587: %Eqs. (\ref{SSQ2b})- (\ref{SSQ2bS}), the dashed lines labeled A are the theory
588: %in Eqs. (\ref{SSQaPlusQb})- (\ref{SSQ2aSinc}), whereas the solid lines
589: %labeled C are the short range screening result, Eq. (\ref{Q2short}). \ (a)
590: %GG, beams $a$ and $b$ are Gaussians. \ Beam parameters are $a=1,b=100,K=0.01$.
591: %\ (b) SG, beam $a$ is a Sinc, beam $b$ is a Gaussian. \ Beam parameters are
592: %the same as in (a). \ \ (c) DG, beam $a$ is a Disk, beam $b$ is a Gaussian.
593: %\ Beam parameters are $a=1,b=10,K=0.04$. \ In all three examples displayed
594: %here the $b$ field was chosen to be a Gaussian in order to emphasize that the
595: %agreement with the theory represented by line B does not depend on the nature
596: %of the $a$ beam. \ Other short range choices for the $b$ beam, Disk or Sinc,
597: %show equally good agreement.}}{\Qlb{Fig3}}{sssscreening_fig3.tif}%
598: %{\special{ language "Scientific Word"; type "GRAPHIC";
599: %maintain-aspect-ratio TRUE; display "USEDEF"; valid_file "F";
600: %width 2.4102in; height 4.9986in; depth 0pt; original-width 3.9712in;
601: %original-height 8.2849in; cropleft "0"; croptop "1"; cropright "1";
602: %cropbottom "0";
603: %filename '../../../Current/SSSScreening/SSSScreening ScientificWord/SSSScreening Figs/SSSScreening_Fig3.tif';file-properties "XNPEU";}%
604: %}}%
605: %BeginExpansion
606: \begin{figure}
607: [pt]
608: \begin{center}
609: \includegraphics[
610: natheight=8.284900in,
611: natwidth=3.971200in,
612: height=4.9986in,
613: width=2.4102in
614: ]%
615: {SSSScreening_Fig3.pdf}%
616: \caption{$\left\langle Q^{2}\right\rangle /R$ vs. $R$ for composite sources in
617: which both the $a$ and $b$ beams have short range correlations. \ The thick
618: curves are the exact result in Eq. (\ref{Q2David}). \ The solid lines labeled
619: B are the theory in Eqs. (\ref{SSQ2b})- (\ref{SSQ2bS}), the dashed lines
620: labeled A are the theory in Eqs. (\ref{SSQaPlusQb})- (\ref{SSQ2aSinc}),
621: whereas the solid lines labeled C are the short range screening result, Eq.
622: (\ref{Q2short}). \ (a) GG, beams $a$ and $b$ are Gaussians. \ Beam parameters
623: are $a=1,b=100,K=0.01$. \ (b) SG, beam $a$ is a Sinc, beam $b$ is a Gaussian.
624: \ Beam parameters are the same as in (a). \ \ (c) DG, beam $a$ is a Disk, beam
625: $b$ is a Gaussian. \ Beam parameters are $a=1,b=10,K=0.04$. \ In all three
626: examples displayed here the $b$ field was chosen to be a Gaussian in order to
627: emphasize that the agreement with the theory represented by line B does not
628: depend on the nature of the $a$ beam. \ Other short range choices for the $b$
629: beam, Disk or Sinc, show equally good agreement.}%
630: \label{Fig3}%
631: \end{center}
632: \end{figure}
633: %EndExpansion
634:
635:
636: We illustrate the above in Fig. \ref{Fig3} . \ As can be seen, Eqs.
637: (\ref{SSQ2bG})-(\ref{SSQ2bS}) are good approximations to the exact results,
638: whereas Eqs. (\ref{SSQaPlusQb})-(\ref{SSQ2aSinc}) are good only to order of
639: $10\%$.
640:
641: We note that the well defined steps in this figure provide striking visual
642: confirmation of the fact that there are two widely different length scales.
643: \ The first step starts, as expected, at $R\sim1/b$, the second at $R\sim1/a$.
644:
645: \section{BEAM $a$ EXHIBITS LONG-RANGE SCREENING, BEAM $b$ EXHIBITS SHORT-RANGE
646: SCREENING}
647:
648: As discussed in the previous section, in region II, $1/b<R<1/a$, $\left\langle
649: Q^{2}\right\rangle $ is dominated by $\left\langle Q^{2}\right\rangle _{b}$,
650: Eqs. (\ref{SSQ2bG})- (\ref{SSQ2bS}). \ In region III, where $R>1/a$, beam $a$
651: exhibits long-range screening, and $\left\langle Q^{2}\right\rangle _{a}$ is
652: given by Eq. (\ref{LongRangeQ2}) with $p=a$. \ Neglecting again cross
653: screening of $a$ and $b$ charges, the total charge variance in region III is
654: approximated by the sum of $a$ and $b$ beam contributions, Eq.
655: (\ref{SSQaPlusQb}).
656:
657: In Fig. \ref{Fig4} we plot $\left\langle Q^{2}\right\rangle /R$ vs. $R$ for RG
658: and DG, again obtaining in the region of short range screening, region B, a
659: plateau that is in good agreement with the calculated value for $\left\langle
660: Q^{2}\right\rangle _{b}$. \ As expected, in region A, the region of long range
661: screening, $\left\langle Q^{2}\right\rangle /R$ grows linearly with $\ln R$.
662: \ We note that here not only is the calculated slope, Eq. (\ref{LongRangeQ2}),
663: in close agreement with the exact result (thick curves), but also that Eq.
664: (\ref{SSQaPlusQb}) provides a rather reasonable description of the data.
665: \ Similar good agreement is obtained for RS (not shown).%
666: %TCIMACRO{\FRAME{fbpFU}{2.4189in}{3.4394in}{0pt}{\Qcb{$\left\langle
667: %Q^{2}\right\rangle /R$ vs. $R$ for composite sources in which the $a$ beam has
668: %long-range and the $b$ beam short-range correlations. \ The thick curves are
669: %the exact result in Eq. (\ref{Q2David}). \ The solid lines labeled B are the
670: %short-range screening theory in Eqs. (\ref{SSQ2b})- (\ref{SSQ2bD}), the dashed
671: %lines labeled A are the long-range screening result in Eq. (\ref{LongRangeQ2}%
672: %). \ (a) RG, beam $a$ is a Ring, $b$ a Gaussian. \ (b) RD, beam $a$ is a Ring,
673: %beam $b$ a Disk. \ Beam parameters in both (a) and (b) are
674: %$\ a=1,b=100,K=0.01$.}}{\Qlb{Fig4}}{sssscreening_fig4.tif}%
675: %{\special{ language "Scientific Word"; type "GRAPHIC";
676: %maintain-aspect-ratio TRUE; display "USEDEF"; valid_file "F";
677: %width 2.4189in; height 3.4394in; depth 0pt; original-width 3.9868in;
678: %original-height 5.6861in; cropleft "0"; croptop "1"; cropright "1";
679: %cropbottom "0";
680: %filename '../../../Current/SSSScreening/SSSScreening ScientificWord/SSSScreening Figs/SSSScreening_Fig4.tif';file-properties "XNPEU";}%
681: %}}%
682: %BeginExpansion
683: \begin{figure}
684: [pb]
685: \begin{center}
686: \includegraphics[
687: natheight=5.686100in,
688: natwidth=3.986800in,
689: height=3.4394in,
690: width=2.4189in
691: ]%
692: {SSSScreening_Fig4.pdf}%
693: \caption{$\left\langle Q^{2}\right\rangle /R$ vs. $R$ for composite sources in
694: which the $a$ beam has long-range and the $b$ beam short-range correlations.
695: \ The thick curves are the exact result in Eq. (\ref{Q2David}). \ The solid
696: lines labeled B are the short-range screening theory in Eqs. (\ref{SSQ2b})-
697: (\ref{SSQ2bD}), the dashed lines labeled A are the long-range screening
698: result in Eq. (\ref{LongRangeQ2}). \ (a) RG, beam $a$ is a Ring, $b$ a
699: Gaussian. \ (b) RD, beam $a$ is a Ring, beam $b$ a Disk. \ Beam parameters in
700: both (a) and (b) are $\ a=1,b=100,K=0.01$.}%
701: \label{Fig4}%
702: \end{center}
703: \end{figure}
704: %EndExpansion
705:
706:
707: \section{BEAM $a$ EXHIBITS SHORT-RANGE SCREENING, BEAM $b$ EXHIBITS LONG-RANGE
708: SCREENING}
709:
710: In region II, $1/b<R<1/a$, we again set $a=0$ in $W^{\left( \text{T}%
711: _{a}\text{T}_{b}\right) }$, and obtain for beam $b$ a Ring,%
712: \begin{equation}
713: \left\langle Q^{2}\right\rangle _{b}\approx\frac{Kb^{2}}{4\pi}\int_{0}%
714: ^{2R}\sqrt{4R^{2}-r^{2}}\frac{J_{1}^{2}\left( br\right) }{1-J_{0}\left(
715: br\right) }dr. \label{Q2bLR}%
716: \end{equation}
717: As before, we have assumed $K\ll1$, $\left[ KW_b^{\left( \text{T}\right)
718: }\right] ^{2}\ll W_b^{\left( \text{T}\right) }$. \ Proceeding as in
719: [$19$, (Eqs. 40-48)], we obtain
720: \begin{subequations}
721: \label{myQ2b}%
722: \begin{align}
723: \left\langle Q^{2}\right\rangle _{b} & \approx\dfrac{KbR}{2\pi^{2}}\left[
724: {\cal{F}}+\ln(bR) \right] ,\label{myQ2bLine1}\\
725: {\cal{F}} & =\pi {\cal{I}}+\gamma+5\ln2-3\approx4.84258, \label{myQ2Line2}%
726: \\
727: {\cal{I}} & =\int_{0}^{\infty}\frac{J_{0}\left( x\right) J_{1}^{2}\left(
728: x\right) }{1-J_{0}\left( x\right) }dx\approx1.20946. \label{myQ2bLine3}%
729: \end{align}
730:
731:
732: We illustrate the above in Fig. \ref{Fig5}. \ In region B, $10^{-2}<R<10^{-1}%
733: $, for small $K$ ($0.015$) Eqs. (\ref{myQ2b}) provide a good description of
734: the exact result, Fig. \ref{Fig5}(a), whereas when $K$ is no longer small
735: ($0.2$) the expected deviations appear, Fig. \ref{Fig5}(b).%
736: %TCIMACRO{\FRAME{fbpFU}{2.3938in}{3.4973in}{0pt}{\Qcb{$\left\langle
737: %Q^{2}\right\rangle /R$ vs. $R$ for composite sources in which beam $a$\ is a
738: %Gaussian with short-range correlations, and beam $b$ a Ring with long-range
739: %correlations. \ Beam parameters are $\ a=1,b=100$. \ The thick curves are the
740: %exact result in Eq. (\ref{Q2David}). \ The solid lines labeled B are the
741: %long-range screening theory in Eqs. (\ref{myQ2b}). \ (a) $K=0.015$. \ (b)
742: %$K=0.2.$}}{\Qlb{Fig5}}{sssscreening_fig5.tif}%
743: %{\special{ language "Scientific Word"; type "GRAPHIC";
744: %maintain-aspect-ratio TRUE; display "USEDEF"; valid_file "F";
745: %width 2.3938in; height 3.4973in; depth 0pt; original-width 3.9444in;
746: %original-height 5.7839in; cropleft "0"; croptop "1"; cropright "1";
747: %cropbottom "0";
748: %filename '../../../Current/SSSScreening/SSSScreening ScientificWord/SSSScreening Figs/SSSScreening_Fig5.tif';file-properties "XNPEU";}%
749: %}}%
750: %BeginExpansion
751: \begin{figure}
752: [pb]
753: \begin{center}
754: \includegraphics[
755: natheight=5.783900in,
756: natwidth=3.944400in,
757: height=3.4973in,
758: width=2.3938in
759: ]%
760: {SSSScreening_Fig5.pdf}%
761: \caption{$\left\langle Q^{2}\right\rangle /R$ vs. $R$ for composite sources in
762: which beam $a$\ is a Gaussian with short-range correlations, and beam $b$ a
763: Ring with long-range correlations. \ Beam parameters are $\ a=1,b=100$. \ The
764: thick curves are the exact result in Eq. (\ref{Q2David}). \ The solid lines
765: labeled B are the long-range screening theory in Eqs. (\ref{myQ2b}). \ (a)
766: $K=0.015$. \ (b) $K=0.2.$}%
767: \label{Fig5}%
768: \end{center}
769: \end{figure}
770: %EndExpansion
771:
772:
773: In Fig. \ref{Fig5}(a) a plateau appears for $R>1$, apparently consistent with
774: the fact that short-range screening $a$ field charges become important in this
775: region. \ But this apparent plateau is misleading, because the long-range
776: screening of the $b$ field charges can never saturate: regardless of the
777: nature of the $a$ field, $\left\langle Q^{2}\right\rangle /R$ for the $b$
778: charges must diverge logarithmically for large $R$. \ The rate (slope) of this
779: logarithmic divergence, however, is $K$ dependent, being small, Fig.
780: \ref{Fig5}(a) (large, Fig. \ref{Fig5}(b)) for small (large) $K$.
781:
782: \section{BOTH BEAM $a$ AND BEAM $b$ EXHIBIT LONG-RANGE SCREENING}
783:
784: The case of two Rings is illustrated in Fig. \ref{Fig6}. \ As before, Eq.
785: (\ref{myQ2b}) holds for the $b$ field charges. \ But now the $a$ field charges
786: also exhibit long-range screening, and $\left\langle Q^{2}\right\rangle /R$
787: exhibits the expected large $R$ logarithmic divergence also for small $K$
788: ($0.015$).%
789:
790: %TCIMACRO{\FRAME{fhFU}{2.3843in}{1.7495in}{0pt}{\Qcb{$\left\langle
791: %Q^{2}\right\rangle /R$ vs. $R$ for a composite source in which both beam
792: %$a$\ and beam $b$ are Rings. \ Beam parameters are $\ a=1,b=100$, $K=0.015$.
793: %\ The thick curve is the exact result in Eq. (\ref{Q2David}). \ The solid line
794: %labeled B is the long-range screening theory in Eqs. (\ref{myQ2b}).}%
795: %}{\Qlb{Fig6}}{sssscreening_fig6.tif}{\special{ language "Scientific Word";
796: %type "GRAPHIC"; maintain-aspect-ratio TRUE; display "USEDEF";
797: %valid_file "F"; width 2.3843in; height 1.7495in; depth 0pt;
798: %original-width 3.928in; original-height 2.8677in; cropleft "0";
799: %croptop "1"; cropright "1"; cropbottom "0";
800: %filename '../../../Current/SSSScreening/SSSScreening ScientificWord/SSSScreening Figs/SSSScreening_Fig6.tif';file-properties "XNPEU";}%
801: %}}%
802: %BeginExpansion
803: \begin{figure}
804: [h]
805: \begin{center}
806: \includegraphics[
807: natheight=2.867700in,
808: natwidth=3.928000in,
809: height=1.7495in,
810: width=2.3843in
811: ]%
812: {SSSScreening_Fig6.pdf}%
813: \caption{$\left\langle Q^{2}\right\rangle /R$ vs. $R$ for a composite source
814: in which both beam $a$\ and beam $b$ are Rings. \ Beam parameters are
815: $\ a=1,b=100$, $K=0.015$. \ The thick curve is the exact result in Eq.
816: (\ref{Q2David}). \ The solid line labeled B is the long-range screening
817: theory in Eqs. (\ref{myQ2b}).}%
818: \label{Fig6}%
819: \end{center}
820: \end{figure}
821: %EndExpansion
822:
823:
824: \section{SUMMARY}
825:
826: Screening in speckled speckle produced by composite sources has been studied
827: in the limit of two widely different characteristic length scales in the
828: speckle pattern and two widely different intensities in the sources. \ The
829: interesting combination, emphasized here, is an intense field with a long
830: length scale (the $a$ field) perturbed by a weak field that has a short length
831: scale (the $b$ field).
832:
833: Exact, and approximate, results have been presented for all four combinations
834: of short- and long-range screening. \ When the correlations in both fields are
835: short-range, cross screening between $a$ and $b$ field singularities appears
836: to be of only secondary importance, and the results of the exact calculation
837: are found to decompose, approximately, into a sum of screening contributions,
838: one for each field, Fig. \ref{Fig3}. \ A similar decomposition is found to
839: hold if the $b$ field exhibits short-range screening and the $a$ field
840: exhibits long-range screening, Fig. \ref{Fig4}. \ If the $b$ field exhibits
841: long-range screening, however, then no decomposition is valid, Figs.
842: \ref{Fig5} and \ref{Fig6}.
843:
844: The unusual screening properties of speckled speckle are most pronounced when
845: the ratio ($b$ field to $a$ field) of length scales is $100:1$ or greater, and
846: the ratio of optical powers in the sources (the parameter $K$) is $1:100$ or
847: less. \ Less extreme parameter ratios can also yield useful results, however,
848: as in Fig. \ref{Fig3}(b), so that the experimental study of screening in speckled
849: speckle appears to be entirely feasible.
850: \end{subequations}
851: \begin{acknowledgments}
852: D. A. Kessler acknowledges the support of the Israel Science Foundation.
853: \end{acknowledgments}
854:
855: \begin{center}
856: {\large \textbf{References}}
857: \end{center}
858:
859: \hspace{-0.15in}[1] J. F. Nye and M. V. Berry, \textquotedblleft Dislocations
860: in wave trains,\textquotedblright\ Proc. Roy. Soc. London Ser. A \textbf{336,}
861: 2165 (1974). For additional sources see online citation databases for the
862: numerous papers that reference this work.
863:
864: \hspace{-0.15in}[2] J. F. Nye, \emph{Natural Focusing and the Fine Structure
865: of Light} (IOP Publ., London, 1999).
866:
867: \hspace{-0.15in}[3] J. W. Goodman, \emph{Speckle Phenomena In Optics} (Roberts
868: \& Co., Englewood, Colorado, 2007).
869:
870: \hspace{-0.15in}[4] B. I. Halperin, \textquotedblleft Statistical mechanics of
871: topological defects,\textquotedblright\ in \emph{Physics of Defects}, R.
872: Balian, M. Kleman, and J.-P. Poirier, eds. (North-Holland, Amsterdam, 1981),
873: p. 814-857.
874:
875: \hspace{-0.15in}[5] F. Liu and G. F. Mazenko, \textquotedblleft Defect-defect
876: correlation in the dynamics of first-order phase
877: transitions,\textquotedblright\ Phys. Rev. B \textbf{46,} 5963-5971 (1992).
878:
879: \hspace{-0.15in}[6] B. W. Roberts, E, Bodenschatz, and J. P. Sethna,
880: \textquotedblleft A bound on the decay of defect-defect correlation functions
881: in two- dimensional complex order parameter equations,\textquotedblright%
882: \ Physica D \textbf{99,} 252-268 (1996).
883:
884: \hspace{-0.15in}[7] I. Freund and M. Wilkinson, \textquotedblleft
885: Critical-point screening in random wave fields,\textquotedblright\ J. Opt.
886: Soc. Am. A \textbf{15,} 2892-2902 (1998).
887:
888: \hspace{-0.15in}[8] M. V. Berry and M. R. Dennis, \textquotedblleft Phase
889: singularities in isotropic random waves,\textquotedblright\ Proc. Roy. Soc.
890: London A \textbf{456,} 2059-2079 (2000); ibid., p. 3048.
891:
892: \hspace{-0.15in}[9] M. V. Berry and M. R. Dennis, \textquotedblleft
893: Polarization singularities in isotropic random vector waves,\textquotedblright%
894: \ Proc. Roy. Soc. Lond. A \textbf{457,} 141-155 (2001).
895:
896: \hspace{-0.15in}[10] M. R. Dennis, \textquotedblleft Polarization
897: singularities in paraxial vector fields: morphology and
898: statistics,\textquotedblright\ Opt. Commun. \textbf{213,} 201-221 (2002).
899:
900: \hspace{-0.15in}[11] I. Freund, M. S. Soskin, and A. I. Mokhun,
901: \textquotedblleft Elliptic critical points in paraxial
902: fields,\textquotedblright\ Opt. Commun. \textbf{208,} 223-253 (2002).
903:
904: \hspace{-0.15in}[12] G. Foltin, \textquotedblleft Signed zeros of Gaussian
905: vector fields - density, correlation functions, and
906: curvature,\textquotedblright\ J. Phys. A: Math. Gen. \textbf{36, }1729-1742 (2003).
907:
908: \hspace{-0.15in}[13] M. R. Dennis, \textquotedblleft Correlations and
909: screening of topological charges in Gaussian random fields,\textquotedblright%
910: \ J. Phys. A: Math. Gen. \textbf{36,} 6611-6628 (2003).
911:
912: \hspace{-0.15in}[14] M. Wilkinson, \textquotedblleft Screening of charged
913: singularities of random fields,\textquotedblright\ J. Phys. A: Math. Gen.
914: \textbf{37,} 6763-6771 (2004).
915:
916: \hspace{-0.15in}[15] G. Foltin, S. Gnutzmann, and U. Smilansky,
917: \textquotedblleft The morphology of nodal lines - random waves versus
918: percolation,\textquotedblright\ J. Phys. A: Math. Gen. \textbf{37,}
919: 11363-11372 (2004).
920:
921: \hspace{-0.15in}[16] B. A. van Tiggelen, D. Anache, and A. Ghysels,
922: \textquotedblleft Role of mean free path in spatial phase correlation and
923: nodal screening,\textquotedblright\ Europhys. Lett. \textbf{74,} 999-1005 (2006).
924:
925: \hspace{-0.15in}[17] I. Freund, R. I. Egorov, and M. S. Soskin,
926: \textquotedblleft Umbilic point screening in random optical
927: fields,\textquotedblright\ Opt. Lett. \textbf{22,} 2182-2184 (2007).
928:
929: \hspace{-0.15in}[18] R. I. Egorov, M. S. Soskin, D. A. Kessler, and I. Freund,
930: \textquotedblleft Experimental measurements of topological singularity
931: screening in random paraxial scalar and vector optical
932: fields,\textquotedblright\ Phys. Rev. Lett. \textbf{100,} 103901 (2008).
933:
934: \hspace{-0.15in}[19] D. A. Kessler and I. Freund, \textquotedblleft Short- and
935: long-range screening of optical phase singularities and C
936: points,\textquotedblright\ Opt. Commun. (2008), doi:10:1016/j.optcom.2008.05.018.
937:
938: \hspace{-0.15in}[20] I. Freund and D. A. Kessler, \textquotedblleft
939: Singularities in speckled speckle,\textquotedblright\ Opt. Lett. \textbf{33,}
940: 479-481 (2008).
941:
942: \hspace{-0.15in}[21] I. Freund and D. A. Kessler, \textquotedblleft
943: Singularities in speckled speckle: Statistics,\textquotedblright%
944: \ arXive:0806365; Opt. Commun. (submitted).
945:
946: \hspace{-0.15in}[22] J. W. Goodman, \emph{Statistical Optics} (John Wiley, New
947: York, 1985).
948:
949: \hspace{-0.15in}[23] M. Berry, \textquotedblleft Disruption of wave-fronts:
950: statistics of dislocations in incoherent Gaussian random
951: waves,\textquotedblright\ J. Phys. \textbf{A 11,} 27-37 (1978).
952:
953:
954: \end{document}