next up previous
Next: 6 Discussion Up: HeuristicGravFluctuations Previous: 4 Correlation functions


5 Interferometric detection of the fluctuations.

The fluctuations introduced in the previous sections may - at least in principle - be detectable in interferometers built to search for gravitational waves [11,12] or higher order QED effects [13]. To estimate how much the fluctuations influence the propagation of light we proceed as follows: if we take a photon with frequency $\nu _{0}$ and wavelength $\lambda _{0}$ in a region where the gravitational potential vanishes, then its wavelength in a region with a (negative) gravitational potential $\Phi $ is approximately

\begin{displaymath}
\lambda ({\vec{x}},t)\approx \lambda _{0}\left( {1+{\frac{{\Phi ({\vec{x}},t)}%
}{{c^{2}}}}}\right) .
\end{displaymath}

while the frequency does not change (see [14,15]). The gravitationally loaded vacuum behaves as a refractive medium with refractive index

\begin{displaymath}
n({\vec{x}},t)\approx \left( {1-{\frac{{\Phi ({\vec{x}},t)}}{{c^{2}}}}}\right)
\end{displaymath}

the optical path length changes, and the resulting phase change along a path ${\mathfrak{L}}$ is
$\displaystyle \varphi _{L}(t)$ $\textstyle \approx$ $\displaystyle {\frac{2\pi }{\lambda _{0}}}\int\limits_{{%
\mathfrak{L}}}n{\left[ {{\vec{x}}(t+\ell /c-L/c),t+\ell /c-L/c}\right] \mathrm{d} %
\ell }$  
  $\textstyle \approx$ $\displaystyle {\frac{2\pi }{\lambda _{0}}}\int\limits_{{\mathfrak{L}}}\left( {1...
...\vec{x}}(t+\ell /c-L/c),t+\ell /c-L/c)}}{{c^{2}}}}}\right) {%
\mathrm{d} \ell }$ (9)

where $\ell$ is the path length, $L$ is the total path length and 9 is the phase change at the end of the path at time $t$ (notice that the wave enters the path at an earlier time $t-L/c$). Thus in a two-arm interferometer with distinct paths ${\mathfrak{L}}_{1}$ and ${\mathfrak{L}}_{2}$ and a zero-field phase difference $\phi_{0}$ between the arms, the phase difference due quantum fluctuations is
$\displaystyle \Delta \varphi (t)$ $\textstyle =$ $\displaystyle (\varphi _{2}(t)-\varphi _{1}(t))-\varphi _{0}$  
  $\textstyle =$ $\displaystyle {\frac{{2 \pi}}{\lambda _{0}}}\left[\,\int\limits_{{\mathfrak{L}}...
...l /c-L/c),t+\ell /c-L/c}\right]
}}{{c^{2}}}}}\right)
\mathrm{d} \ell }
\right .$  
    $\displaystyle \qquad
\left .
-\int\limits_{{\mathfrak{L}}_{1}}{\left( {1-{\frac...
...+\ell /c-L/c}\right] }}{{c^{2}}}}}\right)
\mathrm{d} \ell }\right]-\varphi _{0}$  
  $\textstyle =$ $\displaystyle {\frac{{2 \pi}}{{\lambda _{0} c^{2}}}}\left\{ {\int\limits_{{\mat...
...}\right) ,t+{\frac{
\ell }{c}}-{\frac{L}{c}}}\right] \mathrm{d} \ell }}
\right.$  
    $\displaystyle \qquad
\left.
-\int\limits_{{\mathfrak{L}}_{1}}{\Phi
\left[ {{\ve...
...\right) ,t+{\frac{
\ell }{c}}-{\frac{L}{c}}}\right] \mathrm{d} \ell }\right\} ;$ (10)

then we find the total irradiance at the end of the two paths:
$\displaystyle I(t)$ $\textstyle =$ $\displaystyle I_{1}+I_{2}+2\sqrt{I_{1}I_{2}}\cos \left[ {\varphi _{0}+\Delta \varphi
(t)}\right]$  
  $\textstyle \approx$ $\displaystyle I_{1}+I_{2}+2\sqrt{I_{1}I_{2}}\cos \varphi _{0}-2\Delta
\varphi (t)\sqrt{I_{1}I_{2}}\sin \varphi _{0}.$  

Therefore the variance of the irradiance noise is (we set $\int \int _{1,2} \mathrm{d} \ell _{1} \mathrm{d} \ell _{2}
\equiv
\int \int _{...
...mathfrak{L}}_{2},{\mathfrak{L}}_{2}} \mathrm{d} \ell _{1} \mathrm{d} \ell _{2}
$)
$\displaystyle \left\langle \left\vert I\right\vert ^{2}\right\rangle$ $\textstyle =$ $\displaystyle 4I_{1}I_{2}\sin ^{2}\varphi
_{0}\left\langle {\ \left\vert \Delta \varphi \right\vert ^{2}}\right\rangle$  
  $\textstyle =$ $\displaystyle {\frac{
{16 \pi ^{2} I_{1}I_{2} \sin ^{2}\varphi _{0}}}{{\lambda ...
...c{{\ell _{2}}}{c}}}\right) ,
{\frac{{
\ell _{2}}}{c}}}\right] }\right\rangle }.$ (11)

The variance 11 does not depend on $t$ because of the stationarity of the phase noise 10, so that we can set $t=L/c$ and simplify the notation; likewise the correlation function of the irradiance is given by

$\displaystyle R_{I}(\tau )$ $\textstyle =$ $\displaystyle \left\langle {\Delta I(t)\Delta I(t+\tau )}\right\rangle$  
  $\textstyle =$ $\displaystyle 4I_{1}I_{2}\sin ^{2}\varphi _{0}\left\langle {\Delta \varphi (t)\Delta
\varphi (t+\tau )}\right\rangle$  
  $\textstyle =$ $\displaystyle {\frac{{16 \pi ^{2} I_{1}I_{2} \sin ^{2} \varphi _{0}}}{{\lambda ...
...}}}{c}}+\tau }
\right) ,{\frac{{\ell _{2}}}{c}}+\tau }\right] }\right\rangle },$  

and, finally, the spectral density of the irradiance fluctuations, from the Wiener-Kintchine theorem, is
$\displaystyle \left\langle {\left\vert \Delta I(f)\right\vert ^{2}}\right\rangle$ $\textstyle =$ $\displaystyle \int\limits_{-\infty }^{+\infty }{R_{I}(\tau )\mathrm{e} ^{-2\pi \mathrm{i} f \tau }\mathrm{d}
\tau}$  
  $\textstyle =$ $\displaystyle 2\int_{0}^{\infty }{R_{I}(\tau )\cos (2\pi f\tau )\mathrm{d} \tau }$  
  $\textstyle =$ $\displaystyle {\frac{{16 \pi ^{2} I_{1}I_{2}\sin ^{2}\varphi _{0}}}{{\lambda _{...
...ht),{\frac{{\ell _{2}}}{c}}+\tau }\right] }%
\right\rangle
\cos (2\pi f\tau ) .$  


next up previous
Next: 6 Discussion Up: HeuristicGravFluctuations Previous: 4 Correlation functions

Stefano Ansoldi