= IceCube ice-transparency calibration experiment
{c}
In the 2013 calibration study, six horizontal <light-emitting diodes> per module on string 63 flashed sequentially; other strings recorded charges and <photon> arrival times. A global fit inferred depth-dependent <scattering of light> and <absorption coefficients>. Dust-logger measurements checked layer structure; independent <muon> events tested simulated time patterns. <Photomultiplier> response and refrozen ice near modules remained limitations. Thus a late hit need not imply a late interaction: <scattering of light> changes the path.<Measurement of South Pole ice transparency with the IceCube LED calibration system by IceCube Collaboration>[Sections 3, 8–11; Figs. 14, 20–23: flashing, cross-checks and limitations]{ref}
An original limiting calculation isolates this issue. Suppose a compact source emits at time $t_0$ into homogeneous, nondispersive <ice> with stipulated <refractive index> $n=1.32$. At distance $r$, the unscattered light arrives at
$$
t_{\mathrm{direct}}=t_0+\frac{nr}{c},
$$
where $c$ is the vacuum <speed of light>. For chosen $r=100\,\mathrm m$, the delay is $440.3\,\mathrm{ns}$. A scattered path $30\,\mathrm m$ longer adds $132.1\,\mathrm{ns}$ even with perfect clocks. These chosen values are not calibration measurements. Real <ice> requires wavelength-dependent propagation, module response and a spatially varying light-transport model; replacing that response by one arrival time discards information.
Back to article page
