Do not use refractive index of ice at LOFAR/Auger

[sjoerd-bouma]

The old NuRadioReco.utilities.ice ice 'model' is still used internally by some modules, including the voltageToEfieldConverter which is used also by LOFAR. This leads to unwanted behaviour for antennas with z-coordinates < 0.

In future, this module should probably be deprecated because it also doesn't do the right thing for deep in-ice antennas (or any ice where the assumption that n_ice=1.3 is wrong). But this would be slightly more work, depending on how we want to implement it (e.g. should air-to-ice raytracing be included).

[cg-laser]

I agree, we should probably deprecate this module. The fix for now looks good.

And yes, the efieldToVoltageConverter only works for local antennas. Two assumptions need to be fulfilled

• straight line propagation. i.e. constant index-of-refraction
• same efield in all antennas

so this only works for very compact antenna clusters or the dual-polarizided antennas used in air-shower arrays.