Angular-acceptance simulations for three azimuthal angles with displaced DOM

[fiedl]

Summer has requested the following set of simulations:

Create angular-acceptance plots for incoming plane waves from three different azimuthal directions. The bubble column with given radius and given geometric scattering length needs to be displaced relative to the DOM ("offset").

offset [cm] radius [cm] scattering [cm]
15  7.5 10  
15  15  10
15  15  70
15  30  70

• Implement wrapper script for these simulations
• Run simulations on the cluster
• Run gauging simulation with optimal conditions
• Produce angular-acceptance plots
• Produce some videos with steamshovel
• Rerun cluster with more photons

https://www.youtube.com/watch?v=M_QGNdGG9Ew

[fiedl]

[fiedl]

What are the polar angles?

The string spacing is delta_x = 20.0 meters.
The vertical DOM spacing is 10.0 meters.
The string length is at least 300 meters.
Also, the DOM above and the DOM below may emit light towards the central DOM.

import numpy as np
delta_x = 20.0
delta_z_range = np.arange(-150, 151, 10)
incoming_polar_angles = [0] + [90 + np.arctan(1.0 * delta_z / delta_x) / np.pi * 180 for delta_z in delta_z_range] + [180]

[fiedl]

What is perfect bulk ice?

• bulk ice without scattering and absorption? This does not work in simulations, because the photons would escape into infinity, which causes the simulation not to finish.
• homogeneous ice

Or, not use perfect bulk ice, but regular bulk ice.

[fiedl]

Regular bulk ice:

Starting photons from 20m horizontal distance at z=500 relative to the IceCube center, most photons are scattered away before reaching the receiving DOM.

The "ice paper", 2013, "Measurement of South Pole ice transparency with the IceCube LED calibration system", shows the depth dependence of scattering and absorption:

The two translate into each other that the ice surface is at z = 1948.07 m (https://wiki.icecube.wisc.edu/index.php/Coordinate_system). z=500m equals a depth=1448.07.

This is the very top of the detector. The effective scattering length should be about 14m there according to the ice-paper figure. This corresponds to a geometric scattering length of 84cm, which explains the observed behaviour.

At the center of the detector, it's even worse. One would have to be lucky to sit in a more transparent layer. Thus, just keep this ice and start more photons, or start them from a closer distance to the receiving DOM.

[fiedl]

Submitting a test job to the cluster:

[2018-10-18 19:17:13] fiedl@kepler00 /afs/ifh.de/group/amanda/scratch/fiedl/hole-ice-study
▶ git pull

[2018-10-18 19:19:08] fiedl@kepler00 /afs/ifh.de/group/amanda/scratch/fiedl/hole-ice-study/scripts/AngularAcceptanceFromSeveralAzimuthalAngles
▶ qsub -l gpu -l tmpdir_size=10G -l s_rt=4:0:00 -l h_rss=2G -m ae -t 1-1 batch-job.sh

In this configuration, the simulation takes about 2mins. But this is still for only two polar angles.

[fiedl]

Submit array job to cluster:

[2018-10-18 19:38:33] fiedl@kepler00 /afs/ifh.de/group/amanda/scratch/fiedl/hole-ice-study/scripts/AngularAcceptanceFromSeveralAzimuthalAngles
▶ git pull
▶ qsub -l gpu -l tmpdir_size=10G -l s_rt=4:0:00 -l h_rss=2G -m ae -t 1-12 batch-job.sh

[fiedl]

Gauging simulation: Count the number of photons arriving for the same starting distance and plane extend when photons are coming from below and no hole ice is present.

[2018-10-18 19:48:10] fiedl@kepler00 /afs/ifh.de/group/amanda/scratch/fiedl/hole-ice-study/scripts/FiringRange
▶ $ICESIM/env-shell.sh
▶ rm -r tmp
▶ ./run.rb --distance=3.0 --number-of-photons=1e7 --number-of-runs=1 --number-of-parallel-runs=1 --angle=0 --plane-wave --no-hole-ice