Read IceCube’s discovery paper, one figure at a time

Figure 1: read the rule before the result

Start with the detector drawing in arXiv:1311.5238v2. Figure 1 shades the regions used to reject events whose first light appears at the boundary. The question is whether a particle entered from outside or an interaction began inside. The drawing explains a selection rule, rather than showing the discovered population.

Follow a hypothetical incoming muon with your finger. Ask which sensor region should light first and what would make the veto miss it. The dusty layer deserves attention: an optical blind spot can change which physical trajectories remain in the sample. Instrument geometry is part of the argument.

Then locate the two observing seasons in the caption. A detector can change during construction, so one picture need not describe every day equally. Before comparing counts with a prediction, both must use the configurations and useful observing time of the analyzed sample.

Sources: [1] · [2]

Figure 2: a point is a reconstructed event

Figure 2 places deposited energy against declination, the north–south coordinate on the sky. Crosses mark track-like events and filled circles mark showers. First read the legend and the uncertainty convention; equally sharp printed symbols do not imply equally precise directions.

First read the events’ directions without comparing their energies; then read their energies without comparing directions. This separates two questions that a vivid scatter plot can blur: where the reconstruction points and how much energy the detector recorded.

Visible energy is not automatically the incident neutrino energy. Escaping particles can carry energy away. Treat a large energy reading as a recorded deposit before interpreting it as a source property. A gap between points also needs a statistical test; a small sample naturally leaves empty intervals.

Sources: [1] · [3] · [4]

Figure 3: why plot the square of radius?

Figure 3 compares the first-light positions with the geometry of entering muons. Its radial coordinate is squared. This choice matters because equal-width radial rings contain different volumes: a ring farther from the cylinder axis contains more material.

For a uniform cylindrical target, equal intervals of squared radius contain equal transverse area. A flat distribution in that coordinate can therefore be compatible with uniform interaction density. That is a geometric check, not a special preference for one part of the detector.

Ask which regions have been excluded and whether the prediction has those same exclusions. The informative comparison is with plausible background geometry after selection. A visually even pattern by itself cannot assign a distant origin to every remaining event.

Sources: [1] · [2]

Figure 4: compare recorded counts with recorded-count predictions

Figure 4 compares the observed energy and direction distributions with atmospheric components and an added extraterrestrial component. Read the colored components separately, then inspect the total. A component curve is not an additional set of measured events.

Before deciding which curve fits, check the plotted quantity and the binning. A source flux must pass through interaction probability, acceptance and energy response to become expected detector counts. Comparing a flux curve directly with an event histogram would omit that conversion.

Finally read the significance calculation in the text. The treatment of the earlier two events and the atmospheric model matters. Subtracting one background number from 28, or attaching a Gaussian error bar to that difference, does not reproduce the published inference.

Sources: [1] · [4] · [3]

Figure 5: a source search asks another question

Figure 5 maps the strength of a point-source fit across the sky. Its most significant cluster had an 8% final chance probability after accounting for the sky search; the paper did not identify a neutrino source from it. The population evidence and source-clustering test therefore support different statements, even within one paper.

In the NGC 1068 paper, look for the tested source catalog, fitted signal contribution and treatment of searches across many positions. The fitted contribution is a statistical excess, not a list of events certified to have left that galaxy. Angular and energy information enter together.

Keep a short reading record: publication version, sample dates, selection, axis meanings, competing hypotheses and the sentence the evidence supports. This makes it possible to compare later papers without silently transferring one sample’s assumptions into another. The figure numbering above refers specifically to the 2013 v2 paper.

Sources: [5] · [4] · [1]

Primary sources & revision

1. IceCube Collaboration · 2013 discovery paper, arXiv v2 (16 December 2013)
2. IceCube Collaboration · Instrumentation and Online Systems (2017; arXiv v3)
3. Francis Halzen · Astroparticle Physics with High Energy Neutrinos: from AMANDA to IceCube (2006)
4. Particle Data Group · Statistics (2025)
5. IceCube Collaboration · Evidence for neutrino emission from NGC 1068 (2022; arXiv v2, 2024)

First published 2026-10-11; last revised 2026-10-11. Original explanatory text and diagrams by Sandbox Physics. Illustrations are schematic; no experimental event records are reproduced here. This is an independent educational publication, not an official Nobel or experiment collaboration publication.

Continue the story