IceCube’s muons could track Antarctic atmospheric change—if detector drift is controlled

IceCube could potentially provide an independent check on long-term temperature changes above the South Pole, according to a study by Tanvi Krishnan, Gabriela Barenboim and Carlos A. Argüelles. Their central finding is also a warning: a very small, unnoticed change in the detector’s counting rate could look like atmospheric cooling or warming.

The arXiv preprint is a feasibility study using simulated muon rates. It does not report a temperature trend measured with IceCube. The supplied source does not establish peer-reviewed publication.

How particles become atmospheric probes

IceCube’s optical sensors sit deep in Antarctic ice. Alongside neutrino events, the observatory records atmospheric muons produced when cosmic rays strike the atmosphere.

Temperature affects how many of these muons reach the detector. In a warmer, more expanded atmosphere, particles called pions and kaons have a greater chance of decaying into muons before colliding again. Consequently, atmospheric temperature changes alter the muon rate.

The relevant quantity is an effective atmospheric temperature: a weighted average through the atmosphere, not a thermometer reading at one altitude. In the researchers’ calculation, about 94% of the weighting comes from pressures below 300 hectopascals, corresponding to altitudes above roughly eight kilometres. The measurement is therefore mainly sensitive to the stratosphere, but cannot resolve temperatures at individual heights.

Testing the seasonal connection

The team used monthly South Pole temperature profiles from the NCEP/NCAR atmospheric reanalysis, which combines observations with weather modelling. They checked the resulting effective temperatures against 217 months of local balloon profiles. The estimates tracked monthly variations closely, with a correlation of 0.9975 and a root-mean-square difference of 1.22 kelvin. That agreement supports the input data, but does not eliminate uncertainty in long-term trends.

Separately, the researchers calculated particle showers using MCEq software and twelve representative monthly atmospheres. Their nominal simulation produced a temperature-response coefficient of 0.745, close to IceCube’s published seasonal value of 0.750. The latter means that a 1% change in effective temperature corresponds to approximately a 0.75% change in muon rate.

However, the calculation assumes vertical muons and a single energy threshold rather than IceCube’s full detector response. Varying that threshold produced coefficients from 0.673 to 0.805, illustrating why the nominal agreement is not a precision validation.

Seasonal warming also has a different vertical pattern from the longer-term changes in the reanalysis. Testing those patterns suggested that the long-term response is about 3% stronger than the seasonal response—a correction relevant to a future precision measurement.

Stability matters more than counting

The team generated simulated monthly rates using reanalysis temperatures, the published seasonal coefficient and counting fluctuations corresponding to IceCube’s approximately 2,150 events per second. They fitted temperature, solar variability and detector drift, accounting for correlations between neighbouring months.

At that event rate, counting noise is tiny. Atmospheric variability limits statistical sensitivity, while detector drift creates a separate systematic ambiguity. An unmodelled rate change of 1% per decade could imitate a temperature trend of about 2.8 kelvin per decade.

To keep that false signal below the roughly 0.9-kelvin-per-decade cooling indicated by atmospheric records over 2011–2025, the rate would need stability of approximately 0.3% per decade. Longer observations cannot, by themselves, distinguish steady detector drift from a genuine atmospheric trend.

Snow accumulation is another complication. Increasing the ice overburden reduces the muon rate. Depending mainly on the assumed effect of ice flow, the study estimates apparent cooling equivalent to 0.24–0.51 kelvin per decade. This contribution must be accounted for rather than mistaken for atmospheric change.

What a longer record could establish

For a 14-year record, the estimated one-sided 95% statistical sensitivity is 1.6 kelvin per decade—insufficient to constrain the approximately 0.9-kelvin cooling trend at that confidence level. Under the study’s assumptions, roughly 22 years of total measurement would be needed.

Those estimates exclude IceCube’s actual non-counting rate variability, which the authors could not assess without its time series. Tests also could not rule out longer-lasting atmospheric correlations that would weaken sensitivity.

Finally, the inferred atmospheric trend depends strongly on the chosen time window and can change sign. The result is therefore a conditional route to checking a specified historical trend, not a forecast of future Antarctic temperatures. Applying it will require direct analysis of IceCube’s rate record and independent evidence that detector changes are sufficiently controlled.