CO2 Concentrations and the Freakish Heatwaves at the Poles: A Preliminary Analysis

In March of 2022, AP News and other news outlets reported that the Earth's poles were experiencing simultaneous heatwaves.  Portions of the Arctic were more than 30 degrees Celsius warmer than expected, while some locations in Antarctica were  40 degrees Celsius warmer than average.  While some suspect this freakish outcome results from human-induced climate change, others have scoffed at this suggestion.  With this attribution uncertainty in mind, this paper seeks to understand the drivers in hourly temperatures at the Ny-Ålesund station (NYA) in Svalbard (Latitude: 78.9227,  Longitude: 11.9273)  and the Neumayer Station (GVN) in Antarctica (Latitude: -70.6500, Longitude: -8.2500).

 Based on one-minute data from the Baseline Surface Radiation Network (BSRN), the hourly temperature data for NYA and GVN from January 1, 1999, through December 31, 2019, were calculated.  Hourly averages of the following radiation variables were also calculated: Short-Wave Downward (SWD), Long-Wave Downward(LWD),  Short-Wave Upward (SWU), and  Long-Wave Upward (LWU).   The hourly net radiation flux at the Earth's surface was then calculated as SWD + LWD – SWU – LWU.  This variable is of interest because it is recognized as an important driver of the weather and climate system.  The analysis also uses the hourly CO₂ concentration data for Svalbard reported by the  Integrated Carbon Observation System (ICOS) from January 1, 2004 through December 2019.

The analysis proceeds by employing the Vector Autoregressive Regression (VAR) method.  The general approach considers K variables specified as linear functions of p of their lags and p lags of the other K - 1 variables.  Using this methodology, one can subsequently test for Granger Causality.  The methodology is based on whether the lagged values of some variable X are useful in predicting the current value of some variable Y. Because of its focus on the lagged values, the methodology does not contest the truism that the correlation between two contemporaneous variables does not imply causation.

The VAR/Granger methodology is first applied here to model the possible relationship between hourly CO₂ concentrations and the hourly net radiation flux levels at  NYA.      In this case,  there is strong statistical evidence that hourly CO₂ concentrations at NYA have Granger Causal implications for the hourly net radiation flux at NYA.  Consistent with this finding, the out-of-sample hourly net radiation flux predictions for NYA are more accurate than a persistence forecast when the lagged CO₂ concentrations are included in the analysis.   

The following evidence is also presented: the hourly net radiation flux at NYA has Granger Causal implications for the hourly temperature at NYA, the hourly net radiation flux at NYA in the Polar region has  Granger Causal implications for the net radiation flux at the GVN station in Antarctica, and the hourly temperatures at NYA and GVN exhibit two-way Granger Causality.  In short, the analysis in the paper supports the view that the atmospheric and meteorological conditions at any location are highly interrelated with conditions elsewhere and that the pace of freakish weather conditions is likely to increase as CO₂ concentrations continue to rise.