El Niño cuts December heating demand in the Midwest and Northeast, a preprint finds

Why it matters

Heating degree days (HDDs) are the basis of traded weather derivatives, yet the authors know of no earlier study that measured the ENSO–HDD link quantitatively. The monthly detail matters because seasonal averages can hide signals that occur in only one month. The authors also turn the signals into fair prices for HDD put options, so the findings can be read in market terms.

What they did

The authors used monthly HDDs for NOAA climate divisions and daily station data for the 13 cities with CME HDD contracts. The data cover cold seasons from October 1979 to March 2026. They removed linear trends and recentered the data to the present. They compared 16 El Niño and 16 La Niña winters with all years, and compared days in four North American weather regimes with all days. They tested significance by permutation and controlled false discoveries. They also computed fair put-option prices as the average historical payout.

Key findings

• The strongest signal is the December El Niño drop in HDDs and in daily HDD spread across the Midwest and Northeast. El Niño effects are modest in January and reverse sign in November.
• La Niña signals are weaker. They are mainly HDD increases in late winter and early spring in the Northwest and northern California.
• Pacific Trough is more frequent in December during El Niño. Pacific Ridge is more frequent in late winter during La Niña. Each regime’s HDD pattern resembles the matching ENSO pattern.
• At the stations, December HDD decreases during El Niño are significant in Chicago, Cincinnati and Minneapolis–St. Paul. February–March HDD increases during La Niña are significant in Portland and Sacramento.
• Fair prices of December put options in Chicago and Minneapolis–St. Paul are higher in El Niño years. The only statistically significant El Niño–La Niña asymmetry is the December–February reduction in variability.

Limitations

• Only 16 El Niño and 16 La Niña winters are available, so statistical power is low. The failure to detect asymmetry may simply reflect this.
• The authors cannot say whether the December concentration of the El Niño signal is sampling variability or a physical mechanism that would appear in independent data.
• The link between ENSO and HDDs through weather regimes is shown as consistency, not proof. This is a preprint that has not been peer reviewed.

Glossary

• Heating degree days (HDDs): The number of degrees by which a day’s average temperature falls below 65°F, summed over a period as a proxy for heating demand.
• Weather regime: A large-scale, recurring and persistent circulation pattern, such as Pacific Trough, used to group daily weather.
• Put option (HDD): A contract that pays $20 per HDD when observed HDDs fall below a strike value, protecting against a mild winter.
• Interquartile range (IQR): A measure of spread: the gap between the 25th and 75th percentiles of the values.