GFS vs ECMWF Demand-Hub Outlook: 2026-06-22T00:00Z

Executive Summary

The 15-day outlook shows a material disagreement between models, with the GFS forecasting stronger national cooling demand (159.5 wCDD) than the ECMWF (151.8 wCDD). This divergence is driven primarily by the GFS forecasting significantly more heat in the South and East, while the ECMWF is warmer in the Midwest. Confidence in the outlook is tempered by a substantial bearish shift in the latest GFS run (-20.3 wCDD), which moved it closer to the more stable ECMWF, and a declining model convergence score.

Hub-Level Weather

The primary drivers of the national demand forecast are the high-weight hubs in the Midwest and East, which show conflicting signals.

• Midwest (Chicago, 35% weight): This is a key area of disagreement where the ECMWF is hotter, forecasting 130.2 cumulative CDDs versus the GFS's 114.8. The GFS model shows a sharp heatwave around June 28-July 1 followed by a significant cooldown, whereas the ECMWF maintains more consistent warmth.
• East (New York City, 30% weight): The GFS is hotter here, projecting 141.6 CDDs against the ECMWF's 128.7 CDDs. The most significant divergence occurs in the June 28-30 period.
• South (Atlanta, 10% weight): This hub shows the most extreme model divergence. The GFS is substantially hotter, with a forecast of 270.9 CDDs, far exceeding the ECMWF's 168.9 CDDs, particularly in the Day 8-15 range.

Demand Implications (Natural Gas)

The late-June forecast is entirely dominated by cooling demand, which drives power sector natural gas consumption.

• The GFS's hotter solution implies stronger gas burns, especially across the southern and eastern U.S.
• The ECMWF presents a more moderate, though still robust, cooling demand scenario.
• Crucially, the GFS trended significantly cooler versus its prior run, shedding 20.3 wCDD. The ECMWF was far more stable, with a minor shift of -1.9 wCDD. This major GFS correction reduces the credibility of its remaining heat and suggests the overall forecast risk is moving towards the cooler ECMWF solution.

Model Agreement & Caveats

Confidence in the 15-day forecast is moderate and has decreased since the last model cycle.

• The convergence score is 0.8858, which indicates fair agreement in the near term but highlights the growing model differences in the Day 8-15 period.
• The score has declined by 0.0612 from the previous cycle, signifying that the models are diverging, which lowers overall forecast confidence.
• The primary caveat is the volatility of the GFS model. Given its large run-over-run change, its hotter long-range outlook for the South and East should be treated with caution.
• Data sourced from Open-Meteo. Analysis is based on specific hub locations and may not perfectly represent broader regional averages.