GFS vs ECMWF Demand-Hub Outlook: 2026-09-06 00Z

Executive Summary

The 15-day outlook features significant model disagreement, with the ECMWF forecasting a materially warmer and more demand-supportive scenario than the GFS. The ECMWF projects 134.1 weighted Cooling Degree Days (CDDs), a stark contrast to the GFS's 117.8. This divergence stems from conflicting solutions for the eastern half of the U.S. in the second week of the forecast. Confidence in either outcome is low, as reflected by a poor convergence score (0.57) that has degraded significantly since the prior model cycle.

Hub-Level Weather

The primary drivers of the national forecast are the major population centers in the Midwest and East, which show substantial model disagreement.

• East (New York City & Boston): This region is the epicenter of the forecast divergence. The GFS model introduces a sharp, early-season cold shot around Sep 16-20, generating a cumulative 8.2 HDDs in NYC and a substantial 40.1 HDDs in Boston. In contrast, the ECMWF maintains significant warmth through Sep 15, forecasting robust CDD totals (120.4 in NYC, 71.3 in Boston) and minimal heating demand.

• Midwest (Chicago): While both models see a transition to cooler weather, the timing differs. The GFS sees a brief cool-down mid-month (Sep 14-17), while the ECMWF holds onto warmth longer before a stronger cold pattern sets in at the very end of the forecast period (Sep 19-20). This results in the ECMWF forecasting both more cumulative CDDs (79.5 vs. GFS's 71.8) and more HDDs (16.6 vs. GFS's 5.5) due to the colder finish.

• South Central (Houston): Both models forecast strong cooling demand, but the ECMWF is consistently hotter, projecting 279.3 CDDs versus the GFS's 244.7. The GFS shows a more pronounced dip in temperatures mid-forecast that the ECMWF does not.

Demand Implications (Natural Gas)

The conflicting weather outlooks present very different scenarios for Natural Gas consumption.

• ECMWF Scenario: The higher CDD forecast (134.1 weighted total) would extend strong cooling demand and support power burn for another week, a bullish signal for NG prices. This outcome would delay the seasonal decline in demand.

• GFS Scenario: The GFS's cooler outlook (117.8 weighted CDDs) suggests a more rapid end to significant cooling demand. While it introduces the first notable heating demand of the season (8.4 weighted HDDs), this is not yet enough to offset the loss of CDDs, presenting a more bearish case for total demand.

The run-over-run model trends highlight the uncertainty: the GFS shifted significantly cooler (-24.1 CDDs), while the ECMWF trended substantially warmer (+14.5 CDDs) versus the prior cycle.

Model Agreement & Caveats

Forecast confidence is low due to substantial disagreement between the GFS and ECMWF models.

• Convergence Score: The current score is 0.57, indicating poor model agreement. This has worsened considerably from the previous cycle, falling by 0.22. This growing divergence suggests instability in the forecast, particularly for the 8-15 day period.
• Primary Disagreement: The models diverge most significantly on the strength and timing of a trough moving into the Eastern and Midwestern U.S. after September 14th.
• Source Limitations: This analysis is based on a select group of demand hubs provided by Open-Meteo. It serves as a directional proxy for national demand but is not a comprehensive, population-weighted national total.