GFS vs ECMWF Demand-Hub Outlook: 2026-09-17T00:00Z

Executive Summary

The 15-day outlook shows significant model disagreement, with the ECMWF forecasting a colder, more heating-demand-intensive period than the GFS. The ECMWF projects 31.9 weighted HDDs versus the GFS's 24.0, driven by a colder solution in the Midwest and East. Conversely, the GFS is warmer in the South, projecting slightly more national cooling demand (80.1 weighted CDDs vs. 77.4). A very large colder shift in the GFS this cycle (+8.8 HDDs) and a low overall convergence score (0.45) indicate high uncertainty in the forecast.

Hub-Level Weather

The national forecast is a tale of two regions, with key hubs showing material disagreement:

• Midwest & East (Chicago, NYC): These high-weight hubs are the primary source of model divergence. The ECMWF is significantly colder, projecting 36.2 cumulative HDDs in Chicago and 37.3 in New York City, compared to the GFS's forecast of 25.4 and 18.4, respectively. This accounts for the ECMWF's more bullish national heating demand outlook.
• South (Houston, Atlanta): The GFS projects stronger and more persistent late-season heat across the South. It forecasts 320.0 cumulative CDDs in Houston and 168.2 in Atlanta, compared to the ECMWF's 298.2 and 141.9, respectively.
• Northeast (Boston): In contrast to other eastern hubs, the GFS is notably colder in Boston, forecasting 96.2 HDDs against the ECMWF's 80.3.

Demand Implications (Natural Gas)

The current outlook suggests a transition from cooling to heating demand, typical for the mid-September shoulder season, but the models differ on the pace and intensity.

• Heating Demand: The ECMWF's colder forecast for the Midwest and East implies a stronger call on natural gas for residential and commercial heating, particularly in the 8-15 day period. The GFS is less supportive of heating demand but has trended significantly in this direction.
• Cooling Demand (Power Burn): The GFS outlook, with its stronger heat in Texas and the Southeast, supports more robust power burn for air conditioning load. This would help offset some of the model's weakness on the heating side.
• Run-over-Run Shift: The GFS model saw a substantial shift colder versus its prior run, adding 8.8 weighted HDDs and shedding 10.3 weighted CDDs. The ECMWF remained very stable (+0.2 HDDs). This suggests the GFS solution is volatile and may be attempting to correct towards the ECMWF's colder outlook, lending more credibility to the European model's forecast.

Model Agreement & Caveats

Confidence in this 15-day forecast is low due to significant model disagreement.

• Convergence: The model convergence score is 0.45, well below levels that would indicate high confidence. This is driven by a weighted mean absolute HDD difference of 1.25 per day between the models.
• Volatility: The major run-over-run change in the GFS highlights forecast instability. While the ECMWF has been consistent, the GFS's large correction introduces considerable uncertainty for NG markets pricing in this weather outlook.
• Source: This analysis is based on GFS and ECMWF model data from Open-Meteo, which is a valuable tool but not a substitute for specialized, commercial weather forecasting services.