GFS vs ECMWF Demand-Hub Outlook: 2026-06-28T00:00Z

Executive Summary

The 15-day outlook is firmly bullish for U.S. natural gas cooling demand. The ECMWF model is materially hotter than the GFS, projecting 26.3 more demand-weighted Cooling Degree Days (CDDs). This divergence is driven by a stronger and more persistent heat forecast in the high-demand Midwest and East hubs, particularly in the 6-10 day period. Both models trended hotter versus their prior cycle, lending confidence to the overall warm theme.

Hub-Level Weather

The national forecast is primarily driven by the two largest demand-weighted hubs, the Midwest and the East, where the models show significant disagreement.

• Midwest (Chicago, 35% weight): The ECMWF is substantially hotter, forecasting 250.3 CDDs versus the GFS's 204.6 CDDs (+45.7 CDDs). The largest divergence appears late in the forecast period (July 9-12).
• East (New York City, 30% weight): A similar story unfolds, with the ECMWF at 272.1 CDDs against the GFS's 224.7 CDDs (+47.4 CDDs). The key difference is timing, with the ECMWF showing a much more intense heatwave during the July 3-7 period.
• South/South Central (Houston & Atlanta): In contrast, the GFS is the hotter model in the southern hubs, partially offsetting the cooler risk it shows elsewhere. The GFS projects a combined +80.4 CDDs more than the ECMWF across these two locations.
• Heating Demand: Heating Degree Days (HDDs) are negligible across the forecast, as expected for this time of year.

Demand Implications (Natural Gas)

The forecast implies strong, above-normal power burns to meet air conditioning load through the first half of July.

• The ECMWF scenario, with its sustained heat in the population-dense Midwest and East, represents a significantly more bullish outlook for gas demand. This would stress power grids and likely support regional cash prices.
• The GFS, while still showing robust heat, suggests a breakdown of the most extreme temperatures in the East after the July 4th holiday, which would ease demand pressure relative to the ECMWF.
• The run-over-run delta is notable: the GFS added a significant +19.2 weighted CDDs, while the ECMWF added +6.1. This bullish trend across both models suggests increasing confidence in a hot start to July.

Model Agreement & Caveats

Confidence in the near-term (1-5 days) is high, but significant uncertainty exists in the 6-15 day period due to model divergence in key regions.

• Convergence: The overall convergence score is high at 0.9596. However, this metric is skewed by universal agreement on near-zero HDDs and does not fully capture the material CDD differences noted above.
• Key Disagreement: The primary uncertainty is the intensity and longevity of heat in the Midwest and Northeast from July 3rd onwards. The ECMWF's forecast for a prolonged heat event in these regions is the main bullish risk.
• Source Limitations: This analysis is based on a limited set of demand hubs provided by Open-Meteo and may not capture all regional weather variations impacting the total U.S. natural gas demand picture.