GFS vs ECMWF Demand-Hub Outlook: 2026-09-08 00Z

Executive Summary

The 15-day outlook features a significant divergence between models, with the GFS forecasting a substantially hotter and more demand-intensive period than the ECMWF. The national GFS forecast projects 149.03 weighted CDDs, a notable +30.1 CDDs above the ECMWF's 118.93. This bullish GFS outlook is primarily driven by persistent heat in the South and has become more pronounced in this cycle, adding over 14 weighted CDDs since its prior run.

Hub-Level Weather

The key drivers of the national forecast are the Midwest, East, and South Central hubs, which together account for 80% of the demand weighting.

• South Central (Houston) & South (Atlanta): These hubs are the primary source of model disagreement. The GFS is dramatically hotter, projecting 326.8 CDDs in Houston vs. the ECMWF's 241.1 (-85.7 diff), and 302.4 CDDs in Atlanta vs. the ECMWF's 217.9 (-84.5 diff). This points to a fundamentally different evolution of the late-summer pattern across the southern U.S.
• East (New York City): A secondary driver of divergence, the GFS is also significantly warmer here, forecasting 115.8 CDDs compared to 84.8 CDDs from the ECMWF.
• Midwest (Chicago): As the largest single hub (35% weight), it shows strong model agreement with GFS at 90.5 CDDs and ECMWF at 88.4 CDDs. Both models see a cooldown toward the end of the forecast period.

Demand Implications (Natural Gas)

The current forecast is entirely dominated by cooling demand, with heating load remaining negligible. The implications for Natural Gas demand are bifurcated based on model preference.

• GFS Scenario (Bullish): A forecast of 149.03 weighted CDDs implies a strong extension of summer cooling demand. This would support power burn and be constructive for NG prices, delaying the onset of the fall shoulder season.
• ECMWF Scenario (Bearish): A forecast of 118.93 weighted CDDs points to a more seasonal and faster moderation in cooling demand, particularly across the South. This would imply a quicker transition to lower shoulder-season demand levels.

The GFS model's +14.1 weighted CDD shift hotter since its prior run is a significant bullish change, while the ECMWF's smaller +2.9 CDD addition shows more stability.

Model Agreement & Caveats

The overall model convergence score is 0.76, indicating only moderate agreement. This score is weighed down by the major forecast differences in the southern hubs, especially in the 6-15 day range. This lack of agreement reduces confidence in the long-range outlook and suggests volatility is likely as models resolve this key difference in subsequent runs.

It should be noted that this analysis uses deterministic runs from Open-Meteo. Ensemble model runs would provide a better sense of probabilistic outcomes and the range of possibilities within each model's forecast system.