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

Executive Summary

Both GFS and ECMWF models project a strong cooling demand signal over the next 15 days, with the ECMWF forecasting a slightly higher national load (+6.2 weighted CDDs). Both models trended significantly warmer compared to their prior runs, adding a combined 18.3 weighted CDDs. While confidence is high in the near-term warmth, material disagreements emerge mid-forecast (around June 16th) in the heavily-weighted Midwest, introducing uncertainty in the daily demand profile.

Hub-Level Weather

The national forecast is primarily driven by the two largest hubs, the Midwest (Chicago, 35% weight) and the East (New York City, 30% weight).

• Midwest (Chicago): This is a key area of disagreement. While cumulative 15-day CDDs are similar (GFS: 99.0, ECMWF: 101.1), the models diverge on timing. The GFS shows a significant cool-down around June 16th, forecasting 4.0 HDDs, while the ECMWF remains firmly in cooling territory with 4.0 CDDs on the same day.
• East (New York City): Models are in excellent agreement, showing sustained cooling demand. Cumulative CDDs are nearly identical (GFS: 156.4, ECMWF: 157.4).
• Northeast (Boston): This hub shows the largest model divergence. The ECMWF is substantially warmer, forecasting 95.8 CDDs versus only 45.3 CDDs from the GFS. The GFS also projects 14.5 HDDs over the period, while the ECMWF sees only 2.1 HDDs.
• South Central & South (Houston/Atlanta): Both hubs show consistent, strong cooling demand. The GFS is slightly warmer in Houston (+7.7 CDDs), while the ECMWF is warmer in Atlanta (+12.4 CDDs).

Demand Implications (Natural Gas)

The outlook points towards robust, above-normal cooling demand, which will support strong natural gas consumption for power generation (power burn).

• National Load: The ECMWF presents a slightly more bullish case with 153.1 weighted CDDs versus the GFS at 146.9. The GFS's forecast for 2.8 weighted HDDs is negligible at a national level.
• Run-over-Run Trend: The forecast has shifted materially warmer versus the prior 6-hour cycle. The GFS added +5.9 weighted CDDs and the ECMWF added a significant +12.4 weighted CDDs. This unified bullish trend adds confidence to the near-term heat.
• Timing Risk: The primary risk to the daily demand forecast lies in the Midwest around June 16-18. The GFS's cooler solution would imply a notable dip in mid-continent power burn, whereas the ECMWF's warmer forecast suggests more resilient demand through that period.

Model Agreement & Caveats

The forecast confidence is mixed, decreasing significantly in the second half of the 15-day period.

• Convergence: The overall convergence score of 0.83 is reasonably high, supporting the general theme of strong cooling demand. However, this score has decreased slightly (-0.005) from the previous cycle, reflecting the growing model differences in the outer range.
• Caveats: The primary uncertainty stems from the forecast for the Midwest and Northeast after Day 7. Traders should monitor subsequent model runs for either convergence towards a cooler GFS solution or a warmer ECMWF solution to gain clarity on mid-month demand.
• Source Limitations: This analysis is based on a 5-hub proxy using GFS and ECMWF data from Open-Meteo. The degree day calculations use a 65°F base. This is not a substitute for a full population-weighted national forecast but serves as a strong directional indicator.