Weather Daybreak Update, Saturday 18 July 2026

Weather Daybreak Update, Saturday 18 July 2026 ECMWF 00Z | Generated 10:30 UTC --- 00Z vs 12Z: What Shifted The overnight run made one meaningful change and confirmed everything else. The principal revision is a continued cold tilt for central Europe in the day-5 to day-8 window, with the Frankfurt ensemble now showing an 87% probability of temperatures more than one standard deviation below normal at day 5, up from what was already a strong cold signal in yesterday's 12Z. Amsterdam holds at 62% cold probability for the same period. The ECMWF IFS prints Frankfurt week 2 at 19.4°C versus week 1 at 15.0°C, which looks like a sharp warming recovery after the near-term cold trough clears, but that week 2 spread (17.4–25.3°C) is wide enough that traders should not bank on the warm tail yet. Elsewhere, the 00Z is quiet. UK and French temperatures are essentially unchanged: London 17.8°C ten-day average with negligible week-to-week variation, Paris holding 18.0–18.6°C across weeks 1 and 2. No meaningful revision to US heat, Dallas remains locked at 35.2°C average over 15 days with 252.5 accumulated CDDs, and Houston continues to print 32.5°C. Japan is also steady: Tokyo at 27.9°C, Osaka 29.8°C, Nagoya 30.1°C, consistent with the 12Z picture. The forecast change data is dominated by wind revisions over southeastern Australia and, to a lesser extent, near-term wind spikes over the Mediterranean. Adelaide sees peak wind forecasts trimmed by 3–4 km/h across multiple August dates while the near-term window (July 21–22) picks up slightly. These are boundary-layer refinements at extended range and carry limited signal for NEM dispatch planning this week. --- Run-to-Run Momentum The Frankfurt cold signal is now tracking across at least two consecutive ECMWF runs, and the 00Z strengthens rather than moderates it. This is not a single-run anomaly. The ensemble probability at day 5 moved higher overnight, and the EC46 week 2 Frankfurt range (17.4–25.3°C, central 21.2°C) shows the models have not resolved the spread, but the central estimate is clearly warmer than week 1, consistent with a cold trough arriving and then lifting, rather than a sustained cold pattern. The shape is: near-term cold spike, week 2 recovery. That structure has been stable across the last two runs. For European gas, the direction of travel is converging: a below-normal temperature event arriving across Germany and the Netherlands in roughly days 4–7, followed by warmer-than-normal conditions in week 2. The week 2 warming tail matters for storage injection pace, any week-1 demand uplift followed by week-2 warmth keeps the injection-season trajectory broadly on track, but the cold event itself will register in daily LDZ demand figures and is now confirmed enough to position around. Houston and the US Gulf Coast show no run-to-run drift, three consecutive runs have printed the same tight hot band, CDDs accumulating at a consistent pace. Japan similarly shows no revision momentum; the Nagoya/Osaka/Tokyo heat is steady and the week 2 ensemble pull (Nagoya central 27.6°C versus week 1 at 29.2°C) is a standard late-summer taper that has been present across multiple runs without deteriorating further. Australian NEM wind revisions, the most numerically prominent changes in the forecast-change table, show no consistent directional drift. Some August dates trimmed, the near-term window marginally higher. Read these as noise at extended range rather than a signal about generation mix for the coming week. --- Bottom Line The 00Z confirms and slightly intensifies the central European cold trough signal for days 4–7, with Frankfurt now carrying an 87% ensemble probability of a >1 standard deviation cold anomaly at day 5. This is the second consecutive run reinforcing the same structure, traders with TTF and NBP exposure should treat this as an established near-term demand event rather than a model outlier. The 12Z run today is the one to watch for whether the week 2 recovery (Frankfurt printing 21.2°C in the EC46 week 2 central estimate) firms up or continues to carry its current wide spread. US heat and Japan heat are confirmed and stable, no revision needed to those positions. The 00Z changed nothing for the southern hemisphere winter load picture.