“By 2025 the data problem for weather prediction will be solved. Observational errors as we know them today will have been eliminated. Global weather prediction models with 1 km horizontal resolution will have reached the theoretical limits of predictability theory. Numerical predictions in the 0-2 day time frame will be essentially perfect. One week forecasts will be as accurate as 2-3 day forecasts are today. Weather information will be hundreds of times more useful to society than today and at less cost per person.”

I made the above “predictions” in my 1999 presentation Global Weather Services in 2025. It is now 5 years into this 26-year forecast, and my talk evaluates how my quarter-century forecast is verifying (and not verifying) and updates it as appropriate, giving examples of recent progress in the areas of observations, scientific advances and modeling.