Just exactly how much better are the slew of supercomputers than Bert Ford with a crayon and a couple of thermometers used to be? When I was a kid Bert and his ilk used to get wrong just once a day, as far as I was concerned, about 6.30 after the early evening news and if I managed to stay up, well he might get it wrong again after the 9 o'clock news. But it was consistent wrongness, the forecast wouldn't change from 6.30 to 9.30. But now we have cutting edge technology to get it wrong in real time. Maybe the likes of air traffic control get the good stuff from IBM's best and I just get the junk from a ZX Spectrum but the only information you can trust, be it from the beeb, Met Office or Metcheck sites is how the weather is now. This is not a forecast, I have windows in my house, I can see what the weather is now. Maybe I should start a weather forecasting site. It's hot, it's fairly humid, it's summer; there is a high probability there are going to be heavy showers, maybe thunder storms sometime this week here. There, that's my forecast, and that's the professionals forecast as well except that they have changed the detail of the forecast from hour to hour, from rain on Friday to rain on Thursday and Friday, back to rain on Friday and light rain Monday, to rain on Friday and heavy rain on Monday all within the past 36 hours. This is not what I understand to be 'forecasting', it is weather guesswork, as it always was and there seems to be no more certainty now than there was 35 years ago beyond about the next 24 hours but now you can refresh your uncertainty as often as you desire.