colweather.org.uk

So, after all the various forecasts of rain on Bank Holiday Monday, it turned out to be a dry day here in Farnborough (and across most of the country from what I have seen). It was, however, overcast. We had an enjoyable walk as planned.
I appreciate the forecasters' honesty earlier in the week, when the computer models were bringing the low from Germany westwards quite quickly. However, even when the forecasts were going for rain on the Bank Holiday, I was telling my wife with considerable confidence that it was unlikely to rain and we should plan for our walk. I don't have any magic insight, but I do have decades of experience of looking at the UK weather, and although I have seen lows move from east to west, I have never seen them do so quickly. That reverse jet pattern is always slow moving. With two days to go, the forecasters agreed!
I don't want to bask in any glory for getting it right, but the point is that computer models are no substitute for a weather forecaster with experience!

Greg Gruner wrote:So, after all the various forecasts of rain on Bank Holiday Monday, it turned out to be a dry day here in Farnborough (and across most of the country from what I have seen). .... [snip] but the point is that computer models are no substitute for a weather forecaster with experience!

... All forecast agencies now (not just the Met Office) plug the 'apps'-style of getting the forecast out. I've now given up trying to tell people I meet that they shouldn't rely on the fancy, apparently detailed, forecasts they can pull up on smart-phones etc., but to no avail. Even listening to various sports commentaries (e.g., TMS, Tennis) the computer-generated output is extensively quoted as 'gospel'.

Recently, I was carefully looking at the BBC (based on Met Office 'BestData' output) graphics for a particular day: within the space of 6 hours, the output for the next day varied from cloudy/dry OR broken cloud/few light showers OR bright sun/light showers, ocnl heavy showers OR lots of heavy showers with little sun.

Totally useless for planning purposes - unless to emphasise that detailed forecasting 24/36 hours ahead is still, in my opinion, still the most difficult challenge in modern forecasting.

Martin.