Today and tomorrow are the next days of interest for #kenringwatch. “Snow Jan 9th & 10th, snow in midlands” was the prediction. Notably, Mr Ring modified that to “Moderate snowfall possible (nothing is ever definite in weather forecasting) within a day or so of Jan 9th & 10th in midlands” in the comments on my last post.
Of course nothing is definite in weather forecasting. Nobody said otherwise. Ken Ring claims an accuracy rate of 80% to 85%, though. Which would be impressive if it’s borne out. That’s up there with the last accuracy reports relating to Met Éireann. Mr Ring asks why I’m not tracking their progress, in comparison to his. Well, the main reason is that to compare like with like, Met Éireann would need to make long range weather forecasts for the whole year ahead, specifying events such as snow or rain on specific days (or, perhaps, “within a day or so” of specific days). And as that’s scientifically impossible, they simply don’t.
Mr Ring also questions why I’m not checking Met Éireann’s past accuracy, given that I pay for their services. Well, not my job. But somebody does – they’ve been audited in the past by the state’s Auditor General. They say in part:
“For ice warnings and severe weather alerts, there are three possible outcomes which can be monitored
- a hit: where the forecasted weather actually occurs
- a miss: where the forecaster fails to predict an exceptional weather event
- a false alarm: where adverse weather is forecast but fails to materialise.”
“A special exercise to consider the accuracy of severe weather alerts (fog, wind, snow and rain) in the period November 1998 to January 1999 indicated a hit rate of 92%.”
Not bad. Admittedly, those figures are for 15 years ago. But with improvements in instruments and software, I can’t imagine that Met Éireann has become less accurate over time. It is an issue that Met Éireann’s last issued annual report (which also contains accuracy figures) was for 2007, and that they’re no longer reporting on their accuracy due to staff and budget cuts. I don’t think that’s useful at all – how can you improve if you don’t know where you’re starting from?
But they are still providing short term forecasts that, on the face of it, seem reasonably accurate to me, rather than being jarringly inaccurate. They’re also providing and contributing to scientific studies such as this one, on global climate change (pdf).
Oh – Mr Ring doesn’t believe in climate change, either…
But speaking of 2007, it turns out that someone else was checking Ken Ring’s accuracy back then. Oh dear…