The Met Office single site forecasts
I have been meaning to write a line or two about the Met Office single site forecast. It’s the one that pops up in the Met Office app on your mobile device, or from the Met Office website in your desktop browser. After over a year of using it for Strathpeffer, I find the forecast that it gives is usually wrong more times than it’s right. That maybe because of the local proximity to mountains which lie not too far to the north, south and west. I know the rain shadow effect is something to behold here, and being in a steep sided strath means we do have an interesting microclimate when it comes to frost. I did a survey when we lived in Devon to try and put some statistics together to see how effective the single site forecasts where, but in a fit of pique I deleted the article I wrote after many months of data collecting. All I can remember from the findings is that things were far from perfect down in Devon, even though the forecast was produced less than 10 miles away in Exeter! My personal feeling is that temperature forecasts are seldom cold enough at night, missing many of the frosts we have, and not warm enough in the day during the summer months. Temperatures often don’t rise as quickly as they are forecast to in winter, probably because Strathpeffer is so well sheltered and frosts linger, sometimes all day. Cloud amount forecasts are generally too pessimistic, with occasional forecasts of seven days of cloudy skies. I’ll be the first to admit this is not the sunniest place in the UK, but cloudy skies for seven days in a row is pushing it a bit.
Take this morning as a perfect example of how just how wrong both the computer and text part of the forecast can be. Mostly cloudy in a region that spans the whole of Northern Scotland from the Hebrides in the west to the Moray Firth in the east. We woke up this morning to blues skies, with just an okta of lenticular medium level cloud in the sky. The clue is in the word lenticular, lenticular is synonymous with lee wave activity and mountains and we do have a few of those around here, so it would be natural to conclude by someone with even basic meteorological training that although it may be cloudy in Stornoway today, the cloud may well be well broken, even to small amounts on the other side of the northwest Highlands in the east. The fact that whoever wrote it included “Windy, with local gales” does give you an idea that lee wave activity was in their mind, but he maybe decided to stick with cloudy because that’s what the hourly cloud amounts from the single site model was telling him – who knows? Here’s what the visible satellite image evidence shows.
I would say it’s safe to assume that the hourly chance of rain for today in Strathpeffer is 0%, I don’t think I have ever seen a weather radar image devoid of any coloured pixels before, not surprising when the MSLP is between 1032 and 1050 hPa across IONA at the moment.
Man machine Out of synch
The other annoying thing is that the hand crafted text forecast is more often or not out of synch with the hourly single site forecast – just which one are you supposed to believe? Something that’s updated every three hours or the text forecast that may be revisited every 12 hours or so? I know I am not the first person to find problems with the single site forecast but at least I’ve got it off my chest. Maybe things will improve, but that hasn’t happened for many years in my opinion, perhaps it’s looked on as good enough for it’s target audience.