I've seen some big seasonal and mid-range forecast model failures in my time but this one really takes the biscuit.
My post here gives a sense of how poor the 2nd week modelling proved to be:
https://x.com/peacockreports/status/1758450803331744096?s=20 So
why has this been the case? This is a very important question in the field of meteorology, because the usual sources of large error haven't been in play this time:
- The MJO did in fact propagate across the tropical Pacific at high amplitude
- Atmospheric angular momentum has increased to high levels as anticipated
- The polar vortex has been no stronger than average and is now undergoing a strong weakening (at least a minor SSW, possibly major).
All of this points toward establishment of high-latitude blocking with a focus to the NW of the UK. Yet this has barely managed to get going this month and that looks to remain the case going forward. So what gives?
To answer that, I believe we must turn to the North Atlantic sea temperature anomaly pattern:
https://psl.noaa.gov/map/images/sst/sst.anom.gif In the past couple of weeks, the tropical and eastern subtropical regions have been near or at record-warm levels for the time of year.
Meanwhile, importantly, the remainder of the basin has been largely near average. This is a change from earlier in the winter, when anomalous warmth dominated the northern reaches, making for a 'tripole' pattern conducive to a -NAO, which we did see at times in Jan.
Without that, the equator-pole temperature gradient has become exceptionally strong. That means we simultaneously have an increased atmospheric gradient, which accelerates the jet stream, and a lot more rising motion in the tropics, where the rising component to the Hadley cell is located.
My take is that this combo has both strengthened the Azores high and made it more difficult for Atlantic troughs to become negatively tilted and 'disrupt' against ridges that are attempting to become high-latitude blocking features. This means they've nearly always overrun such ridges, pushing them back toward the subtropics, when otherwise this probably wouldn't have occurred so much of the time.
The magnitude of this effect has well exceeded what, it appears, anyone had envisioned, although that is admittedly assuming that the higher CET estimates in here are tied to a general distrust of extended range modelling - forgive me if I'm wrong there.
Even the Met Office were forecasting increased likelihood of cold weather for mid-Feb until relatively short notice - and we know they are not ones to do so lightly.
CET competition wise, this is going to be a tough one to take, as not only is it a large error, but I unwittingly took far more of a gamble than I thought I would be, having expected to see a lot more below average estimates for the month. Due to the lack of a cap on the error added each month, this might be the point that, for the first time in 16 years, I have to throw in the towel because it would take a minor miracle to recover to a respectable position from such a heavy hit.
If you have any problems or queries relating to TWO you can Email
[email protected] 🙂
https://twitter.com/peacockreports 2021's Homeland Extremes:
T-Max: 30.4°C 21st Jul | T-Min: -6.8°C New Years Day! | Wettest Day: 34.1mm 2nd Oct | Ice Days: 2 (27th Jan & 8th Feb)
Keep Calm and Forecast On