Profile avatar
peacockreports.bsky.social
Head Meteorologist at MetSwift, working to revolutionise delivery of weather information. LinkedIn: https://t.co/pUCAQfAfRP
884 posts 344 followers 54 following
Regular Contributor
Active Commenter

Signs are, next Wed-Thu will probably go down as a 'close call' regarding very hot conditions affecting south-eastern England. Most model runs of the past 24 hours have predicted that an airmass conducive to max temps at 2 m into the 30s °C will move through overnight Wed-Thu.

Frequency distribution of official UK highest & lowest temperatures during spring 2025, versus a 2009-2024 climatology. The max temps are strongly skewed toward higher values, while min temps are slightly skewed toward lower values.

A big source of forecast uncertainty for later next week onward in Europe is what the Madden-Julian Oscillation (MJO) gets up to. Projections are widely spread regarding which phase out of 7, 8, or 1, an active MJO reaches before becoming indistinct.

Sunday's looking more showery in central & southern UK compared to what I saw yesterday, but still well-scattered. Mon-Tue sees high pressure bring largely dry weather so the S half of the UK & perhaps Ireland, while further north has to contend with a bit of rain at times.

The overnight runs weakened the low that crosses central UK tomorrow, which has toned down both the intensity of heavy showers & the thunderstorm risk. ECMWF's det. was putting down 10-20 mm quite widely, now 5-10 mm with higher totals more localised.

At 9.4 mm through 5th, the month-to-date rainfall local to me is above the long-term average for the first time since January! 21% above, a figure expected to increase today (which has already seen 2.1 mm) & tomorrow (could deliver over 10 mm), before a dry spell of ~4 days.

Finally found a moment to check the ECMWF ensemble clusters from last night's run. They express high confidence in the deterministic run's take on Tue-Thu. Yet if I consider many forecast models, I can find scenarios that are quite different to it, such as the 00z JMA run 🤔

Predictions by ECMWF, GFS, & ARPEGE for up to 1 pm BST today, versus observed. ECMWF overdid the swathe of higher totals in Dorset-Hampshire, otherwise not bad overall. GFS widely underestimated. ARPEGE performed best overall. All underestimated Brecon Beacons & surrounds.

Broadly, there's not bad agreement now between forecast models on the UK & Ireland rainfall totals through to the end of this week, but still quite a bit of variation for southernmost England. Widely 'useful' amounts regardless, after the dry spring.

The run of brisk westerlies continues to erode the anomalous warmth to the sea surface temps to the west of the UK, but they are showing some impressive resilience. Where it's stayed less windy to the southwest & north of the UK, SSTs remain ~1.5-3°C above the long-term average.

Our study just out in Nature: the extreme North Atlantic heat in summer 2023 was mainly due to weak winds causing shallow mixed layers, which heat up more quickly. Global warming is causing a trend toward shallow mixed layers, making such extremes more likely. 🌊 www.nature.com/articles/s41...

Steady on, UKMET forecast model… This sort of outcome, with hot air reaching the UK & perhaps Ireland too by next Wednesday, has been appearing increasingly often over the past 36 hours - but it remains something to keep watch on, not expect to happen.

The overnight ECMWF ensemble is essentially divided in two regarding some very warm or hot weather visiting at least southern UK next week; ~53% have it sticking around for 3+ days, the rest sweep it aside by Friday. The det. run produced one of the quicker sweep-asides.

Still a variety of path scenarios for the low crossing the UK on Saturday. Also apparent, an issue with the ECMWF model in such unstable setups: It simulates (very) heavy showers, but with questionably large sizes. In my experience this generally leads to overdone accumulations.

Totals 9 pm yesterday through 5 pm today. Decent amounts in many western parts, sporadically useful amounts further east with the locally higher totals in the south being wherever the squall line held together. It didn't here in Verwood, just 1.4 mm for the day.

The GFS & UKMET forecast models just can't seem to get along of late. They continue to disagree regarding how much the low crossing the UK on Saturday deepens, & strongly differ on how close a large low is located next Monday.

Cool temperatures in the northwest (fresh wind off Atlantic) & southeast (weather front) of the UK as of 2 pm BST. Warmer in between where it's drier with sunny spells, some spots even reaching appreciably above the long-term average.

Just happened across this on OpenStreetMap... I'd sure like to know what kind of flow dynamics are going on there (no indication of a bridge or tunnel).

Hunga Tonga undersea volcanic eruption cooled Earth by −0.38 W/m2 due to scattering of sunlight by aerosol particle emission with negligible warming from injected water vapour based on simulations 15 January to 15 July 2022: doi.org/10.1029/2024...

Looks like Thursday might not be as dreary as it could have been in Wales & England, thanks to the frontal rain moving through the preceding night. Sunshine & showers, some heavy, moving through swiftly on a brisk wind.

Here's a puzzler. Both GFS & ECMWF det. runs now have the low that crosses southern UK on Thu not deepening much, yet GFS predicts far less rainfall than ECMWF. The difference seems too large to be explained by model bias (In my experience the ECMWF det. tends to go a bit OTT).

Scroll through this video & watch southernmost England to see a very clear example of the sea breeze 'front' advancing inland. It's where the sea breeze meets a differing gradient wind, such that air converges & rises, assisting the development of cumulus clouds.

Today is expected to be the last of an extraordinary run of predominantly dry weather that began here on 27th February. Of the 96 days, 73 have recorded zero rainfall, placing 2025 between 1995 (70) & 2011 (76). Just 3 days have seen 4+ mm rainfall, which is equal to 2011.

A frequency distribution is a handy way to get a sense of what's 'normal' for daily max temperatures during a certain time of year. The 1991-2020 average max in my local area is 20.1°C, yet the distribution peaks at 19°C & is slightly skewed negative relative to the mean.

Here's spring 2025 versus 1943-2024, looking incredible for mean daily max yet totally unremarkable for mean daily min, here in NE Dorset. Much of CS England has likely seen similar. Anomalies v. 1991-2020 emphasise the contrast: Mean daily max: +2.84°C Mean daily min: -0.14°C