iiinova — where your digital self keeps evolving

James Peacock @peacockreports.bsky.social · 17h

Interesting way of putting it, which for some reason has reminded me of when I was once casually asked what the local record was for "the sunniest day on record". I failed to keep a straight face.

010

James Peacock @peacockreports.bsky.social · 06/10/2026

A high-based shower lit by the last rays of the sun made for a rare sort of sunset scene just now

070

James Peacock @peacockreports.bsky.social · 06/10/2026

There's basically no cross-model agreement regarding how much rain falls where tonight through tomorrow in England & Wales. Most areas have 10+ mm in one or two of the scenarios shown here but much less in the others. Guessing game!

160

James Peacock @peacockreports.bsky.social · 05/10/2026

Impressive deep convection occurring between Menorca & southern France this afternoon, as a low attempts to organise into a 'medicane'. Not saying it will do so - apart from one UKMO run last week, forecast modelling hasn't been interested in it succeeding as it heads eastward.

060

James Peacock @peacockreports.bsky.social · 05/10/2026

To the other side of the Americas, we find a Pacific that's been very active in the northern hemisphere. The East Pacific basin is chasing frequency records, with 2026 one of just four seasons on record to produce 3 or more category 5 hurricanes.

010

James Peacock @peacockreports.bsky.social · 05/10/2026

Hurricanes meanwhile tended to be detected more reliably (but not perfectly) many decades ago, so we have a fairer comparison. Makes 2026's lack of any so far all the more remarkable - it remains absent from the tracker. Chance of finally appearing on it later this week.

120

James Peacock @peacockreports.bsky.social · 05/10/2026

While notably quiet overall, 2026 is still demonstrating why a record-low Atlantic hurricane season for named storms is arguably impossible in today's world with its much improved detection of weak tropical storms well away from land when compared to many decades ago.

140

James Peacock @peacockreports.bsky.social · 05/10/2026

The last two ECMWF det. runs took the warm route with a pattern that in summer would produce record-challenging heat. In mid-October it depends on exact wind direction & speed as to whether date / weekly high temperature records can occur. Easy to end up with lingering fog.

030

James Peacock @peacockreports.bsky.social · 05/10/2026

After a roller-coaster later this week, next week's temperature theme for the UK has been mysterious lately with forecast models swinging between wintry chill & a return of warm-feeling afternoons. Latest ECWMF ensemble favours the latter but not emphatically. More runs needed.

160

James Peacock @peacockreports.bsky.social · 03/10/2026

2026 fell just short of last year for the longest unbroken run of days reaching at least 20°C here in NE Dorset, but when looking at 19°C it broke last year's record by 15 days, & at 18°C it's 14 days clear of 2022's old record & likely to add 4 more (including today).

070

James Peacock @peacockreports.bsky.social · 03/10/2026

That hint of another warm spell also exists in the ECMWF ensemble (to estimate the daily max temps add 1-2°C to the at-noon peaks shown here). The mean suggests something more modest, with Reading reaching 17-18°C during 12th-15th. That'd still be decently warm for mid-October.

020

James Peacock @peacockreports.bsky.social · 03/10/2026

This possibility sticks out from an otherwise still rather dry outlook for the next week or so, with only subtle suggestion of change on approach to mid-October. IFS has also been hinting lately at a new round of warm weather (after current one ends Tue) later in week 2.

120

James Peacock @peacockreports.bsky.social · 03/10/2026

Watching Wednesday for some potential hydrological drought-easing rainfall in SE England as a weak front interacts with a low over the near continent & reinvigorates. Past experience tells me this could easily end up too far southeast but who knows, it might not.

360

James Peacock @peacockreports.bsky.social · 02/10/2026

Another round of splendid evening sky

040

James Peacock @peacockreports.bsky.social · 02/10/2026

This has also been the sunniest year on record up to this point in my local area, just a little ahead of 2014 but well clear of the 3rd sunniest.

000

James Peacock @peacockreports.bsky.social · 02/10/2026

I find it hard to get my head around the fact that local to me, 1963's mean daily max up to this point was nearly 7°C lower than this year's.

110

James Peacock @peacockreports.bsky.social · 02/10/2026

Temperature wise, this year through the end of September has been well outside of reliable records for many southern parts of the UK. E.g. local to me it's over half a °C clear of the prior record for mean daily min, & over a °C clear for mean daily max. Huge margins for a 273-day mean.

282

James Peacock @peacockreports.bsky.social · 01/10/2026

Just noticed that yesterday's UKMO 12z run almost developed a 'medicane' just south of France early next week. Not seen it in any deterministic runs since, from this model or the other most reputable ones, so it seems to have just been an interesting 'rogue scenario'.

120

James Peacock @peacockreports.bsky.social · 01/10/2026

Deeper context: The 7-day smoothed 1991-2020 average (centred on target date) begins at 17.2°C for 1st October & drops by 1.1°C during the opening week. It's a big reduction month, the smoothed average for 31st being just 13.3°C. Biggest start-finish drop of any calendar month.

130

James Peacock @peacockreports.bsky.social · 01/10/2026

Looking at 19-22°C daily max temperatures here in NE Dorset for the first 6 or 7 days of October. Having to remind myself that the 1991-2020 average for this time of year here is only 16-17°C!

271

James Peacock @peacockreports.bsky.social · 01/10/2026

Since the start of spring, 2026 has been the driest on record by a margin of 19.3 mm over 2025's prior record, & sunniest by a margin of 124.8 hours over 2024's prior record. Yes really, over 100 hours beyond the old record for total sunshine. What a time for solar power here.

020

James Peacock @peacockreports.bsky.social · 01/10/2026

September continued the extraordinary run of much warmer than average months here in NE Dorset. Meanwhile it had 51.2% of the average rainfall, while sunshine was almost exactly average (100.2%). It felt rather cloudy due to having 125 hours less than the very sunny August.

151

James Peacock @peacockreports.bsky.social · 30/09/2026

Looking at higher up in the troposphere, we can see a cool signal in the ECMWF ensemble mean for above Reading as an example, but again that's a lot of scatter around it. So my current take is "probably feeling more seasonal, at least intermittently".

010

James Peacock @peacockreports.bsky.social · 30/09/2026

I've noticed a cool signal in the GFS ensemble for 11-15 days time in the UK & Ireland drawing some attention lately, but I'm not convinced at this stage. It's only slight, with a lot of scatter, & ECMWF's ensemble is more evenly spread around the long-term average.

140

James Peacock @peacockreports.bsky.social · 30/09/2026

Last night could have recorded a low in the 18s had the weather front cleared just a couple of hours later. The low for 18:00-06:00 UTC was 18.4°C. Had the front been some 6 hours later in time, I reckon one in the 19s would have been possible.

010

James Peacock @peacockreports.bsky.social · 30/09/2026

Observed an overnight low of 17.6°C here in NE Dorset. This would be by far the latest-in-year on record for a low in the 17s if not for 13th October 2018. That date saw an equally warm night half a month further into autumn, an extraordinary feat.

120

James Peacock @peacockreports.bsky.social · 29/09/2026

Perhaps - it depends how far north high pressure locates; some runs keep it far enough south that the UK generally sees unusually warm but breezy conditions (& with rain chances in the north).

010

James Peacock @peacockreports.bsky.social · 29/09/2026

ECMWF's modelled soil moisture status indicates widespread low amounts in western & central Europe, including much of southern England & the SE of Wales. A long-running hydrological drought continues, & the forecast looks anomaly dry widely until at least a week into October.

1123

James Peacock @peacockreports.bsky.social · 28/09/2026

For the UK, there's decent support for this high pressure to be of a 'warm' character... at least aloft. As October progresses, overnight temperature inversion development (cooler near surface than aloft) is an increasingly strong possibility if winds are light.

100

James Peacock @peacockreports.bsky.social · 28/09/2026

Looking at the 8-14 day range in the ECMWF ensemble clusters, I'm struck by how many feature high pressure over NW Europe. ~86% support for that days 8-10, still ~37% on day 14 which is notably high considering the pattern persistence that would entail.

140

James Peacock @peacockreports.bsky.social · 28/09/2026

Today & tomorrow look to reach the 30s °C in parts of SW France, perhaps the final occurrences within that country this year. Notable warmth spreads north-eastward tomorrow & Wednesday, with a wide swathe of mid-high 20s °C, impressive for the end of September.

040

James Peacock @peacockreports.bsky.social · 28/09/2026

Which is to say, the likelihood isn't zero, but it is very small. Nearly all potential 'medicane' storms fail to organise sufficiently due to land interaction, upwelling of cooler waters beneath, or wind shear.

030

James Peacock @peacockreports.bsky.social · 28/09/2026

In this model run, Malta experiences the passage of an eyewall, a relatively clear & calm eye, then another eyewall. Is something like this actually possible sometime & somewhere in the basin this October? Honestly, I think it is, but only if atmospheric conditions are just-so.

110

James Peacock @peacockreports.bsky.social · 28/09/2026

Seems this is a thing that the GFS model does now: Occasionally simulating a hurricane-style windstorm in the Mediterranean (an unusually strong example of a 'medicane'). Likely a manifestation of the bias seen with tropical cyclone development; too eager to spin things up.

180

James Peacock @peacockreports.bsky.social · 27/09/2026

Somehow even more remarkable here in NE Dorset from a ranking perspective (1943-present), sitting higher each of Feb-Apr: Jan 22nd Feb 2nd Mar 3rd Apr 5th May 1st Jun 1st Jul 1st Aug 2nd Sep 2nd or 3rd

030

James Peacock @peacockreports.bsky.social · 27/09/2026

Beyond that, wet weather looks to become restricted to scattered showers in some spots for the remainder of the week. Overall, a lot of variation to the weekly totals is expected - some areas seeing quite a bit while others don't see much by autumn standards.

030

James Peacock @peacockreports.bsky.social · 27/09/2026

The next wet weather arrives Wed-Thu courtesy of a deep low to the west. Initially the front looks to fizzle out while crossing the UK, but there are signs of a 'wave' developing along it Wed night/Thu morning, resulting in some heavy rain - something to keep an eye on.

130

James Peacock @peacockreports.bsky.social · 27/09/2026

Good example tomorrow of how weather conditions aren't all about how low or high the sea-level pressure is. A trailing front, though decaying, will bring some heavy spells of rain to parts of northern France, Belgium, & The Netherlands. Perhaps SE England in the evening.

191

James Peacock @peacockreports.bsky.social · 25/09/2026

Still seeing the potential for a notably warm 29th September in parts of England, though nothing exceptional unless the deep low to the west is adjusted to approach more slowly in future model runs. Also uncertain exactly how much cloud will be around to limit the warmth.

030

James Peacock @peacockreports.bsky.social · 25/09/2026

Gonzalo's forecast behaviour is somewhat unusual, a movement almost due north across Cabo Verde as a tropical storm today-tomorrow, before strong wind shear forces it to become extra-tropical early next week.

010

James Peacock @peacockreports.bsky.social · 25/09/2026

What's this, two named storms at the same time in the North Atlantic basin? Unremarkable in almost any season, but not this one. Both are forecast to stay below hurricane strength, though Fay might briefly get to within ~15 mph of the required max. sustained wind speed.

261

James Peacock @peacockreports.bsky.social · 24/09/2026

Just a little disagreement between UKM & GFS regarding how much low pressure to the west of the UK deepens early next week...! In the past few runs GFS has really gone off the idea of a favourable interaction with the jet stream. UKM has held steady.

021

James Peacock @peacockreports.bsky.social · 24/09/2026

Apples to oranges comparison - the potential deep low to the W or SW of the UK early next week is non-tropical in nature, which imposes limits on how rapidly it can deepen. Not saying that ECMWF run is unrealistic though, just that it likely represents the high end of the possible range.

000

James Peacock @peacockreports.bsky.social · 24/09/2026

For context, the latest-in-year 25+ on record here currently stands as 2nd October when 2011 reached 26.2°C (after 26.5°C the previous day). The latest 24+, 3rd October, when 1959 reached 24.4°C. Honourable mention to 1978 which reached 23.5°C on 12th October.

000

James Peacock @peacockreports.bsky.social · 24/09/2026

Only "for fun" but the end of the GFS 06z run is something I'd very much like to experience for the novelty of it. 25°C here in NE Dorset 9 days into October would be quite something. Alas, way too far ahead to take this scenario seriously.

100

James Peacock @peacockreports.bsky.social · 23/09/2026

Despite a brisk wind & a fair bit of cloud around, such a scenario would probably raise 2 m temperatures in England into the mid-20s widely. Potentially further north too depending on the timing of frontal systems (too slow for northern UK on this run).

010

James Peacock @peacockreports.bsky.social · 23/09/2026

With this potentially powerful windstorm exists another interesting possibility: An exceptional northward surge of very warm (or hot) air across western Europe. Nicely demonstrated by the latest UKMO run - something it's been quite consistent on (but not any other model).

110

James Peacock @peacockreports.bsky.social · 23/09/2026

I suspect the overnight ECMWF det. run got carried away here, but it predicted a genuinely hurricane-force windstorm just west of Ireland early next week. The potential for a powerful windstorm is legit, but this is probably near the top end of what's possible.

3123