Three lights maintaining triangular formation – Wolverhampton, UK – 23 Aug 2026, 22:01 BST

[jon_eir]

Some suggestions:

- https://earth.nullschool.net/ lets you view a model based on historical wind data, at any time and place, and a choice of altitudes. Because the air movements are displayed as animated streams, you can sort of eyeball where something might have come from.

- I recall that the high altitude balloon hobbyists use some involved wind tracking tools for balloons they've lost contact with. Someone might have to trawl the "spy balloon" threads to find it, or maybe use a more direct web search.

The biggest problem here is that this sighting does not give a lot of clues to the altitude of the objects, and that introduces a source of error into the calculations.

i will for sure check this out-its my first real time looking a wind data etc. the data I had access to gave a pretty consist figure from 10-80m and 80-200m that's the ranges I modeled. with "take off" and "landing" and movement tru these ranges incorporated also. there may still a problem with the data i had

 

[JMartJr]

I'd advise caution when using wind data and how wind might have impacted an object within a few hundred feet of the ground (or of the tallest up-wind obstruction) unless you are dealing with wind coming over a wide expanse of water or an open prairie or something very smooth like that. (In the latter cases, I'd only be somewhat cautious.)

I have no source for this other than a few decades experience trying to deal with winds at lower altitudes, and comparing them to what is/was actually experienced on-site. What I will call "micro-local" conditions can have a lot of impact on how fast the wind moves, how smooth or turbulent it is, and what it's direction is at a specific point in time. Wind can vary a great deal from one end of the kite field to the other at a given moment, including in speed and direction. (Somewhere I have a photo from back in my pre-digital days of an even at an American football field, with flags at different ends of the field blowing directly towards one another -- I doubt I can dig it out at this point but if I can lay hands on it I'll post it.

Looks like much of the terrain around Wolverhamppton is reasonably even, but there are some hills -- and from the OP video it looks like there are some lights on a hill off in the same direction as the triangle of lights. So while the prevailing wind at that time is worth noting, just keep in the back of your minds that it is possible that wind AT THAT SPECIFIC POINT AND TIME may have been funneling around terrain or other obstructions, and may not be quite what the overall flow of wind across the area was.

 

[John J.]

My concern is the Open-Meteo data is an extrapolation of others' data, e.g. for the UK it might use data primarily from the Met Office.
The Met Office uses a supercomputer, but it doesn't make hyper-local predictions, so it might be unlikely that extrapolations made by others with less specialised IT will be able to do so with any degree of confidence. If such models were reliable, I suspect the Met Office would use them to predict bad weather on a more localised basis than it currently does. (Anecdotally, many people in Britain regularly moan about the weather they're experiencing not being what was forecast, but on a regional scale Met Office predictions are probably pretty accurate).

The Open-Meteo returns for a given location might have internal model consistency with the data gathered at other locations, but not accurately describe the actual weather at that specific location and time (unless that location coincides with meteorological monitoring equipment).

 

[jon_eir]

Some suggestions:

- https://earth.nullschool.net/ lets you view a model based on historical wind data, at any time and place, and a choice of altitudes. Because the air movements are displayed as animated streams, you can sort of eyeball where something might have come from.

- I recall that the high altitude balloon hobbyists use some involved wind tracking tools for balloons they've lost contact with. Someone might have to trawl the "spy balloon" threads to find it, or maybe use a more direct web search.

The biggest problem here is that this sighting does not give a lot of clues to the altitude of the objects, and that introduces a source of error into the calculations.

Thanks, useful suggestions. I've now checked Earth Nullschool at 21:00 UTC on 23 August. GFS gives about 55° at 10 km/h at the surface and 65° at 16 km/h at 1000 hPa near the observation location. That's broadly consistent with the Open-Meteo data I've been using, which also gives an ENE wind and therefore generally WSW passive drift. I wouldn't regard this as an independent measurement—both are numerical weather models—but it is a useful cross-check.
I've also been looking at the reverse-trajectory problem. The altitude issue you mention is important, but I've found the unknown range is at least as problematic. Putting hypothetical objects at 1, 2, 4 and 6 km along the estimated LOS produces substantially different possible source positions even when the assumed flight time is restricted to 5–20 minutes.
I'll have a look at the balloon prediction tools as well.
I think I've found out for myself what the limitations of the approach is now- using nullschool should proably be my first port off call in the future and unless I can derive some more accurate assumptions or the wide potential launch area maybe useful dont go any further -thanks for the input

 

[jon_eir]

I'd advise caution when using wind data and how wind might have impacted an object within a few hundred feet of the ground (or of the tallest up-wind obstruction) unless you are dealing with wind coming over a wide expanse of water or an open prairie or something very smooth like that. (In the latter cases, I'd only be somewhat cautious.)

I have no source for this other than a few decades experience trying to deal with winds at lower altitudes, and comparing them to what is/was actually experienced on-site. What I will call "micro-local" conditions can have a lot of impact on how fast the wind moves, how smooth or turbulent it is, and what it's direction is at a specific point in time. Wind can vary a great deal from one end of the kite field to the other at a given moment, including in speed and direction. (Somewhere I have a photo from back in my pre-digital days of an even at an American football field, with flags at different ends of the field blowing directly towards one another -- I doubt I can dig it out at this point but if I can lay hands on it I'll post it.

Looks like much of the terrain around Wolverhamppton is reasonably even, but there are some hills -- and from the OP video it looks like there are some lights on a hill off in the same direction as the triangle of lights. So while the prevailing wind at that time is worth noting, just keep in the back of your minds that it is possible that wind AT THAT SPECIFIC POINT AND TIME may have been funneling around terrain or other obstructions, and may not be quite what the overall flow of wind across the area was.

Thanks, that's a very useful point. I think I found the limitations of my approach the hard way The more I tried to narrow the tracks, the clearer it became that the uncertainties—especially at low altitude—quickly overwhelm any attempt at a precise source location.