Climate of Britain · 5000–2000 BC Graphs (desk)

Stonehenge, Wiltshire

The same winters as today, and an Avon too soft to carry stone

Mid-Holocene Wessex sat in the same maritime temperate climate as the modern Plain. Dragging a forty-tonne sarsen across a frozen River Avon would have needed close to a metre of solid ice, and nothing in this record produces winters cold enough to grow it. The Beaker-period ancestry replacement is real, but it follows the stones rather than preceding them — a climate crash timed to explain either event has the order backwards. The full set of proxy graphs behind these figures is on the desk page, built for a wider screen.

Annual mean8.6–11.1 °Ctoday 10.4 °C
July15.6–18.5 °Ctoday 17.1 °C
January1.8–5.3 °Ctoday 4.6 °C
Annual rain591–1,001 mmtoday 783 mm · wide band

These ranges include the full published reconstruction uncertainty across four millennial slices (5050, 4050, 3050 and 2050 BC): Mauri et al. (2015) pollen-grid anomalies for the 1° cell over Wessex, added to the Boscombe Down 1971–2000 climate normals. Annual temperature error is borrowed from the summer standard error, since no annual figure is published. Annual rainfall assumes spring and autumn track the mean of winter and summer. None of this is a weather diary for any single year.

Gold's formula, applied to the Avon

P = A × h²

The bearing capacity of floating ice is still commonly estimated from Gold's formula (Gold 1971), the standard rule of thumb behind Canadian ice-road practice: P is the safe load in kilograms, h is the ice thickness in centimetres, and A is a constant of about 3.5 for good-quality river ice under a conservative margin. Solving for a forty-tonne sarsen, close to the heaviest stones at Stonehenge, gives a required thickness of about 107 cm. Growing a metre of solid ice takes several thousand freezing-degree-days of sustained sub-zero weather. The reconstructed January means for this window run from about 2 °C to 5 °C — a southern English winter, not the conditions that build ice roads. A dusting of lying snow on the downs, as some accounts propose (after Hill 1961), is a smaller and separate claim from a load-bearing river crossing. Glacial transport of the stones themselves by moving ice sheets belongs to a much earlier period again, tens of thousands of years before this record begins, and these proxies have nothing to say about that question either way.

The dates do not line up into a single story

  1. ~3000 BCStonehenge I: the ditch and bank, and the first bluestones, raised by Neolithic builders.
  2. ~2500 BCThe sarsen circle and trilithons go up — still several centuries before the 4.2 ka label.
  3. ~2450–2400 BCBeaker arrivals begin. Ancient-DNA studies find a substantial ancestry replacement across Britain from this point; that is a genetic finding, not one this climate data makes or explains.
  4. ~2200 BCThe 4.2 ka interval, usually framed as a global drought. Roland et al. (2014) find no coherent event in the British and Irish peat record; the eastern-England tree-ring record (Bebchuk et al. 2025) describes this period in wet, inundation terms rather than a freeze.

Three claims, kept separate

A frozen Avon

Not supported as a working assumption for moving twenty- to fifty-tonne sarsens. Gold's formula asks for close to a metre of ice; these winters do not produce it.

Crash, then stones, then Beakers

The order is wrong. The large sarsens were already standing well before 2200 BC, and the Beaker-period replacement follows them. No British climate catastrophe in this record explains either event.

Delivery by ice sheet

A Pleistocene question, not a Holocene one. These proxies cover 5000–2000 BC, long after the ice sheets that could have moved stone were gone, and have nothing to add to that separate debate.

The strongest British signal across this interval is moisture balance, not temperature, and there is no thermometer on Salisbury Plain from the Neolithic. What exists instead is four millennial pollen slices set against a stack of peat, chironomid, oak and speleothem records — enough to rule out a frozen river and a climate-driven collapse, without pretending to be a weather diary for any single year.

Open the proxy explorer Source on GitHub Citations