Battle of Hastings 1066AD - A1 - Both Fleets, One Weather System

The storm that struck William’s fleet as it left Dives-sur-Mer in September 1066 is documented on page D — Sailing, Dives sur Mer to St Valerie. That page establishes that the Norman fleet lost approximately 83 ships on the coastal run to St Valérie, that William suppressed news of the casualties, and that the storm involved a counter-clockwise low pressure system tracking up the Channel.

This page proposes a wider meteorological model. The storm was not an isolated event. It was one component of a larger, persistent atmospheric pattern — an Omega block — that governed the entire campaign of September 1066.

That single pattern simultaneously:

• Guided Harald Hardrada’s fleet south from Norway to Yorkshire with following winds
• Trapped William’s fleet at Dives with persistent contrary winds for weeks
• Generated the counter-clockwise low that damaged both Harold’s and William’s fleets in mid-September — from opposite sides of the same 25 miles of water
• Finally collapsed to produce the light southerly that carried William across on 27–28 September

Three invasions. Three fleets. One weather pattern. One outcome — the Norman Conquest of England.

This is a theoretical model. It has been submitted to the Met Office National Meteorological Library for expert review. Comments from meteorologists or maritime historians are welcomed via the contact page.

An Omega block is a well-documented atmospheric configuration in which the jet stream splits around a slow-moving or stationary high pressure ridge, with a low pressure system on each side. The pattern takes its name from its resemblance to the Greek letter Ω. In the North Atlantic and European region it can persist for days or weeks, fundamentally altering weather patterns across a large area.

The model proposed here places an Omega block over the British Isles and adjacent seas in August and September 1066, configured as follows:

• Low 1 — a counter-clockwise low centred over or near the northern North Sea, approximately in the region of Edinburgh or southern Norway
• High ridge — a slow-moving area of high pressure trapped between the two lows, centred over central or southern England
• Low 2 — a counter-clockwise low developing in or tracking up the English Channel

Each component of this pattern affected the three fleets differently — determined entirely by their position relative to the system.

Harald Hardrada assembled his fleet at Solund in the Sognefjord, Norway, and departed in early August 1066 with approximately 300 longships. His route took him west to Shetland, south to Orkney where he collected the Jarls Paul and Erlend Thorfinnsson and additional forces, then south down the east coast of Scotland and England to the Humber, and finally up the Ouse toward York.

This route requires explanation. The initial westward leg from Norway to Shetland requires an easterly wind. A simple blocking high over northern England would not produce this — it would produce southerlies on the eastern flank, which would be useless or contrary for a westward passage. However, a counter-clockwise low centred over the northern North Sea or near Edinburgh would produce exactly the easterly needed on its southern flank, shifting to a northerly on its western flank as Hardrada turned south from Orkney.

This is the critical meteorological observation: Hardrada’s route implies Low 1, not a simple high. The low gave him the easterly for the Norway-to-Shetland leg, then the northerly for the Orkney-to-Yorkshire leg — following winds throughout.

The historical record confirms Hardrada’s remarkable speed. John of Worcester places him at Tynemouth on 8 September. Orderic Vitalis places the fleet at the Humber on 18 September. Scarborough was burned around 16 September. The fleet reached Riccall (8 miles south of York) around 18–19 September, and fought the Battle of Fulford on 20 September. This rate of progress is consistent with persistent following winds throughout the coastal passage.

Between the two lows, the Omega block configuration traps a high pressure ridge over central England. It is important to understand that William was not simply sitting under calm, light winds beneath this high. The high itself was not the direct cause of his difficulties. It was what the high did to the Channel that imprisoned him.

The high over central England acts as a wall along the northern side of the English Channel. Atlantic low pressure systems — which would normally track north-east across the British Isles — are deflected by this wall and instead find the path of least resistance: eastward along the Channel corridor itself. The gap between the blocking high to the north and any low pressure to the south creates a natural funnel, and small depressions are squeezed through it one after another, each producing westerly or south-westerly winds along the Norman coast.

For William at Dives, the result was a procession of small Channel lows, each arriving from the west and driving conditions that made a northward crossing to England impossible. A fleet of 700 vessels cannot beat west or north-west into a sustained westerly, nor can it safely cross a Channel running with steep westerly seas. Each time conditions eased briefly, another depression was funnelled through before William could act.

This mechanism explains several features of the historical record that a simple northerly headwind does not:

• The extreme length of the wait — weeks, not days. A single unfavourable wind direction might persist for days and then shift. A procession of funnelled lows arriving repeatedly from the same direction is a fundamentally different problem, and one that cannot be predicted or planned around from the beach at Dives.
• Why William moved to St Valérie rather than waiting at Dives — Dives is directly exposed to westerly funnelling along the Channel. St Valérie, in the sheltered Somme estuary set back behind the Normandy/Picardy headlands, offers considerably more protection from Channel westerlies. Moving there was not simply about a better departure angle — it was about finding a harbour that gave the fleet some shelter from the repeated Channel lows.
• Why his departure on 12 September was into deteriorating conditions — William moved in what appeared to be a brief window between two of these funnelled systems, only to be caught by the next one as it entered the Channel behind him. The fleet was still strung out along the Normandy coast when Low 2 arrived.

It also explains why William prayed publicly at St Valérie for a change in the wind and granted a charter to the church there in thanks when it finally came. He would have understood, as any experienced medieval sailor would, that the wind had been against him in a way that went beyond ordinary bad luck. Something large and persistent was funnelling conditions against him, and only a fundamental change in the large-scale pattern would end it.

As the Omega block began to weaken in early to mid September, Atlantic low pressure systems — previously deflected north by the block — began to find a route into the Channel. The configuration proposed here places Low 2 tracking along or up the English Channel around 12–13 September.

A counter-clockwise low in the Channel produces winds that rotate around its centre. The critical observation is that the same low produces opposite wind directions simultaneously on opposite sides of the Channel:

• On the French (Norman) coast — north-western side of the low: North-westerly winds, driving vessels east along the Normandy coast and onto the Alabaster Coast cliffs between Dieppe and Le Tréport
• On the English coast — northern side of the low: South-easterly winds, driving vessels onto a lee shore as they approached the Thames estuary from the west

The Channel is approximately 25 miles wide at the Strait of Dover. Both effects occur simultaneously, 25 miles apart, from the same rotating system. This is not two separate storms recorded in two separate sources. It is one storm, recorded from two different sides by chroniclers who had no knowledge of each other’s observations.

The storm map on page D shows the counter-clockwise rotation and its effect on the Norman fleet along the Alabaster Coast. Harold’s fleet was simultaneously on the opposite side of the same system, experiencing south-easterly conditions in the Strait and Thames approaches.

The collapse of the Omega block may itself have intensified Low 2. As the trapped high began to weaken, the sharp contrast between the stable air it had maintained over England and the incoming Atlantic air from the west would have steepened pressure gradients and potentially deepened any developing low, producing a more sudden and damaging storm than would otherwise have occurred.

Harold had maintained a fleet watching the Channel from the Isle of Wight throughout the summer of 1066. The Anglo-Saxon Chronicle records that the fleet was stood down on 8 September — the Nativity of St Mary. Supplies were exhausted and the men could not be held any longer. The fleet was ordered back to London.

There is only one Chronicle reference to losses on this passage, so it is not corroborated and is shown in red:

The Anglo-Saxon Chronicle (manuscript C) records: ‘the ships were driven to London; but many perished ere they came thither.’ Harold’s fleet suffered losses and damage on the return voyage from the Isle of Wight to London. Ships were wrecked and men were drowned.

The route from the Isle of Wight to London covers approximately 200 to 250 miles: east along the Sussex coast past Beachy Head and the Seven Sisters, through the Strait of Dover, around the North Foreland, and south-west into the Thames estuary. At typical sailing speeds of 3 to 5 knots this passage takes three to five days for a large fleet.

Leaving around 8 September, Harold’s fleet would have been in the region of the Dover Strait on approximately 10 to 12 September — within 20 to 30 miles of the French coast. This is precisely the moment William’s fleet was leaving Dives.

Harold’s decision to stand down the fleet on 8 September has often been criticised as premature or misjudged. But the same blocking pattern that had kept William trapped in Normandy all summer was also making conditions in the Channel difficult and unpredictable. Harold’s commanders may have had good practical reasons to move the fleet before conditions deteriorated further — and the fleet sailed directly into the developing storm.

William left Dives-sur-Mer on approximately 12 September 1066, heading north-east for St Valérie-sur-Somme — a coastal passage of around 150 miles. The fleet had been waiting since early August for a favourable wind. The trapped high of the Omega block had denied him that wind for weeks. When he finally moved, it was into the deteriorating conditions of Low 2 developing in the Channel.

There is only one Chronicle source for this passage, so it is not corroborated and is shown in red:

William of Poitiers, Gesta Guillelmi: ‘Presently the whole fleet, equipped with great foresight, was blown from the mouth of the Dives and the neighbouring ports, where they had waited for a south wind to carry them across, and was driven by the breath of the west wind to moorings at Saint-Valéry. There too the leader...concealed (as far as he could) the loss of those who had drowned, by burying them in secret; and by daily increasing supplies he alleviated want.’

As documented on page D, the discrepancy between the Dives ship list (777 vessels) and Master Wace’s figure for St Valérie (694 ships) implies approximately 83 ships were lost — over 10% of the fleet. The losses were predominantly among the slower, non-oared transport and supply vessels. The faster oared warships carrying troops made St Valérie. This left William with most of his army intact but critically short of food, fodder and equipment.

William’s wait at St Valérie was therefore not simply a wait for a favourable wind. It was also a wait for replacement supplies to be gathered from the surrounding area. He needed both before he could sail.

The Omega block configuration affected all three fleets simultaneously, but differently — determined entirely by each fleet’s position relative to the pattern:

The connection between Harold’s losses and William’s losses has never previously been made in the historical literature. Norman chroniclers wrote about William’s campaign. Saxon chroniclers wrote about Harold’s. Neither had reason — or possibly knowledge — to connect the two events. The Norman sources actively suppressed the storm damage. The link has therefore remained invisible for nearly a thousand years, simply because no one was reading both sets of sources simultaneously and asking whether the weather connecting them was the same weather.

For William to receive a southerly wind at St Valérie-sur-Somme, the following sequence is required:

• Low 2 exits the Channel eastward, tracking into Belgium or the southern North Sea
• The trapped high migrates north or north-east behind the departing low — moving toward Denmark or southern Scandinavia
• With the high now to the north or north-east of St Valérie, the clockwise circulation of the high produces a southerly at that location

This migration of the high north-east toward Scandinavia is consistent with the known behaviour of Omega blocks as they collapse. The jet stream’s seasonal southward shift in late September — as autumn establishes itself in the North Atlantic — is a plausible trigger for the breakdown of the block and the north-eastward displacement of the trapped high.

Crucially, the wind that gave William his crossing was light — historians estimate Force 2 to 3. This is consistent with the slack pressure gradient of a high that has recently moved into position. Heavy winds would have endangered the horse transports on which William’s cavalry depended. The weather did not merely allow William to cross — it gave him precisely the conditions he needed.

William of Poitiers confirms the wind shifted on 27 September and the fleet sailed immediately. The crossing was made overnight and the fleet landed at Pevensey on 28 September 1066.

In 2026, Professor Tom Licence of the University of East Anglia published research arguing that Harold’s famous 200-mile march from York to Hastings never happened — that Harold returned from Stamford Bridge largely by sea, using the fleet gathered at the Humber, which by that point included approximately 300 vessels captured from Hardrada’s Norse army at Stamford Bridge.

Prof Licence argues that the Anglo-Saxon Chronicle phrase describing Harold’s ships as having “come home” was misread by Victorian historians as a disbanding of the fleet. He contends the ships remained operational throughout, were used against Hardrada in the north, and were then used for Harold’s return south and a planned naval blocking action against William.

This is an important revision. However the meteorological model raises a problem that Prof Licence does not address.

If the Omega block collapsed in late September to produce the southerly that carried William across on 27–28 September, then Harold sailing south from the Humber at the same time would have been heading directly into that same southerly — against the wind that was filling William’s sails. The captured Norse ships, though capable open-water vessels, would have been rowing or tacking into contrary winds with exhausted crews on a 300-mile passage down the North Sea and around the Kent coast.

Prof Licence’s own evidence shows the English fleet arrived too late. The meteorological model may provide the missing explanation: Harold’s sophisticated naval strategy was undone by the same weather system that gave William his crossing.

There is a narrow exception. If Harold had sailed immediately after Stamford Bridge on 25 September — into the dying northerly tail of the block, before the wind finally shifted southerly on 27 September — he might have had a following wind south for one or two days. It was a window of perhaps 48 hours. He appears to have missed it.

Prof Licence’s counter-argument is discussed by Marc Morris at marcmorris.org.uk.

If the same low pressure system struck both fleets simultaneously in mid-September 1066, several previously puzzling aspects of the campaign fall into place:

Why Harold stood down his fleet on 8 September. The decision has often been criticised as premature. But if a major storm was already building in the western Channel, Harold's commanders may have had good reason to get the fleet moving before conditions deteriorated further. The timing was not poor judgement — it was a race against the weather that was nearly won.

Why William's wait at St Valerie was so prolonged. The received explanation is that William was simply waiting for a southerly wind. But as page D argues, he was also waiting for replacement supplies — the lost ships were predominantly the non-oared transport and supply vessels. The storm that drove him to St Valerie left him with most of his army but insufficient food and fodder for an invasion. He needed both the wind and the supplies before he could sail.

Why the two fleets never met in the Strait. Harold's fleet was in the Strait of Dover on approximately 10 to 12 September — the closest point of approach between England and France, the natural interception point for any blocking action. William was still at Dives. Had Harold known, or had the weather been different, the two fleets might have met there. Instead, Harold's fleet was driven through the Strait by the building storm, and William left Dives into that same storm two days later — the two forces passing each other's position separated by a day and 25 miles of open water.

Why neither side recorded the connection. Norman chroniclers were writing about William's campaign. Saxon chroniclers were writing about Harold's. Neither had reason — or possibly knowledge — to connect the two sets of losses. The Norman sources actively suppressed the storm damage. The connection between the two events has therefore remained invisible for nearly a thousand years, simply because nobody was reading both sets of sources simultaneously and asking whether the weather linking them was the same weather.

The meteorological, geographical and chronological evidence is consistent with a single counter-clockwise low pressure system sitting over the Strait of Dover in mid-September 1066, producing north-westerly conditions on the French side that wrecked William's supply ships on the Alabaster Coast, and south-easterly conditions on the English side that drove Harold's fleet onto a lee shore in the Thames approaches.

Both fleets were at sea simultaneously. Both suffered losses. Both were in the same Channel, separated by 20 to 30 miles of water, on opposite sides of the same rotating weather system.

One storm. Two fleets. One outcome — the Norman Conquest of England.

The storm did not simply delay William. It weakened Harold's naval capacity, disrupted his ability to maintain a blocking force in the Strait, drew both armies into the sequential crises — Hardrada in the north, William in the south — that ended at Hastings on 14 October 1066.

This theory invites further examination by maritime historians and meteorologists. The author would welcome contact from anyone able to model Channel storm tracks for September 1066 using palaeoclimatological data.

Primary Sources

• Anglo-Saxon Chronicle — multiple manuscript versions, see page 1, Anglo Saxon Chronicles
• William of Poitiers, Gesta Guillelmi — the only contemporary Norman account of the storm between Dives and St Valerie
• Master Wace, Roman de Rou — ship count at St Valerie (694 ships)
• Domesday Book and Annales Altahenses — references to a possible English sea engagement, October 1066

Secondary Sources

• Licence, T. (2026). Harold's Sea Campaign — UEA Research. University of East Anglia. Available at: UEA website
• See also this site: D — Sailing, Dives sur Mer to St Valerie for the Norman fleet storm analysis and ship loss calculations
• See also this site: B — The Ship List of William the Conqueror for the Dives vessel count of 777 ships