The Boscastle storm of August 2004 and other heavy rainfall events of the last century in the area









The 'North Cornwall' and a comparision to some other similar heavy rainfall events during the last century in the SW of England

On August 17th 2004, after several hours of intense thundery rainfall in the drainage basin (202mm reported at Otterham) combined with a rising tide lead, a catastrophic flash flood ensued on the River Valancy, upstream of the village of Boscastle, N Cornwall, which, with the sheer volume of water (travelling at times up to 40 mph), broke its banks in the village. The force of the raging river swept over one hundred cars and three houses away, into the harbour and the open sea. A major rescue operation was launched by the RAF to airlift people to safety. A synoptic map of the situation (from UKMO) just after the event can be seen here and a satellite image here (courtesy Bernard Burton). Other areas locally such as the harbour village of Crackington Haven, north of Boscastle, were also flooded as well as parts of the River Ottery catchment area flowing east into the River Tamar.


Although the rain eased off in the evening, further rain fell in the W Cornwall area the following day (17th August) hampering the clean up efforts. Up to another 60mm fell in W Cornwall (more especially in the Camborne, Redruth and Helston areas) causing more local flooding with police advising motorists not to travel unless necessary.  

The technical background to the event
The synoptic set up that led to the devastating floods in the Boscastle region was a mixture of the general synoptic set up over the whole of the UK and also the local geography of this part of Cornwall. On the synoptic chart for 00Z on August 16th, a broad unstable S-SW airflow covered the entire SW Pensinsula. The winds aloft (700mb - one of the main steering levels) were strong from the SW. The airmass that covered Cornwall was unstable to modest temperatures. During the morning shower clouds developed inland as temperatures rose past the trigger temperature required to start convection.

The 12Z Camborne ascent (see side panel) illustrated the convective potential; when convection had been initiated tops could easily reach around 38,000 ft (about 250mb). Once high tops were established the storm was able to "vent" quite well as strong winds aloft allowed air to move away from the storms centre whilst new growth could be initiated near the surface.

Whilst temperatures were not that high in the area, they were just high enough to start a slight sea breeze along the north coast. With the prevailing surface wind from the S-SW this, along with the slightest sea breeze along the north coast, was enough to initiate a shower "train" convergence line or zone.

(A shower train is a constant line of showers that continues to develop almost on the same spot whilst cells are advected along the train in the direction of the steering flow). The friction of the coastline in the location was also aligned to the 700mb flow and thus with the abundant moisture aloft, these shower cells quickly grew and released the large quantities of moisture stored aloft.

Radar loops show that the shower cells initiated over the high ground in North Cornwall at around lunchtime on 16th; they grew and then an almost constant 'shower train' sat over the area affected for at least two hours. With rainfall readings being confirmed it seems likely that as much as 200mm fell in the space of 3 hours. This is in line with radar estimates from the afternoon and goes some way to explaining the events on the ground.


12Z tephigram at Camborne, Cornwall circa 30m to SW of the area
(click image to open tephi in large window)

See also a related BBC news report just here. Below is the regional radar (courtesy Paul Blight) for the SW at 1645Z showing the area of heavy rain on the north coast of Cornwall, the red echoes (heavy rain) remaining nearly stationary for some hours over the Boscastle area.  


The history of severe rain fall events in the area


This event, though clearly severe, was not that exceptional for the area and has occurred before in living memory, notably the Lynton -Lymouth disaster of 1952. There have been a number of other local severe summer flooding events near or on the north coast of the South West peninsula.


These have included an event on August 6-7th 1770 when a great flood devastated Lynmouth (Devon); perhaps a forerunner of the 1952 storms and floods, and it seems, anecdotally, that there may have been perhaps even heavier rainfall.

In the more recent past, on the 28 June 1917 Bruton, Somerset, 242.8 mm of rain fell in 24 hours although this wasn't a local event though as such. In Somerset the rain began late in the afternoon of the 28th, peaked in the middle of the night and contined until midday on the 29th. About 215 mm fell at Aisholt, in the Quantocks, 213mm at Timberscombe, Somerset and 150 mm at Street, near Glastonbury. Look at the day's synoptic chart here

On the 16 August 1924 at Cannington, Somerset, 238.8 mm fell in 24 hours. This was clearly a more localised fall with 215mm falling in just five hours. In fact this remains the record daily rainfall for August, anywhere in the UK. Look at the day's synoptic chart here

On the 6th August 1930 at Cheddar, Somerset thunderstorms ensured that 111mm was recorded at Cheddar, Somerset in a short period, and the famous caves were flooded. Look at that days synoptics chart here.

A few years later on the 4th August 1938 at Abbey Park, Torquay 162mm fell in a severe thunderstorm that hit East Devon just before dawn. However it was not that local, as 50 mm of rain fell in 6 hours over a large area from Tintagel to Exmoor and 100 mm was reported in an area lying from Paignton on the south coast of Devon, north to Two Bridges on Dartmoor. Hailstones as large as small walnuts lying 10 cm deep with four hours of almost continuous lightning were reported and severe flash flooding followed. The day's synoptic chart is here

The most infamous flooding event in the area in living memory was on the 15-16 August 1952 at Lynmouth, Devon, when 229.5 mm of rain fell in 24 hours. This was the result of heavy rain on Exmoor on the 15-16th. 228.6 mm of rain fell in 22 hours at Longstone Barrow, draining into the West Lyn river. 275 mm is estimated to have fallen over parts of Exmoor and up to 300 mm is actually claimed to have fallen in a day at Simonsbath. Although the event was preceded by 21 hours of heavy rain starting around noon on the 15th it also followed heavy rainfall over preceding two weeks that had allowed high river levels and very waterlogged ground. A seven hour long intense downpour from an almost stationary 'supercell' occurred in the late afternoon. There was a sudden surge of water here, as seems to have happened in the 2004 Boscastle event. 

The reason for this, as may well be found to be the case at Boscastle, is that boulders and trees had created a large temporary dam upstream that suddenly gave way, and around 200,000 tonnes of huge boulders moved rapidly downstream in the floodwater. The death toll was 34, twenty-eight bridges were destroyed and 93 houses ruined or damaged so badly that they had to be demolished later and 420 people were left homeless; 28 cars were wrecked and another 38 disappeared out to sea. This was clearly on a different scale to the Boscastle event. The article here has a good discussion of the event and the day's synoptic chart is here

A few years later on the 8 June 1957 Camelford, Cornwall, (very close to Boscastle) recorded 203.2 mm in 24 hours. This was perhaps the most similar of those documented here to the recent Boscastle event. A thunderstorm led to 203 mm of rain falling, with 140 mm of it in two and a half hours and about 70mm (nearly 3 inches) in one hour. Evidently there were hail drifts reported of up to 2 feet deep and although some bridges were destroyed there was no loss of life. A synoptic chart for the day is here

Again quite locally, on the 14 June 1965 at Wadebridge, Cornwall 140mm of rain fell in 220 minutes, in another similar event to Boscastle, which is very close. The synoptic set up chart doesn't look all that concerning at first glance just here though, it must be said.

A few years later still, on the 10 July 1968 at Chew Stoke, Mendips, Somerset 175 mm was recorded. This was not too localised however with 125 mm also reported in 17 hours at Bristol,leading to flooding and damage. Many bridges were swept away though in the Somerset area. Incredibly up to 125mm (5 inches) of rain fell on other parts of Somerset and Devon in just 90 minutes and several people in Sidmouth, Devon drowned. The days synoptic chart is here

More recently on the 11-12th July 1982 Bruton, Somerset (nb see June 1917) recorded 113 mm of rain in 16 hours; the River Brue burst its banks, leading to flooding and lightning strikes led to power losses. The daily synoptic chart is here

In the far SW on the 22nd July 1983 Penzance, Cornwall flooded as an area of low pressure brought thundery rain as it moved N from Biscay into very warm air over the UK. The synoptic chart is here

In the last decade on the 9th June 1993 Culdrose, Cornwall  recorded 125mm of rain in the 9 hours to 09Z as a thundery low moved north from Biscay, 92mm of this in the 2 hours to 08Z. The synoptic chart is here

It's also worth giving a mention, in closing, to the infamous 18 July 1955 Martinstown, Dorset event, 279.4 mm in 15 hours - though this was not related to an event near the north coast. This is the British record for daily rainfall, 190mm- or nearly 7 1/2inches -fell in 4.5 hours. The rain came with thunderstorms in two waves, the first starting at 2.30 pm and the second at 9 pm but the heavy rain fell over a large area of Dorset eg Upwey, Dorset, reported 241.3mm, Dorchester 187.5 mm, Weymouth 178.8 mm with flooding. The synoptic chart for this event can be found just here

So it can be seen that though relatively rare, such events as Boscastle are a part and parcel of our climates natural extremes, that we may expect to see once every decade or so in an area like the southwest. 

With grateful thanks to Dr Trevor Harley's site at for some of this information.

Thanks also to Paul Blight for the technical account, tephis and radar and Dave Jameson for general reporting.

Synoptic maps courtesy 'Wettercentrale' archives