The Flood Investigation
Before reconstruction could take place the flood had to be investigated so that the new Lynmouth could accommodate such flooding without a repeat of the disaster. The floods were investigated as soon as possible and before any evidence could be destroyed. The assessment of the peak flow in each part of the Lyn river system was the first hydraulic investigation to be carried out a fortnight after the occurrence. As there were no gauging sites, walking surveys of the river bed were undertaken by engineers of the Ministry of Agriculture and the River Boards. From this information gauging sites were selected and surveyed with a cross section of the channel, the gradient and the measurement of the wrack line. Wrack lines indicate the peak flow where vegetation and loose material have visibly been removed. These may well have been washed down by side flow or alternatively had only indicated a transient surge. Eye-witnesses had emphasised the occurrence of frightening flood waves or surges in the Lyn rivers, which were no doubt caused by recurring ponding and subsequent release of flood waters from temporary reservoirs formed by the jamming of trees and boulders. The high water mark may thus have indicated discharges of short duration only (Dobbie and Wolf, 1953).
Flood Modelling
Sites were selected for their uniformity in cross section, their straightness and their ease of access. The Mannings formula was adopted as being easiest for calculation and as being the standard for both the Ministry of Agriculture and the United States Department of Agriculture. The coefficient value is of paramount importance, making a large difference to the results. For the Lyn river system, values of 0.045 to 0.08 were used (Dobbie and Wolf, 1953). Using four cross sections of each part of the main river and tributaries with clearly defined wrack lines and where there was no doubt about the channel change during the flooding, the rate of flow could be calculated using the well established Mannings formula:
Q = (AR 1/3 S 1/2) / n
Whereby Q = Rate of flow; A = Wetted area; R = Hydraulic radius; S = Water surface slope; n = A variable coefficient dependant on the roughness of the channel. While A,R and S are physical measurements, the value of "n" depends largely upon an experienced judgement, although published guides are available. In these circumstances n = 0.045 in tributaries, 0.06 in heavy wooded sections, 0.08 in rock filled gorges.
Without the very necessary run-off figures it would have been impossible to design with confidence any river improvements. Dobbie recognised the importance of assessing the river flow by two independent methods. Upon discovery of a natural gauging site, Dobbie constructed a 1/48 scale model of a rocky cleft situated 600 yards up the West Lyn from Lynmouth in the Hydraulic laboratory of the Imperial College London. The flow at the site was confined in a steep sided channel of remarkable regularity where the solid rock would have insignificant erosion during the event. Turning in quick succession of two right angles, the flow then emerges over a series of waterfalls. All boulders reaching this channel were quickly moved through this stretch (Dobbie & Wolfe,1953). Flood water had clearly scoured moss from the rock faces, giving a clear indicator of the maximum depth. The wax model reproduced the gorge, waterfall and 350 ft of the approach channel. Tests made using the model showed that the peak flow in the West Lyn had been 8,000 cusecs (226.4 cumecs) compared with 8,630 cusecs (244.2 cumecs) as calculated by the Mannings formula. It was felt that both figures were significantly close to indicate that the combination of the "n' figure used on both the East and West Lyn had produced a reasonably accurate figure for both, (Harris, 1992, 22).
Although felt to be less reliable than the above methods, the Met. Office isohyetal map was used to calculate that the maximum rate of runoff from the combined East and West Lyn, and was calculated as 22,000 cusecs (622.6 cumecs), (Harris, 1992). Dobbie's report built upon information of flood volume and velocity, recommended to the Council the main concerns with the rebuilding and reconstruction of the town. These were Dobbie's recommendations:
- The proposed span and cross-sectional area of all bridges.
- The width and depth of the rivers East and West Lyn, particularly near Lynmouth, Hillsford, Barbrook and Parracombe.
- The construction of dams to trap boulders on the East and West Lyn near Lynmouth.
- The construction of an overflow channel across the Manor House grounds to the sea from the confluence of the Rivers East and West Lyn.
Many of Dobbie's recommendations were rejected by the Roads and Rivers Committees. The construction of boulder dams was felt to be unnecessary as the delta was to be cleared of all obstructions, which, in the event of such a repetition, would supply adequate space to receive further debris. Also deleted was the flood relief channel. It was felt that, should a return event ever exceed the capacity of the widened channel, then the right bank nearest the Manor House should be constructed so that it could breach as happened during the flood, and carry flood water over the Manor House grounds to the sea, preventing damage to village side. The width of the river and bridge spans were completed to the recommendations of the Dobbie Report (Carnegie, 1956).
The Reconstruction
Reconstruction took several years. Plans were drawn up and a model produced, which was used for public debate. The opportunity was used not only to reconstruct but to plan for Lynton and Lynmouth's future as a tourist resort and to cater for the motor car. Roads were widened and new car parks created in both settlements.
It was necessary to completely rebuild the harbour wall. This was undertaken by Devon County Council and Lynton Urban District Council with the engagement of Dobbie as consultant engineer. The harbour wall, as with the new Riverside wall and Tors wall, were constructed over sheet steel toe piles driven into the ground to prevent any possibility of scouring. These were then constructed of concrete faced with local stone washed down in the flood. It was also necessary to rebuild the Rhenish Tower and a new river training wall was added to divert the river from scouring out the harbour, giving shelter to the anchored boats in times of sea storms and river floods (Carnegie, 1956).
Rebuilding the pier, January 1954 (Photo: Dr M Nightingale)
Many roads had been damaged during the flood. A number of these had to be completely re-made including road foundations, embankments and surfacing. The opportunity was taken to modernise, allowing for the requirements of modern traffic. The new Riverside Road was constructed to a width of 22ft. Tors Road was reconstructed to 16ft wide and both main approaches to Lynmouth from Watersmeet and from Countisbury were widened (Carnegie, 1956).
During the temporary phase of reconstruction a large amount of debris was excavated from the rivers to restore their flow. This, along with further material derived from channel widening, had to be removed from Lynmouth. Suggestions were made that the material should be hauled up Countisbury Hill, where it would be tipped over the cliff into the sea. Another suggestion was to deposit it out of sight somewhere on Exmoor. Both suggestions were considered too expensive. Instead it was agreed to tip the 70,000 cubic yards of excavated material at the extreme western end of the Esplanade, where it would make a useful extension to the car park (Carnegie, 1956).
The new channels in Lynmouth were designed and constructed to accommodate flood discharges of 15,000 cubic feet per second for the East Lyn and 9,000 cubic feet per second for the West Lyn, and a total of 23,000 cubic feet per second for the River Lyn below the confluence, these being the peak flow estimations for the flood estimated by Dobbie. It will be noticed that the flow of the River Lyn is less than the total of both the East and West Lyn. The reason for this is the unlikely chance of maximum discharge of the two rivers occurring at the same time, since both have different times of concentration. Additionally, it had proved difficult in designing for a 23,000 cubic feet per second discharge on the River Lyn due to the threshold levels of shops and houses. The parapet wall against the riverside road forms part of this defence and all openings are non-return flap valves (Carnegie, 1956).
Re-aligning the Lyn and constructing the new Esplanade, 1954 (Photo: Dr M Nightingale)
Special treatment was given to the Manor Grounds bank, where a two stage channel was created for the combined rivers. The bed was stepped on two levels to avoid the appearance of an empty channel: the lower 60 ft wide channel conveying the water during dry times of the year and the higher to be available for the higher flow levels on extreme occasions. The 12ft wide terrace was cobbled on a concrete bed. From here to the river is a 1 in 3.5 slope of reinforced concrete bed, and from the terrace to the Manor Grounds a further 1 in 1 slope, giving a total width of 100ft - a vast increase from the original 35 ft (Carnegie, 1956; Harris,1992).
During the flood, the Lynmouth Street Bridge over the West Lyn was completely destroyed without a trace. Lyndale bridge over the East Lyn was underscoured and the Prospect Corner Bridge was undamaged but filled to the roof with debris. In response to the Dobbie report the Lyndale bridge was reconstructed to give a clear span of 80 ft, and the new road over the West Lyn had a clear span of 50 ft. These were constructed from pre-stressed concrete beams on mass concrete abutments faced with local stone and of sufficient widths not to be blocked by trees and debris (Carnegie, 1956). Further upstream, a different design of bridge was used. The alternative to the expensive concrete structures was the relatively cheap, low wooden bridges which are deliberately built to fail in the event of a flood. In such times the wooden structure will simply lift off the stone base and be swept away ensuring that they do not become dammed (Keene and Elsom, 1990). Further measures taken to ensure future safety were to install flood markers further up stream where, during excessive rain periods, close studies can be kept and warnings could be issued if so required.
Watching demolition of the Lyn Valley Hotel from one of the new bridges over the West Lyn
(Photo: Dr M Nightingale)