Selwick Bay: hold the line or managed retreat?
Selwick Bay has a 1.8 km frontage of low cliffs and a sand beach protecting 420 residents and 60 properties. Historical survey stakes show the cliff toe retreating at 1.2 m per year, and the beach has narrowed by 30 m since 1998. A concrete sea wall would cost $14 million to build and $250,000 per year to maintain, with a design life of 20 years. Managed retreat would require relocating 60 properties at an estimated compensation cost of $9 million and would allow 12 ha of saltmarsh to develop behind a new set-back embankment. Tourism in the bay is worth about $2.1 million per year and depends on the beach; 70% of visitors say they would not come if the beach were replaced by rock armour. The local council has a capital budget of $10 million and cannot fund both options. The shoreline management plan prefers 'no active intervention', but the parish council wants protection.
1. Using the figures, compare the 20-year financial cost of the sea wall and managed retreat. Explain one reason why the cheaper option may not be the better option.
Read the worked answer
Sea wall: capital $14.0 million + maintenance ($0.25 million × 20 years = $5.0 million) = $19.0 million over 20 years. Managed retreat: $9.0 million compensation, assuming no major new maintenance. Managed retreat is therefore about $10.0 million cheaper over the period. The sea wall also exceeds the council's $10 million capital budget by $4 million, so it would need an external grant or borrowing.
However, cheaper is not automatically better. Managed retreat displaces 60 households and may break social networks, reduce the local tax base and cause stress for elderly residents. The tourism estimate is also important: if the beach is lost or access is reduced, $2.1 million per year is at risk, which over 20 years is far larger than the difference in capital cost. A good judgement compares who bears each cost - the council, households, visitors or environmental groups - and how certain the tourism loss actually is.
2. Recommend which option the council should choose. Justify your answer with reference to stakeholders, sediment processes and at least one limitation of the data.
Read the worked answer
I would choose managed retreat with compensation, but only if the set-back embankment is designed to maintain a usable beach and public access. The sea wall protects property in the short term but interrupts the sediment cell: the cliffs currently supply sand to the beach, and holding the line cuts that supply. The beach would narrow further, putting the $2.1 million annual tourism income at risk. Managed retreat allows 12 ha of saltmarsh to develop, which can absorb wave energy and provide habitat.
Stakeholders conflict: residents want security, visitors want the beach, environmental groups want the marsh, and the council has only $10 million. The main data limitation is that 1.2 m per year is an average; erosion is episodic and may accelerate with sea-level rise. The tourism figure is also a stated preference, not a guaranteed loss. The recommendation is therefore conditional: monitor the set-back and beach width, and review if erosion exceeds the design rate.
3. Describe two pieces of field data you would collect to test whether the 1.2 m per year erosion rate is reliable.
Read the worked answer
First, I would measure the horizontal distance from fixed benchmarks - such as concrete posts or mapped building corners - to the cliff toe at repeated intervals over at least 12 months. Measurements should be taken at the same tide stage and with a tape or ranging poles, and repeated at several points along the 1.8 km frontage to capture local variation.
Second, I would compare historical maps, aerial photographs or old survey stakes with a known date and scale, georeferencing them to current positions. This gives a longer record than one year. I would also record storm events, because a single winter storm can move the cliff toe more than the annual average. The rate is only reliable if both short-term measurements and longer-term records agree, and if the measurement points are representative rather than chosen because they are easy to reach.