What if a coast were made of hard rock?

Waves crash on the beach, again and again. Little by little, they carry sand away.

What if the coast were made of hard rock instead of sand? Would the waves wear it away too?

Every wave that crashes onto a coast carries energy. Over and over, waves pick up sand and pebbles, wash them away and wear the land back. This is called erosion.

Some coasts are soft sand, some are clay cliffs, and some are solid rock. In this lab, a small island faces the same ordinary waves every day. It starts as bare sand. What happens if you build the island from hard rock instead?

Coastal erosion depends on the energy delivered by waves and on the resistance of the material they attack. Bedrock, consolidated sediments and loose sand span a huge range of resistance.

This lab holds the waves fixed (1.5 m high, 8 s apart) and removes plant cover, so only the material changes. It compares a sand island with an identical island of hard rock, and then asks how much bigger the waves would need to be for rock to wear as fast as sand.

Our island is made of sand. It has no plants to hold it together.

The waves wash a little sand away every year. Slowly, the island gets smaller.

In the model, ordinary waves 1.5 meters high roll in every 8 seconds. Each meter of wave front carries about 8.8 kilowatts of power.

The island is bare sand, so the model's erosion index is 100: the scale is set by exactly this case. After 10 model years, about 94 percent of the island is left. After 50, about 72 percent. The side facing the waves wears away fastest.

Deep-water wave power is P = ρg²H²T ÷ (64π) ≈ 0.49 × H² × T kW/m, so 1.5 m, 8 s waves give 8.8 kW per meter of crest. The model's erosion index is 100 × (P ÷ P_ref) × softness × (1 − 0.7 × plants); with P = P_ref, sand (softness 1) and no plants, it is 100.

In the 50-year run the island keeps 93.7% at 10 years and 72.3% at 50. The model's 'years' and 'share left' show a pattern, not real meters or dates.

Now Olly makes the island out of hard rock. The same waves crash against it.

This time, the island hardly changes at all!

Choose 'hard rock'. The waves are exactly the same, with the same 8.8 kilowatts per meter. But the erosion index drops from 100 to 3: in the model, hard rock wears away about 33 times more slowly than sand. After 50 model years, about 99 percent of the rock island is still there.

Now try the biggest, longest waves the lab allows: 6 meters high, 15 seconds apart. The rock island still keeps about 75 percent after 50 years, more than the sand island kept under ordinary waves.

With softness 0.03 the index is 3 and the island keeps 99.0% after 50 model years. Clay cliffs (softness 0.35) sit between: index 35, 89.2% left.

To wear rock as fast as sand under ordinary waves, the wave power would need to be about 33 times greater. The slider's maximum, H = 6 m and T = 15 s, gives 265 kW/m, 30 times P_ref, and an index of only 90; the rock island keeps 74.7% after 50 years, against sand's 72.3% under ordinary waves. Storm waves on sand, by contrast, give an index of 1,111.

Sand grains are loose. A wave can pick them up easily.

Hard rock is all stuck together. Waves have to wear it away very slowly, bit by bit.

Sand is made of loose grains that a wave can lift and carry away. Clay cliffs wear back more slowly, and hard rock is solid, so in the model the waves can only wear it down a tiny bit at a time.

In the model, this is a simple rule: each material gets a 'softness' number, and the erosion speed is the wave power times the softness. Those numbers put the materials in the right order but are not real measurements, and the model's island is made of only one material at a time.

The NPS contrasts resistant bedrock, which forms steep cliffs under strong waves, with easily eroded rock, which forms gentle slopes. The model compresses all of this into a softness factor (sand 1, clay 0.35, hard rock 0.03) chosen to order the materials, not measured.

Its other limits: deep-water wave power with no shoaling, refraction or breaking; no sediment supply, so lost land never returns; no tides, storms or sea-level change; and a single material per island. Real coasts mix rock types, which is exactly what produces headlands and bays.

On rocky coasts, waves carve caves and arches. Some leave tall rock pillars called sea stacks.

Soft cliffs wear away much faster. Some lose a few meters every year.

The National Park Service explains that rocky coasts made of resistant rock form steep cliffs. Rocky points called headlands stick out into the sea, and waves bend around them and pound their tips. Caves form in the weaker rock on these points, and waves can wear a headland into an arch, then a lonely pillar called a sea stack.

You can see both kinds of coast on one stretch of shore in England. At Flamborough Head, chalk cliffs rise 120 meters, with stacks and caves. At Holderness, on the same council's coast, soft cliffs of clay, pebbles and sand wear back by up to 4.5 meters a year in places.

The NPS notes that resistant bedrock under high-energy waves produces steep profiles, while easily eroded rock in low-energy settings gives gentle ones. Waves refract around headlands, concentrating erosion at their seaward ends; sea caves develop in less resistant rock on promontories; and continued erosion turns arches into sea stacks, which are themselves temporary. Olympic National Park's coast has hundreds of them.

On England's East Riding coast, the council describes 120-meter chalk cliffs with stacks, caves and coves at Bempton and Flamborough Head, while the coast's 48 km of soft glacial till (clay, pebbles and sand) is particularly vulnerable: Holderness has one of North West Europe's fastest average erosion rates, up to 4.5 m a year in parts, and undefended places can lose 20 m or more of cliff at once.