Geology, Landforms and Erosion Rates

Combining evidence about a changing coast

A coastline's shape records the interaction of its geology with marine and land-based processes. Maps help you investigate where different landforms occur, while measurements over time show how quickly the coast is changing. Weather and climate data can then help explain variations in that change.

These sources answer different questions. A geological map identifies the materials beneath the landscape; an Ordnance Survey (OS) map shows its mapped shape and features; repeated coastal measurements reveal retreat. Using them together produces a stronger explanation than relying on one source alone.

Reading a BGS geological map

The British Geological Survey (BGS) Geology Viewer lets you examine bedrock, the solid rock beneath surface material, and superficial deposits, the loose material above it. Both matter: the material exposed at a cliff may include deposits as well as bedrock.

Begin by locating the stretch of coast being investigated. Read the geology legend to identify the mapped formations, then follow their boundaries towards the shoreline. In the online viewer, tapping the map reveals information about the geology. The layer legend updates as you move around the map, and its entries link to further descriptions of the formations.

Compare the geological boundaries with the coastal outline. Ask which formations coincide with projecting headlands, recessed bays or stretches of retreating cliff. Where rock bands meet the shore approximately at right angles, the arrangement is discordant; where they run approximately parallel to it, it is concordant. This describes the arrangement of rocks, not an erosion rate by itself.

A useful interpretation moves from observation to explanation: identify a coastal shape, identify its geology, then explain how the properties and arrangement of that geology could produce the shape. Map colour alone cannot establish resistance to erosion.

A worked geological interpretation: the Dorset coast

BGS mapping of the Jurassic Coast provides a real example of this reasoning. The Kimmeridge Clay Formation, Wealden Group and Oxford Clay Formation tend to coincide with low-lying coastal stretches and bays. More resistant units, including the Portland, Purbeck and Chalk groups, tend to form steeper coastal stretches and headlands. Less resistant materials are worn away more readily, leaving more resistant sections projecting into the sea: this is differential erosion.

The arrangement of the rocks also matters. In the central Isle of Purbeck, steeply dipping or nearly vertical Portland and Purbeck limestones run parallel to the coastline. Their resistance creates a barrier along this concordant stretch. Breaches in that barrier are associated with features including Lulworth Cove and the arch at Durdle Door.

For a map-based explanation, trace the limestone bands alongside the shore, identify where the barrier is interrupted, and compare these places with the coastal outline. The conclusion is not simply that limestone is present: its resistance and coast-parallel arrangement help explain the form of the coast.

Zooming in does not create finer geological evidence. The BGS warns that its general-interest viewer retains geological layers beyond the scale at which their detail is useful. Treat an apparently precise boundary cautiously when comparing it with a small coastal feature.

Recognising landforms on OS maps

Map scale tells you how a distance on the map relates to distance on the ground. On a 1:25,000 map, 1 cm represents 250 m. On a 1:50,000 map, 1 cm represents 500 m. The 1:25,000 map therefore shows the same area at a larger scale, which helps when examining small features.

First trace the coastline and distinguish land from water. A headland projects into the sea; a bay is an indentation in the coast. Compare neighbouring projections and indentations rather than identifying them from a name alone.

A spit is a narrow accumulation of deposited sediment attached to land at one end and extending into the water. To recognise a possible spit, trace its connection to the mainland and check whether its other end is free. A narrow feature joining land at both ends is not a spit simply because it looks elongated.

In this fictional illustration, the 500 m sediment projection would measure 2 cm on a correctly printed 1:25,000 map and 1 cm on a 1:50,000 map. Its attachment to the mainland and free end support identification as a spit. The headland and bay are recognised from the coastline's outline, not from labels alone. These tests transfer to real OS extracts, but this illustration is not evidence of a real location or its geology.

Using symbols, sediment and relief as evidence

The OS Explorer 1:25,000 key distinguishes water, mud, sand and shingle. Sand has a pale peach fill; mud has a muted grey-brown fill; shingle uses a pattern of small rounded marks. These distinctions help you identify the mapped material along a beach or sediment feature rather than treating every pale coastal area as sand.

The key also distinguishes a vertical face or cliff using a line with short tooth-like marks. Combine that symbol with relief evidence: contours close together indicate a steep slope, whereas widely spaced contours indicate a gentler slope. Explorer contours have 5 m and/or 10 m vertical intervals, so check the interval on the particular map. Heights are measured above mean sea level.

A mapped coastal projection with cliff symbols and steep relief supports identification as a cliffed headland. A narrow projection attached at one end, with mapped sand or shingle and little relief, supports identification as a depositional spit. Marsh, reeds or saltings symbols behind a sheltered sediment feature provide additional evidence of a low-lying wet environment, although they do not prove how the feature formed.

Use the particular map's key at either scale. Combine coastal outline, symbols, labels and relief: an outline alone cannot distinguish a steep cliff from a low shoreline. Small features such as caves or arches may not be represented clearly enough to identify confidently, especially at 1:50,000.

Calculating mean erosion rates

Coastal erosion wears away material; coastal retreat is the resulting landward movement of the shoreline or cliff edge. Repeated positions of the same coastal boundary can be compared to measure retreat. Measurements must refer to the same location and the same feature: comparing a cliff edge at one date with a waterline at another would not measure the same change.

For retreat measured in metres over an interval measured in years:

$$\text{Mean erosion rate} = \frac{\text{total coastal retreat}}{\text{elapsed time}}$$

The result is in metres per year, written m/year.

Consider an example dataset. At one fixed measurement line, a cliff edge retreats 3 m from 2018 to 2019, 1 m from 2019 to 2020, 6 m from 2020 to 2021 and 2 m from 2021 to 2022. Total retreat is 12 m over four years, so the mean rate is $12 \div 4 = 3$ m/year.

This does not mean that 3 m disappeared every year. The annual retreat varied considerably, and some movement might have occurred in a short event such as a cliff collapse. The mean summarises the whole period rather than describing every year within it.

Reading a real multi-year erosion record

At Aldbrough on the Holderness coast, repeated terrestrial laser surveys measured changes in cliff volume. Between September 2001 and November 2017, the central 100 m coastal section lost 53,000 m³ of cliff material over approximately 16.2 years.

Volume loss is not a distance. To express it as an equivalent recession, the researchers used a cliff height of 17 m and section width of 100 m. Dividing volume by this cliff-face area gives an equivalent retreat distance, which can then be divided by time:

$$\text{Mean equivalent recession rate} = \frac{53{,}000}{17 \times 100 \times 16.2} \approx 1.9\text{ m/year}$$

This represents the average landward movement of a simplified cliff face that would account for the lost volume. It is not a measured cliff-top retreat rate: the cliff has an irregular shape and material can be lost from different parts of it.

The survey table contains both incremental loss, the material lost since the previous survey, and cumulative loss, the running total since monitoring began. Its intervals are unequal, and no laser-survey data are available for 2008. For a whole-period rate, use the total loss and total elapsed time rather than simply averaging rates for unequal intervals.

Using UK weather and climate data to explain variation

Weather and climate data help investigate why erosion rates vary. Relevant information includes rainfall, wind direction, seasonal conditions and storm frequency. High-energy waves can remove coastal material; rainfall can increase saturation and soften cliff materials, contributing to gravity-driven mass movement.

Match the weather records to the erosion measurements by location and time period. If a survey measures loss over several months, compare it with rainfall or wave conditions over those same months, not just the weather on the survey date. Check whether rainfall figures are totals for equal periods: a longer interval can accumulate more rain without being unusually wet.

Aldbrough monitoring found broad agreement between rainfall and cliff volume loss, with peaks in loss following major landslides. The rainfall records before 2012 were averages from three stations within 23 km of the site, so they were not exact measurements at the cliff. Major landslides identified in August 2004, March 2010 and February 2017 also show why retreat should not be imagined as a smooth annual process.

UK weather records add context. June 2007 brought more than four times the average rainfall at Holderness. During the December 2013 storm event, the nearby Hornsea buoy recorded waves exceeding 6 m and the BGS weather station recorded a maximum wind speed of 20.8 m/s. Such evidence identifies conditions capable of contributing to coastal change, but an extreme event alone does not establish a long-term relationship.

Across several years, wave energy and cliff volume loss showed similar trends. However, the monitoring did not establish a causal relationship, and storm counts did not demonstrate a simple causal effect on volume loss. Geology, groundwater conditions and the timing of landslides also matter.

Weather measurements do not themselves calculate erosion. You need retreat distances—or clearly identified volume measurements—and elapsed time for the rate calculation; weather data help interpret the result. A short multi-year record can reveal variation, but it is not sufficient on its own to establish a long-term climate trend.

BGS geology maps

• Identify bedrock and superficial deposits using the legend and formation information.
• Compare geological boundaries with headlands, bays and retreating cliffs.
• Discordant: rock bands approximately perpendicular to the coast. Concordant: approximately parallel.
• Dorset: less resistant clay units tend to form bays; resistant units tend to form steeper stretches and headlands. Coast-parallel Portland and Purbeck limestones form a barrier breached at features such as Lulworth Cove.

OS landform recognition

• 1:25,000: 1 cm = 250 m. 1:50,000: 1 cm = 500 m.
• Headland: projection; bay: indentation; spit: narrow sediment feature attached at one end.
• Combine shape with the map key, sediment symbols and relief.
• Explorer keys distinguish cliffs, sand, mud and shingle; contour intervals are 5 m and/or 10 m. Check the particular map.

Mean erosion rate

$$\text{Mean erosion rate} = \frac{\text{total retreat (m)}}{\text{elapsed time (years)}}$$

• Unit: m/year; compare the same boundary at the same location.
• A mean hides variation and sudden losses.
• Aldbrough, 2001–2017: approximately 1.9 m/year equivalent recession, calculated from volume loss—not a direct measurement of cliff-top retreat.
• Incremental loss covers one interval; cumulative loss is the running total.

Weather and climate evidence

Compare erosion with rainfall, storm frequency and wave conditions for matching locations and periods. Aldbrough records show broad rainfall–volume-loss agreement and similar multi-year wave-energy trends, but not proof of a single cause. Weather explains variation; it does not replace erosion measurements or establish a climate trend from a short record.