Experiments
Britain's clearest — and murkiest — bathing waters
Turning ten summers of satellite observations into one deliberately honest coastal infographic.
This started with a simple holiday observation: why can the sea around west Cornwall look extraordinarily clear while the water around Liverpool Bay and the Wirral so often looks brown or grey?
The lazy answer is pollution. The more interesting answer is geography, depth, tides, sediment, rivers, biology and the long history of a body of water before it reaches the beach. I wanted to see whether that contrast was strong enough to map—and whether a map built from real observations could still be made legible to somebody who had never opened a NetCDF file.
The question
The experiment was not trying to forecast tomorrow's swimming conditions. It asked a slower question: what does the water around Britain's coast typically look like during the summer holiday window?
The working period was 15 July to 31 August across ten complete summers, 2016–2025. Daily Copernicus Marine ocean-colour observations were reduced into a national climatology, with each summer given equal weight so an unusually observable year could not dominate the map.
The large-scale pattern is emphatic. The South-West Approaches, Isles of Scilly and much of western Scotland are relatively clear. Liverpool Bay, the Bristol Channel, the Wash and the Thames and Humber systems are markedly more turbid.
There is no perfect clarity measurement
The satellite products offered several candidates. Secchi depth estimates how far down a white disk would remain visible. Suspended particulate matter measures the material carried in the water. Kd490 describes how quickly blue-green light attenuates with depth.
Secchi depth and Kd490 were effectively telling the same story, so using both would have counted one optical signal twice. Suspended matter contributed genuinely different information, especially around estuaries and shallow shelf systems.
The final visual index therefore blends 65% log-scaled satellite Secchi depth with 35% inverse log-scaled suspended matter. It is an editorial device, not a recognised scientific index. That distinction matters enough to appear on the finished graphic itself.
Why Cornwall and Liverpool Bay differ
West Cornwall projects into comparatively clear Atlantic water, with deep ocean nearby and many rocky headlands and coarse sandy coves. Liverpool Bay is a shallow, energetic sediment system: strong tides repeatedly stir sand and mud while the Mersey and Dee add estuarine influence.
Around small offshore neighbourhoods, the contrast was strong. Porthcurno measured roughly 7.6 metres of satellite Secchi depth and 0.4 mg/L of suspended matter. The Wirral coast was closer to 2.0 metres and 7.7 mg/L. Those are climatological satellite estimates, not readings taken by somebody standing on the named beach, but the scale of the difference is real.
Clear is also not the same as clean. Official bathing-water classifications describe microbiological safety; this experiment describes optical conditions. The two should never be silently substituted for one another.
Does 100-metre data make the map better?
Yes, but not simply. A Sentinel-2 coastal product revealed sharper sediment plumes, near-shore bands and differences between bays and headlands that the one-kilometre national field could not show. Much of that detail survived a six-summer aggregate.
It also brought fewer usable observations, mosaic boundaries, missing areas and greater risk from shallow-bottom reflection. The conclusion was not to replace the national layer. The useful future shape would be a stable one-kilometre context with higher-resolution coastal detail introduced only where its coverage and behaviour deserve trust.
Designing the infographic
The first idea was an interactive map with pan, zoom, metric controls and point inspection. The better first deliverable was a static object that forced the analysis to become clear. If the story could not survive one page, interaction would only disguise the problem.
The final map colours a nine-kilometre ribbon along the coast rather than filling the whole sea. That keeps attention on the water people are likely to experience. A crisp independent coastline preserves coves, islands and estuaries while the measurement field underneath remains honestly coarser.
Official designated bathing waters provided the comparison set. Of 666 assembled locations, 611 coastal sites had valid source pixels within three kilometres and were eligible for the top-and-bottom ranking. Achmelvich came out clearest on the experimental index; Morecambe North came out murkiest.
What the map does not know
It cannot tell you what a swimmer will see this afternoon. Clouds, tides, rain, waves, phytoplankton and recent storms all matter. Satellite Secchi depth is an algorithmic estimate, and shallow bright seabeds can confuse an optical retrieval.
The index has not yet been independently validated against the historical Cefas dataset or a sufficient set of in-water observations. The coastal rankings should therefore be read as a designed comparison supported by satellite data, not a definitive league table of British beaches.
What I learned
This was a useful example of analysis and design improving one another. The visual question exposed scientific weaknesses: redundant metrics, thin coverage and smoothing that looked more certain than the source. The science, in turn, made the visual simpler.
It also reinforced the value of stopping at the right artefact. An interactive map may still be worth building, but the infographic already answers the motivating question and makes its uncertainty visible. That is a better outcome than shipping controls simply because the data could support them.
Experiment facts
- Status: Complete first publication experiment
- Type: Experiment
- Build mode: AI-assisted research, data processing and visual design
- Primary source: Copernicus Marine North Atlantic ocean-colour observations
- Headline window: 15 July–31 August, 2016–2025
- Working index: 65% Secchi depth and 35% inverse suspended particulate matter
- Important boundary: Clarity is not water quality, cleanliness or a daily forecast