How to Read a River Water Quality Report: A Plain-English Guide to EA and NRW Data
A water quality report is not a verdict. It is a set of measurements, taken at particular places at particular moments, and read against standards that vary from river to river. Once you have that in your head, the spreadsheets get a lot less intimidating. The same handful of numbers comes up again and again, and each one answers a slightly different question. Read them properly and you can judge whether your local stretch is improving or sliding.
Here is how to make sense of the data published by the Environment Agency in England and Natural Resources Wales.
Find your monitoring point first
Before you look at a single figure, work out which stretch the data actually describes. Both agencies publish monitoring results under the Water Framework Directive, and both let you search by river name, place or map.
Then check where the sampling site sits physically. A point half a mile below a sewage works outfall tells a very different story to one on a tributary above it. Note whether it is above or below:
- a sewage treatment works or combined sewer overflow
- a large arable or dairy farm and its drainage ditches
- a weir, which oxygenates the water as it falls
- a town, with its surface water drains and misconnections
Two points five miles apart on the same river can be classified differently. That is not the agencies contradicting each other. It is the river changing. For example, a site on the River Lark above Bury St Edmunds might show good phosphate levels, while the same river below the town's sewage works fails the phosphate standard. The difference is the discharge, not the sampling method.
Also check the sampling dates and frequency. Most WFD monitoring is monthly or quarterly, with some continuous sondes for oxygen and temperature. A single reading from a dry August is not the same as a winter storm sample. The agencies often publish a summary statistic, such as the 90th percentile, which tells you how often a standard is exceeded, not just the average.
Phosphate: the number that usually decides the argument
Phosphate limits plant growth in most of our rivers. Add it and algae and blanket weed respond, which depresses oxygen and smothers the gravels where invertebrates and fish eggs live. In lowland rivers it is often the element that keeps a stretch out of Good status when everything else looks respectable.
Three things to check. First, units. Figures are usually given as milligrams per litre of phosphorus, written mg/l P. If a report says "phosphate" instead, the number may cover the whole phosphate ion, which is roughly three times heavier. A reading of 0.3 mg/l as PO4 is about 0.1 mg/l as P, so mistaking one for the other makes a river look three times better or worse than it is.
Second, magnitude. A clean upland river can sit in the low hundredths of a milligram per litre. An enriched lowland stretch fed by treated sewage and farm runoff can be several times that. As a rough guide, many lowland rivers are expected to be below 0.1 mg/l as P to reach Good status, though the exact standard depends on the river type and altitude.
Third, pattern. A February spike after a storm means something quite different to a stubborn plateau across three summers. Storm spikes often come from soil erosion and combined sewer overflows, while a summer plateau points to continuous discharges from sewage works or groundwater. Look at the seasonal breakdown, not just the annual mean.
Nitrate: a slower, quieter problem
Nitrate moves differently. It leaches through soils into groundwater and can take years to reach the river, so a change in farming practice this spring will not appear in the sample bottle for a long time. It matters for ecology, but it matters most for drinking water, where the regulatory limit is 50 mg/l. If your river feeds a public supply, nitrate may dominate the conversation in ways that have little to do with the invertebrates in the riffle.
Expect higher values in lowland, arable catchments and lower ones in the wet uplands. Judge it on a long run of data, never on a single reading. A river that averages 30 mg/l might be perfectly fine for fish, but it still adds to the cost of treating drinking water. In Nitrate Vulnerable Zones, farmers are required to limit applications, and the data can show whether those measures are working over a decade, not a season.
Dissolved oxygen: the hour matters as much as the place
Dissolved oxygen is reported two ways: in milligrams per litre and as percentage saturation. The second is often the fairer comparison, because cold water holds more oxygen than warm water. A river at 10 mg/l in January is not automatically healthier than the same river at 8 mg/l in July.
Oxygen also swings through the day. Plants and algae produce it in sunlight and consume it all night, so the lowest reading usually arrives just before dawn. Spot samples taken mid-morning will routinely miss the worst of it. A summer bloom can push saturation high by afternoon and crash it before sunrise, which is why continuous monitoring tells a different story to a monthly visit. Whenever you see an oxygen figure, read the time, the temperature and the flow alongside it.
If you only remember one thing: a dissolved oxygen reading without a time stamp is almost meaningless. The same river can be supersaturated at 3pm and hypoxic at 5am.
Ammonia: the sewage and farmyard signal
Ammonia is a useful early warning. It is toxic to fish and invertebrates, and it breaks down in the river, consuming oxygen as it does so. High ammonia usually points to a recent pollution event: a sewage discharge, a combined sewer overflow after heavy rain, or slurry running off a farmyard. The standard is strict, often below 0.5 mg/l as N, and some rivers fail it after every storm.
When you see an ammonia spike, cross-check the date with local rainfall records. If it coincides with a downpour, it is likely a storm overflow or runoff. If it appears in dry weather, it may be a continuous discharge or a misconnection. Ammonia is also reported as unionised ammonia (NH3), which is the toxic form, but most agency data gives total ammonia as nitrogen. Read the footnote.
Turbidity, suspended solids and the look of the river
Turbidity measures how cloudy the water is, and suspended solids measure the weight of particles carried in it. Both are proxies for silt, soil erosion, and organic matter. They may not appear in every WFD classification, but they matter because silt smothers gravels, carries phosphorus, and can clog fish gills. A river that runs brown after every rain is not just aesthetically poor; it is losing habitat.
Look at the pattern. A short spike after a storm is normal in many clay catchments. A persistently high baseline suggests ongoing erosion or a continuous discharge. If you are comparing sites, remember that turbidity depends on flow, so a high reading in a flood is not the same as a high reading in a drought.
What "Good status" actually means
Both agencies classify water bodies as High, Good, Moderate, Poor or Bad. Two caveats save a lot of confusion.
The classification is not a simple average of all the numbers. It works on a "one out, all out" principle for ecological status. That means the single worst element can drag the whole water body down. A stretch with excellent invertebrates, lovely oxygen, and low nitrate will still be Moderate if phosphate is too high. This is why phosphate so often dominates the debate: it is frequently the limiting factor.
Second, the classification is based on a rolling data set, usually three years, and it is compared against standards that are specific to the river type. A chalk stream in Hampshire is judged differently from an upland Welsh brook. So "Good" does not mean pristine. It means the river is close to what an undisturbed river of that type would look like, within the limits of the monitoring. A water body can be Good overall while still having occasional pollution incidents, and a Bad classification may reflect a single persistent problem rather than a dead river.
How to read a trend without fooling yourself
To judge whether things are getting better or worse, you need to compare like with like. Compare the same site, the same month, and similar flow conditions. A wet year will dilute phosphate and nitrate, while a drought year will concentrate them. So a single year of low readings does not prove a cleanup, and a single bad year does not prove decline.
Use the agencies' own trend tools where you can. The Environment Agency publishes a Water Quality Data Explorer that lets you download raw data and see long-term graphs. Natural Resources Wales has its own water quality pages and open data. Look for at least five years of data, and if possible, ten. Pay attention to the 90th percentile for phosphate and ammonia, because that shows how often the river breaches the standard, not just the average.
What to do with what you find
If the report shows a problem, you are not powerless. You can report pollution incidents to the Environment Agency hotline (0800 80 70 60) or Natural Resources Wales (0300 065 3000). If it is a persistent issue, join a local river group or a Rivers Trust, who often have their own monitoring and can press for action. You can also ask your local council and MP for the water company's storm overflow data, which is now published annually.
And remember: a water quality report is a tool, not a final judgment. It tells you what is being measured, where, and when. With a bit of patience, you can read past the jargon and see the river for what it is: a living system under pressure, but also one that can recover when the sources of pollution are tackled.