Data note

What the English river water quality record can and cannot show

Hydrionis holds annual chemistry for thousands of English river monitoring points, computed from millions of laboratory results. It does not publish a national trend line from them, and this note sets out why: the archive begins after the improvement most people want to see, sampling frequency has fallen unevenly, and a large share of results for some determinands sit below the laboratory detection limit.

2026-08-24Version 1.0

4,846

Monitoring points held

255,337

Annual values

2000–2025

Period

61%

Fall in points sampled

The network has shrunk

English river monitoring points contributing an annual value, by year

Counted from the values Hydrionis holds, which require at least eight samples in the year.

View the data behind this chart

Underlying data
Period	Monitoring points with a published annual value (points)
2000	4,513
2001	3,758
2002	4,487
2003	4,566
2004	4,587
2005	4,500
2006	4,456
2007	3,710
2008	3,600
2009	3,410
2010	2,934
2011	3,250
2012	3,428
2013	3,462
2014	2,755
2015	2,520
2016	2,433
2017	1,399
2018	1,426
2019	2,476
2020	114
2021	1,492
2022	1,321
2023	1,817
2024	1,783
2025	1,766

Source: Environment Agency Water Quality Archive, aggregated by Hydrionis.

In 2000 Hydrionis holds annual values for 4,513 monitoring points. In 2025 it holds 1,766 — a fall of about 61%. The lowest year in the record is 2020 at 114 points, when routine sampling was suspended.

Why there is no national trend line here

Three reasons, each sufficient on its own

The archive starts too late. The public interface begins in 2000. The large improvement in English river chemistry — the collapse in ammonia and biochemical oxygen demand that followed the Urban Waste Water Treatment Directive — largely predates that, so the thing most people want to see is outside the window.
The panel is not constant. With the number of points sampled falling by roughly 61%, a national average computed year by year would partly measure which points were still being visited rather than what was in the rivers. Restricting to points with a complete record leaves too few to be representative.
Much of the signal is below the detection limit. For ammoniacal nitrogen, a large share of results are reported as less than the laboratory limit. Those are not missing and they are not the limit either, and a mean taken over the detected minority reads as pollution that is not there.

What the data does support is per-point analysis: how one river reach has changed, against its own record, with the sampling density visible. That is what the monitoring index publishes.

Coverage by determinand

Determinand	Points	Annual values
`ph`	4,629	71,698
`nitrate-as-n`	3,923	57,328
`dissolved-oxygen`	3,679	49,552
`dissolved-oxygen-saturation`	2,365	30,949
`orthophosphate-as-p`	1,899	25,038
`ammoniacal-nitrogen`	1,565	20,772

Data behind this

England River Water Quality Dataset v1.0.0 · 252,857 observations

Every figure above is computed from these datasets at build time rather than written into the text. If the underlying record changes, this page changes with it.

Citation

Hydrionis (2026), What the English river water quality record can and cannot show, version 1.0. Retrieved from https://hydrionis.com/research/what-english-river-quality-data-can-show/