Skip to content
Air quality

MCERTS and indicative sensors: what your air quality data is admissible for

What UK reference, MCERTS-indicative and uncertified sensor data can legitimately be used for under LAQM — and what it cannot.

Three different things get sold as "air quality monitoring". They are not interchangeable, and the difference decides what you can put in an Annual Status Report, what you can put in front of a committee, and what a consultant will strike out of your evidence base.

This page sets out the three classes, what each is admissible for, and where a system like the one Make Sense is building would sit. Make Sense holds no certification. That is stated plainly at the end, and it is the reason this page can be written straight.

What MCERTS is

MCERTS is the Environment Agency's Monitoring Certification Scheme. It certifies instruments, people and laboratories against published performance standards (Environment Agency). It is a formal product certification scheme, run on the Agency's behalf by a certification body accredited by UKAS to ISO/IEC 17065 — Sira Certification Service, now part of CSA Group, which publishes the certified product lists that Defra's technical guidance links to.

MCERTS is not one bar. It is several, and the ambient air standards sit at different heights.

The three classes

1. Reference methods and demonstrated-equivalent methods. The reference method for PM₁₀ and PM₂.₅ is gravimetric — BS EN 12341:2014. For NO₂ it is chemiluminescence, EN 14211 (Defra UK-AIR). Because gravimetric sampling cannot run continuously, automated instruments are used instead, but only where they have been demonstrated equivalent to the reference method under the EC Guidance for the Demonstration of Equivalence of Ambient Air Monitoring Methods (2010). The relevant data quality objective is a maximum uncertainty of 25% for PM (Directive 2008/50/EC, Annex I).

Defra names the approved instruments individually, and the corrections that go with them. LAQM.TG(22) (v2.1, May 2025) tells you that the Met One PM₁₀ Smart Heated BAM 1020 needs its data divided by 1.035, that the Palas Fidas 200 PM₂.₅ needs dividing by 1.06, and that the PM₂.₅ TEOM1400AB, PM₂.₅ Opsis SM200 and unheated PM₂.₅ BAM 1020 "should not be used by local authorities" at all (7.161–7.175). Being an automatic monitor is not the qualification. Being on the list is.

2. MCERTS-certified indicative ambient particulate monitors. A separate, lower standard: Performance Standards for Indicative Ambient Particulate Monitors, Environment Agency, Version 4, August 2017 (LIT 7070). Two identical instruments run beside a reference method for at least 40 paired 24-hour results. The pass criterion is an expanded uncertainty at or below 50% — the data quality objective for indicative measurement, double the fixed-measurement figure.

Note the scope. It covers PM₁₀ and PM₂.₅ only. There is no UK indicative certification for NO₂, SO₂ or CO: TG22 states that electrochemical sensors, DOAS and cavity ringdown spectroscopy "are not currently approved for NO₂, SO₂ or CO, as they have not yet been tested" (7.134). Anyone offering a "MCERTS-certified NO₂ sensor" is describing something that does not exist.

3. Uncertified sensor systems. Everything else. Two European technical specifications now classify these — CEN/TS 17660-1:2021 for gases and CEN/TS 17660-2:2024 for particulates. Both grade a system as Class 1 (data consistent with the DQOs for indicative measurement), Class 2 (objective estimation) or Class 3 (relaxed, "not formally associated with any mandatory data quality objective"). TG22 was written while the PM part was still in development and describes it as such; it was approved by CEN in October 2024. Alongside these, PAS 4023 (BSI, 31 December 2023, Defra-funded) is the code of practice for selecting, deploying and quality-controlling them. It places the burden of proof on the user, not the vendor.

The part most people miss

An MCERTS indicative certificate does not, on its own, entitle you to a number.

Section 5.9 of the standard splits certified operation in two. Used qualitatively — pollution trends, source identification, pollution roses — an instrument may rely on factory calibration alone. Used quantitatively, to produce a concentration carrying the ±50% indicative uncertainty, "each instrument used for measurement has been calibrated on the specific site where monitoring is taking place against a standard reference method for a period of two weeks", with the resulting slope and intercept applied. That calibration must be repeated at least every twelve months, or when the instrument is moved to a different site.

So the certificate is a property of the instrument model. The admissible number is a property of the instrument, the site and the last twelve months of work. Buying a certified box does not buy you the second thing.

The Environment Agency has since narrowed this further. The IAQM reports that MCERTS Product Certification Bulletin PC4, published 1 July 2025, withdrew the previous certification ranges from all indicative instruments and set a typical measurement range up to a maximum of 150 µg/m³ as a one-hour mean, clarifying that indicative certification alone is not sufficient where concentrations may exceed typical ambient levels — construction, demolition, waste and industrial sites being the named examples (IAQM Position Statement, September 2025).

What each class is admissible for

Statutory LAQM work. Local authorities review and assess air quality, declare Air Quality Management Areas and produce Action Plans under Part IV of the Environment Act 1995, as amended by the Environment Act 2021. For compliance against the objectives, you need reference or demonstrated-equivalent data, or NO₂ diffusion tubes used with bias correction.

MCERTS-indicative instruments do not qualify. TG22 is unambiguous: instruments certified under the indicative scheme "are not suitable for measuring PM₁₀ or PM₂.₅ annual mean or the number of exceedances as they are not accurate enough to meet the expanded uncertainty requirements of equivalent instruments" (7.176, repeated at 7.261).

Uncertified sensors are further out again. Defra's LAQM FAQ 148 states they "are not currently accepted as a primary method for statutory LAQM reporting and therefore they cannot be used to formally assess compliance with air quality objectives", and advises against using them in Detailed Assessments, for model verification, or as evidence to declare or revoke an AQMA without additional validation against a reference monitor.

Revocation is the tightest gate of all. TG22 expects a detailed study carrying "the same degree of confidence as was provided for the original declaration", and three consecutive years of compliance evidenced through monitoring (3.57, 3.61). Declaration is deliberately easier — where annual monitoring and local intelligence show a persistent exceedance, authorities are encouraged to declare immediately rather than wait for a Detailed Assessment (1.43–1.44). Asymmetric, and sensibly so.

Everything else. Both Defra and the Environment Agency describe real uses for lower-class data, and they are not consolation prizes:

  • Screening for hotspots before committing capital to a reference-grade site
  • Deciding where the reference instrument should go — network design is the strongest case for cheap spatial coverage
  • Identifying spatial variation where local concern exists
  • Identifying short-term pollution events, and short local campaigns (TG22 7.176)
  • Trend analysis and source identification, including pollution roses (LIT 7070, 5.9)
  • Complementing diffusion tubes and continuous monitors for better spatial or temporal coverage
  • Community engagement and citizen science
  • Internal prioritisation — ranking streets, schools or routes for attention

The honest test for an officer: this data can tell you where to look and what changed. It cannot tell a court, an inspector or a Cabinet member what the concentration was. Cite it for the first. Never for the second.

One gap worth knowing about

Every framework above assumes a fixed site. CEN/TS 17660-1 and -2 both scope themselves to sensor systems "at fixed sites". The MCERTS quantitative condition is site-specific calibration, revalidated whenever the instrument moves. There is currently no UK certification route for mobile ambient measurement, because the certification model rests on co-location at a place the instrument stays.

That does not make mobile measurement worthless. It makes it Class 3 evidence with an unusually good spatial argument, auditable on the same terms as anything else, and it means anyone selling mobile air quality data should be explicit that no scheme certifies it today.

Where Make Sense sits

Make Sense does not hold MCERTS certification of any kind. The first vehicle system is in build.

Any air quality data the system eventually produces would sit in the third class described above: uncertified, mobile, and outside every certification scheme currently operating in the UK. It would be usable for screening, spatial pattern, network design and internal prioritisation. It would not be admissible for LAQM compliance reporting, for declaring or revoking an AQMA, or as the evidential basis of an Air Quality Action Plan. If you were ever to use it, PAS 4023 would be the framework you would be held to, and the burden of demonstrating fitness for your specific purpose would sit with you.

That is the whole position. If it is useful to talk through where indicative or uncertified data would and would not help a specific problem you have — including if the answer is that it would not — get in touch.

Sources

Last reviewed 12 August 2026

Related
Talk to us about this