TEOM PM2.5, Filter Oscillation and Semivolatile Losses

A tapered element oscillating microbalance, or TEOM, reads airborne particulate matter by measuring how a tiny vibrating filter assembly changes as collected particle mass increases. If a PM2.5 chart from a TEOM seems lower than a nearby reference sampler, ask first whether the instrument heated its filter and lost semivolatile material. That question is usually more useful than assuming that one of the instruments is defective.

To translate a TEOM printout into a defensible air-quality statement, identify the size-selective inlet, sampling flow, instrument configuration, averaging interval, filter condition and any flags. A number expressed in micrograms per cubic metre is not a direct count of particles. It is a calculated mass concentration for the air volume entering a particular measurement system, subject to its conditioning and recovery characteristics.

This guide explains the actual oscillator, illustrates concentration arithmetic, distinguishes conventional heated TEOM operation from filter dynamics measurement systems, and walks through conflicting readings. All numerical examples are hypothetical teaching examples rather than observed air-quality data, and readers must consult the actual instrument manual and network protocol when deciding whether a reading meets an official reporting requirement.

Start with the chain from outdoor air to the reported number

The inlet is the first measurement decision. A PM2.5 inlet is designed to transmit a defined fraction of particles as a function of aerodynamic diameter at the prescribed flow. It does not make all particles smaller than 2.5 micrometres pass and every larger particle disappear. The cut point and penetration curve depend on an operating configuration. An incorrect total flow or blocked inlet can shift the size selection before the air reaches the sensor.

After size selection, sample air moves along a line toward a small filter supported at the end of a hollow tapered element. A continuously driven oscillation provides a frequency signal. As additional aerosol mass accumulates, the oscillating system responds through its effective inertia. Instrument electronics convert the frequency change into a mass increment using a calibrated relationship. The instrument can then divide that mass increment by a measured or defined sampled air volume to obtain a concentration.

Because the measurement depends on a small mass change, it is tempting to interpret a negative hourly value as physically negative air pollution. That would be a mistake. The net filter mass can decrease when collected substances evaporate or water desorbs, and the signal can also respond to thermal behaviour and measurement noise. A negative short-interval result is a diagnostic clue about the instrument and aerosol, not evidence that the surrounding air contains negative particles.

The words ‘continuous’ and ‘real time’ also need careful translation. A TEOM gives frequent updated readings, but a value may represent an averaging period or instrument algorithm rather than a perfectly instantaneous sample at the minute printed on a screen. A dataset should specify whether each time stamp marks the beginning, centre or end of an interval, and whether displayed values are hourly averages, rolling averages or raw measurement increments.

Why oscillation can reveal mass

A familiar analogy is a musical oscillator whose pitch depends on effective mass and stiffness. In the simplest ideal oscillator, frequency is proportional to the square root of stiffness divided by mass. Increase mass without changing stiffness and the frequency falls. The TEOM’s actual calibrated relationship accounts for its specialised geometry and sensor behaviour; a user must not substitute an ideal-school-laboratory formula for the instrument manufacturer’s calibration algorithm.

The important causal chain is that aerosol accumulates on the filter, the mechanical system’s oscillation shifts, and software infers the incremental deposited mass. That is fundamentally different from estimating particle mass using the intensity of scattered light. Optical particle measurements must rely on assumptions about size, refractive index, density and shape, whereas TEOM directly senses deposited mass, although transfer and volatile-loss uncertainties remain.

The instrument is therefore not automatically a measure of every component present outdoors before sampling. Molecules can evaporate from particles, deposit on an inlet surface, react, or change with temperature and humidity. Measurement language should distinguish ‘mass arriving at and remaining on the filter under stated conditions’ from ‘all aerosol mass at the ambient location’. That distinction is especially important in atmospheres with ammonium nitrate or semivolatile organics.

Frequency information is powerful only when paired with integrity checks. If the filter is overloaded, the flow drifts, the instrument is unstable or a filter-change procedure resets the measurement, an apparently neat concentration may not be suitable for publication. The correct practice is to look for the status history and quality flags before drawing an environmental conclusion from a graph.

Work a transparent mass-per-volume example

Imagine a hypothetical TEOM whose valid analysed interval shows that 18 micrograms of particulate mass were added to the filter. If the accepted sampled volume for that exact interval is 1.5 cubic metres, the average mass concentration is 18 micrograms divided by 1.5 cubic metres, or 12 micrograms per cubic metre. The numerator is deposited mass for the interval; the denominator is the corresponding air volume. Mixing a one-hour mass increment with a two-hour air volume would silently halve the answer.

Now suppose a log reports 0.9 cubic metres per hour through the relevant measurement path and the valid interval is 40 minutes. The expected volume, assuming this is a stable volumetric flow at the specified reference basis, is 0.9 multiplied by 40/60, giving 0.60 cubic metres. For a 7.2-microgram mass gain, the interval concentration is 7.2/0.60, again 12 micrograms per cubic metre. This is arithmetic for interpreting a measurement, not a replacement for the instrument’s own approved algorithm.

There is an important distinction between flow through an inlet and flow through the filter. Some systems use bypass flow or separate paths to maintain correct inlet operation while sampling a defined portion through the oscillating filter. Do not use an inlet total flow in the denominator unless the method establishes it as the air volume that corresponds to the measured filter deposit. In a manual check, start from the channel specified in that instrument’s data dictionary.

Averaging requires equal attention. Suppose the first valid hour has a concentration of 8 micrograms per cubic metre and the next three-hour period averages 20. The proper time-weighted mean over those four hours is (8 times 1 plus 20 times 3) divided by 4, equalling 17 micrograms per cubic metre. The ordinary mean of 8 and 20 is 14, which gives equal importance to intervals of unequal duration and answers a different question.

Why a heated filter can lose part of the aerosol

Heating is often used to reduce disturbances from changing water content. A sample presented at a reasonably stable condition can yield a steadier mechanical mass signal than one exposed to uncontrolled ambient humidity. Yet heat is not a neutral operation for every aerosol constituent. Ammonium nitrate and various organic compounds can repartition from particulate material to the gas phase, reducing the mass ultimately retained on the filter.

The issue is chemical partitioning rather than simple measurement failure. An air mass may contain a mixture of relatively nonvolatile mineral particles, soot, sulfates, nitrates, organics and water. A heated path can change their equilibrium and evaporation rate. The sampled deposit may therefore differ from the particle composition that existed at ambient temperature. A conventional heated TEOM is especially vulnerable when the local aerosol contains a substantial semivolatile fraction.

Published evaluations establish that operating temperature affects the loss of certain semivolatile organic particulate constituents. It would nevertheless be unsound to assign one universal percentage loss to every location or season. Aerosol composition, humidity, inlet residence time, filter loading, temperature and sampling arrangement all matter. A site with mainly coarse mineral dust need not behave like a cool, nitrate-rich urban atmosphere.

When a TEOM reads lower than a colocated gravimetric sampler, list the plausible mechanisms rather than announcing ‘TEOM underestimates by X per cent’. The comparator may itself lose volatile constituents during handling, pick up adsorption artefacts or use a different sampling and conditioning protocol. A robust analysis compares methods, time coverage, quality controls and meteorology before ascribing the difference to chemical volatilisation.

Understand the filter dynamics measurement approach

Filter dynamics measurement systems, commonly abbreviated FDMS in TEOM contexts, were developed to better account for changes associated with semivolatile material. Their switching or reference measurements estimate filter-mass dynamics that can be missed by a continuously heated collection signal. This does not mean an FDMS simply adds a fixed correction to every hour of data; its correction arises from a configured instrument sequence and algorithm.

A practical reader should seek the instrument’s channel definitions. The reported concentration may combine a base measurement and a reference-related component, and the latter can be positive or negative according to convention. Avoid treating a negative reference term as a negative ambient concentration by itself. The meaning of the sign has to come from the method documentation, not a guess based on whether the final concentration looks reasonable.

In a hypothetical comparison, a conventional channel indicates 28 micrograms per cubic metre while a properly configured corrected channel indicates 36. The absolute difference is 8 micrograms per cubic metre. Relative to the corrected channel, the difference is 8/36, or approximately 22.2 per cent. Relative to the conventional channel, it is 8/28, or approximately 28.6 per cent. Both calculations are arithmetically valid, but they use different denominators and must be labelled.

The corrected result is not beyond criticism simply because it is numerically larger. Its reference periods, flow rates, maintenance, filter behaviour and time alignment still require examination. A large difference might reflect semivolatile chemistry, measurement instability or an instrument setting. An analyst should explain what was compared, why the correction was adopted and which flags were screened before presenting a single final estimate.

Read a disagreement with another particulate instrument

Imagine a daily chart where a heated TEOM average is 19 micrograms per cubic metre, an optical sensor’s humidity-adjusted estimate is 27, and a filter reference sample reports 25. It is premature to declare either 19 or 27 ‘wrong’. First align the day boundaries; next confirm that all three measurements truly represent PM2.5; then examine outages, flow control, zero/span or calibration documentation, conditioning and relative humidity.

The optical estimate is not an independent gravimetric truth. Light scattering responds to particle optical properties as well as number and size. An episode dominated by water-rich particles can cause an optical signal to grow without a proportional increase in dried particle mass. Meanwhile the TEOM may lose some semivolatile material when heated. These effects can act in opposite directions, making an apparent gap larger than a naïve side-by-side comparison suggests.

A valid comparison also needs enough overlapping time. If a filter sampler ran from midnight to midnight but the TEOM lost three hours during a high-pollution episode, an unqualified daily mean from remaining hours answers a narrower question. A network may apply a formal completeness threshold. Readers should record hours observed, hours missing and whether the instrument’s average was reconstructed, suppressed or qualified.

Colocation matters because spatial gradients and local activity can be substantial near roads, construction or exhaust outlets. A monitor a few streets away is not strictly the same sampled air. An instrument comparison should document the separation distance and siting conditions. If the monitors are far apart, the correct language may be ‘nearby observations differ’ rather than ‘measurement methods disagree’.

Troubleshoot negative values and abrupt steps

A sudden downward jump following a routine filter exchange might be an expected change in filter loading and baseline, but it should not be silently treated as an ambient pollution collapse. The maintenance log identifies whether the apparent step aligns with a filter replacement, instrument reset, flow adjustment or operator intervention. Valid concentration output should reflect the method’s handling of such events rather than raw cumulative filter mass.

A brief negative interval at a humid-to-dry transition could be consistent with loss of retained water or semivolatile compounds, but alternative explanations include electrical noise or unstable temperature. Check whether other stations saw a comparable ambient trend, whether the reference channel changed, and whether the relevant status flags marked a problem. Negative values may be retained for some diagnostic analyses even where formal published products apply screening rules.

An extended flatline deserves scrutiny. It may represent genuinely stable concentrations, but unchanging digits over many intervals can also reveal frozen telemetry, a stopped sample flow or rounding. Inspect the number of decimal places, the flow trace and timestamps. Verify that data points update independently rather than repeating the last valid record whenever communication fails.

A sharp positive spike has multiple interpretations. It may be a real smoke or dust plume, filter contamination during servicing, or a flow transient affecting volume normalisation. Time-match the signal with weather, nearby activities, instrument warnings and any colocated measurements. A plume hypothesis is stronger when independent evidence agrees, but a single anomalous peak is not enough to identify its chemical source.

Distinguish quantity, quality and compliance

Mass concentration, data quality and regulatory status are three separate questions. The TEOM reports a value under defined instrumental conditions. Data quality asks whether the value passed controls and is representative of the stated interval. Compliance asks whether a specific jurisdiction recognises the measurement method and how its legally defined averaging period and completeness requirements apply. A perfectly functioning research instrument is not automatically a regulatory reference method.

A PM2.5 value of 22 micrograms per cubic metre over twenty minutes is not directly comparable to a legal twenty-four-hour standard without the prescribed computation. Neither does a high value reveal toxicity or source composition by itself. Different particles can carry different substances, and exposure depends on duration and where people actually breathe. The monitoring record should be interpreted within its designed scientific question.

Measurement uncertainty is not an invitation to ignore the result. It is a reason to state the uncertainty sources explicitly: size selection, flow, mass calibration, ambient conditioning, semivolatile loss, interval completeness and data processing. Some factors create biases that are systematic rather than random. Averaging many biased hours may reduce random scatter without eliminating a heating-related loss.

When preparing a report for the public, use phrases with measured meanings. ‘The corrected one-hour PM2.5 estimate was 36 micrograms per cubic metre during this example interval’ is specific. ‘Pollution was 36’ leaves the unit, pollutant, method and time window hidden. Precision of language prevents a technically correct calculation from becoming a misleading health or regulatory statement.

Apply the method to a second worked case

Consider a sequence of three valid intervals. The first is thirty minutes at 10 micrograms per cubic metre; the second is ninety minutes at 18; the third is sixty minutes at 14. The weighted mean for the full three hours is (0.5 times 10 plus 1.5 times 18 plus 1 times 14) divided by 3, or 46/3, approximately 15.33 micrograms per cubic metre. The arithmetic explicitly preserves duration.

Now assume the ninety-minute interval was actually interrupted by a twenty-minute flow error. If the instrument provides no independently valid replacement measurements for those missing minutes, do not present the earlier 15.33 as an unquestioned complete mean. Recalculate from accepted subintervals if they are available and label the missing portion. If only whole-interval means are stored, determine whether the interrupted interval must be rejected according to the relevant quality protocol.

An analyst may receive mass data as nanograms and volumetric flow in litres per minute. For example, 6,000 nanograms is 6 micrograms. A measured 10 litres per minute for 30 minutes corresponds to 300 litres or 0.300 cubic metres. The hypothetical deposited-mass concentration is then 6/0.300, or 20 micrograms per cubic metre. This exercise exposes two frequent errors: failing to convert nanograms to micrograms and failing to convert litres to cubic metres.

The last check is dimensional. A mass concentration must end in a mass unit divided by a volume unit. If the intermediate answer is ’20 micrograms per minute’, the workflow has mistakenly used time where volume belongs. Write the units through every multiplication and division rather than attaching units only at the end. That habit transfers directly to reading water, soil and laboratory concentration reports.

Reader transfer and independent checks

A site recorded a 24-microgram gain during a valid two-hour period with 0.8 cubic metres of sampled air per hour. The corresponding sampled volume is 1.6 cubic metres, giving 15 micrograms per cubic metre. If another person divides 24 by two and calls the answer 12 micrograms per cubic metre, their denominator is hours, not air volume. The corrected calculation shows exactly why their number is not a concentration.

A second reader reports that an FDMS-derived value of 40 is 25 per cent higher than a conventional TEOM value of 30. This is correct when the comparison is expressed relative to 40? No: the increase relative to the conventional 30 is (40 minus 30)/30, or 33.3 per cent. The missing fraction relative to the corrected 40 is 10/40, or 25 per cent. Always name the reference denominator.

A third reader sees a negative 15-minute value and concludes that fresh air removed suspended particles from the filter. A safer interpretation is that the net measurement mass changed over that period; filter volatilisation, water loss, stability and diagnostic flags must be evaluated. The sampler does not observe every individual particle leaving the surrounding atmosphere, and a negative net filter change cannot describe a negative atmospheric particle concentration.

A fourth reader compares a sensor labelled PM10 with a TEOM labelled PM2.5. Even if both are perfectly calibrated, they measure different size fractions. The first can include larger airborne particles excluded by the second. An exact numerical match is neither required nor expected, and a divergence does not diagnose sensor failure until pollutant definitions and instrument configuration match.

Frequently asked questions and concluding interpretation

Does a TEOM count individual particles? No. Its characteristic measurement is deposited particulate mass inferred from mechanical oscillation. Two air samples can have similar particle counts but different mass because particle sizes and densities differ. Conversely, a few larger particles can contribute substantial mass relative to many ultrafine particles. Particle number and mass are complementary, not interchangeable, measurements.

Is an FDMS-corrected TEOM always more accurate? Not under every condition without qualification. It seeks to address losses and filter dynamics that may affect conventional readings, but its success depends on the complete measurement configuration, calibration, maintenance and aerosol properties. Independent comparison with an appropriate reference and review of uncertainty remain necessary.

Can one infer the source of pollution from a high TEOM value? The time series can suggest when a particle-mass episode occurred, not uniquely identify its source. Source attribution may require weather, spatial comparisons, chemical speciation, dispersion analysis and evidence about emitting activities. A single mass trace cannot reliably distinguish a nearby diesel plume from regional secondary aerosol or other mixtures.

Why does a higher instrument temperature sometimes give a lower mass reading? A higher temperature may help stabilise water effects while driving some semivolatile material from the collected particle phase. The effect varies with composition and exposure conditions. Do not translate a lower reported concentration into ‘cleaner air’ before checking whether the sampling system altered the aerosol.

What should accompany a public chart? Identify the pollutant and size fraction, instrument model and configuration, local time zone, averaging window, units, missing-data treatment, calibration and quality flags. When comparing stations or technologies, explain colocation, timing, flow and volatile-loss limitations. These contextual details turn a graph into evidence that a reader can actually scrutinise.

The useful habit is to translate each observed number backward through the instrument. Ask what air entered, what fraction reached the filter, what mass remained, which oscillator signal changed, what volume and time the calculation used, and which checks establish that the interval is valid. Only then interpret the environmental meaning. This sequence prevents precision-looking instrument output from concealing a very ordinary denominator or conditioning error.

Measurement geometry, inlet losses and routine maintenance

Consider the chain of physical objects between ambient air and deposited mass. The size-selective inlet is an aerodynamic classifier, not a chemically inert portal. Large droplets, insects, dust and liquid water may interact with its surfaces. The inlet requires a prescribed flow and maintenance condition; accumulated material can change performance long before it completely blocks. A report that reproduces the precise model name but ignores inlet cleanliness has not yet documented a reliable PM2.5 measurement.

Tubing introduces another stage. Longer paths and sharp bends can alter particle penetration, especially for particles subject to inertial deposition or diffusion. A system may be designed around particular lengths, internal diameters and thermal conditions. Field operators should not casually improvise a hose extension because the box still displays concentration. The exposure time in a heated line also affects how much volatile material has an opportunity to partition into the gas phase.

At the end of this path sits the collecting filter. Its mass can include deposited particles and absorbed or adsorbed material, but sampling and conditioning may change the composition of that deposit. A nominally identical filter change can shift mechanical baseline and response. When reporting trends across a maintenance event, the analyst should note whether the instrument automatically handled the transition and whether the hours around it met validity requirements.

Calibration deserves a more exact vocabulary than ‘we calibrated it last month’. Instrument checks may independently evaluate flow, mass-response calibration, temperature, pressure or internal electronics. A successful flow check does not prove the oscillator mass relationship, and a mass calibration does not prove size selection at the inlet. Each documented check reduces a different uncertainty. Readers should match the reported claim to the specific thing actually tested.

Zero response and span checks also require interpretation. A zero-air challenge can reveal apparent particulate signal in the absence of incoming aerosol, while an instrument-specific mass verification examines sensor response. These procedures differ between designs and should be drawn from approved manufacturer and monitoring-network instructions. A nonzero response under a zero test is a flag to investigate; it is not automatically an amount that can be subtracted from every outdoor reading.

A site visit should preserve a clear chronology: identify instrument serial number, filter identifier, start and stop of servicing, before-and-after flow checks, alarms, set points, and operator notes. If software supplies an automatic quality code, retain the original as well as any later human flag. Deleting inconvenient records without an audit trail makes it impossible for another analyst to distinguish legitimate correction from selective removal.

A chemical scenario where the missing mass matters

Imagine a hypothetical cool morning during which ammonium nitrate makes a considerable contribution to airborne fine particulate matter. This material reflects an equilibrium involving ammonia and nitric acid in the gas phase and particulate ammonium nitrate in the condensed phase. Heating and removal of gaseous products can encourage the particulate salt to volatilise. What remains on a conventional heated TEOM filter may therefore be less than the ambient particle-phase mass arriving at the inlet.

At midday, a different aerosol mixture dominated by relatively nonvolatile material may show a smaller method difference even at similar ambient PM2.5 mass. A misleading analysis would assign the morning correction percentage to every hour. Instead, pair the time series with temperature, humidity and, where available, composition measurements. The goal is not to force the curves together but to test a plausible physical explanation against independent evidence.

Suppose the conventional TEOM reads 24 micrograms per cubic metre in a morning interval, while a well-validated comparison method reads 34. The 10-microgram difference is around 29.4 per cent of the comparison value. Calling this ‘a 41.7 per cent increase’ instead uses 24 as the denominator. Neither denominator is automatically forbidden, but readers must know whether the question is fraction missing from comparison mass or percentage increase from the smaller signal.

Even perfect arithmetic cannot establish that the entire 10-microgram gap is volatilised nitrate. Differences may arise from organic compounds, filter conditioning, collection efficiency, spatial sampling or reference-method artefacts. A chemical attribution would need species-resolved evidence, such as nitrate and organic measurements collected over matching periods. When evidence is absent, call it an unexplained method difference consistent with several hypotheses.

The seasonality of such a difference is not arbitrary. Temperature and atmospheric chemistry influence gas-particle partitioning, and source mixtures shift across time. A station could see closer agreement in one season and greater divergence in another without either instrument undergoing mechanical damage. One useful analysis is to plot paired difference against temperature and humidity, while remembering that correlation alone is not proof of a specific compound’s evaporation.

Finally, do not confuse water loss with chemical loss. Water bound to hygroscopic particles can change with relative humidity without representing the dry mass defined by a gravimetric reference method. By contrast, volatilised nitrate or organics may be part of the ambient dry particle mass of interest. Instruments and standards handle these processes differently. Label the conditioning basis before interpreting which lost mass is an unwanted artefact.

Design an independent comparison study

A credible comparison starts with a question that can be falsified. For example: ‘Does a heated TEOM read systematically lower than an FDMS channel during high-nitrate periods?’ Decide the averaging interval, co-location criteria, completeness threshold and screening rules before inspecting which result looks favourable. Predefining criteria reduces the temptation to exclude difficult observations merely because they disrupt an expected relationship.

Collect concurrent measurements with compatible time zones and timestamps. If one device stores period-end labels and another stores period-start labels, an apparent lag can be produced by software alone. During a sharply changing smoke or traffic episode, a one-hour offset can reverse the interpretation. Create a short test table containing raw timestamps, interpreted intervals, overlap duration and reported instrument status.

Use paired observations where both instruments have valid data. For each matched interval, calculate difference D = comparison minus TEOM and, if appropriate, a ratio. Summarise median difference as well as the mean, because a small number of extreme episodes can influence the mean. A scatter plot with a one-to-one line reveals whether differences depend on concentration. Bland–Altman-style displays can also expose concentration-dependent disagreement, though proportional or heteroscedastic errors need care.

Avoid collapsing every statistical test into a single correlation coefficient. A correlation of 0.95 can coexist with a fixed 8-microgram bias, and two instruments can share the shape of a pollution episode while disagreeing substantially in level. To quantify agreement, evaluate intercept, slope, dispersion and confidence intervals along with measurement uncertainties. Evidence of a shared trend is not automatically evidence that the reported concentrations are exchangeable.

An example dataset contains four matched hourly readings. TEOM reports 10, 20, 30 and 40 micrograms per cubic metre. Comparison reports 12, 24, 36 and 48. Correlation is perfect in this deliberately simple example, yet the comparison is consistently 20 per cent higher than TEOM. Mean difference is 5 micrograms per cubic metre: differences 2, 4, 6 and 8 average to 5. A reader who says ‘the methods are identical because r equals 1’ has mistaken association for agreement.

Keep concentration units identical. A monitor exporting milligrams per cubic metre gives 0.020 mg/m³ where another prints 20 µg/m³; these figures agree exactly after multiplying by one thousand. A naïve plot of 0.020 against 20 would falsely imply a factor-of-one-thousand discrepancy. Before any comparison, standardise units, volume-reference basis and interval lengths, while retaining each original value for audit.

How to write an uncertainty-conscious conclusion

A good conclusion separates the observations from explanations. ‘The heated TEOM recorded lower paired means than the comparison monitor during the sampled periods’ reports a reproducible difference. ‘This proves nitrate evaporated’ makes an additional chemical claim. To reach the latter, explain which nitrate-specific evidence was gathered, how much of the difference it could account for, and what alternative mechanisms remain possible.

Random uncertainty may shrink when valid independent samples are averaged. A stable systematic bias associated with temperature, however, does not vanish because the study includes thousands of points. Report both where possible. If a paired instrument comparison estimates a bias of minus 4 µg/m³ with substantial uncertainty, resist turning that estimate into a universal correction for every location and aerosol mixture.

Time-coverage uncertainty also matters. Imagine a 24-hour period with 18 accepted hourly measurements. The observed mean describes those 18 hours, not necessarily the missing six. If the omitted intervals were a high-pollution night, the full-day mean might be higher; if maintenance coincided with a clean period, it might be lower. Without an approved imputation protocol, no one can recover the exact full-day average from the remaining observations alone.

An educational report can display a complete worked comparison without implying legal compliance. Give the sample size, overlapping interval definition, descriptive statistics and known limitations. State whether comparisons are exploratory or formally assessed against the relevant method. This protects the reader from thinking that successful colocation automatically qualifies a particular TEOM version for any jurisdiction’s official ambient network.

Communication has a mathematics component too. A 20-per-cent gap in the concentration value does not imply a 20-per-cent change in health risk. Risk is influenced by pollutant composition, duration, population characteristics and other exposures. Similarly, a reduction in the TEOM signal after a maintenance change is not direct evidence that the community breathed cleaner air. Match the language to what the instrument actually measured.

The strongest concluding question is simple: if another laboratory received the raw readings, flow definitions, maintenance history, channel descriptions and calculation steps, could it reproduce the reported number and understand its limitations? That is the real standard of a useful translated data narrative, whether the reader is a field technician, student, researcher or member of the public.