Monitoring records a set of indoor air quality parameters continuously, over weeks or months, rather than capturing them once during a visit. The difference is not simply one of quantity. A continuous record shows how a building behaves across occupancy cycles, plant schedules and seasons, which is a different class of information from a snapshot, and it produces that information using sensor technology whose accuracy and stability differ substantially from laboratory methods. This page sets out how monitoring differs from spot testing, what each class of sensor can and cannot be trusted to do, why drift and calibration determine whether a dataset is usable at all, why carbon dioxide is read as an indicator of ventilation rather than as a pollutant, and what a long record reveals that a single test cannot.
A spot test answers what conditions were during a visit. Monitoring answers how conditions vary, which is usually the more useful question when a complaint is intermittent. Few indoor air quality parameters are stable properties of a room: carbon dioxide follows occupancy, particulate concentration follows outdoor episodes and indoor activity, humidity follows the cooling schedule and the weather, and volatile compound concentrations follow temperature, cleaning and use.
Because of that variability, a single visit has a real chance of missing the condition that prompted the investigation. A complaint arising on Monday mornings, in late afternoon on one side of a building, or during a particular plant changeover, is a pattern rather than a value, and patterns are found by sampling repeatedly over time rather than by sampling well once.
Non-dispersive infrared sensors measure carbon dioxide by absorption at a characteristic wavelength. They are the most dependable of the low-cost classes and are generally suitable for tracking ventilation behaviour, provided the device is a true infrared type. Instruments reporting an equivalent carbon dioxide value inferred from a volatile compound sensor are not measuring carbon dioxide at all, and their output should never be presented as though they were.
Optical particle sensors infer mass concentration from light scattered by particles passing through a beam. They respond well to change and are useful for showing when a particle event began and ended, but the conversion from scattered light to mass rests on assumptions about particle density, shape and refractive index that a mixed indoor aerosol does not satisfy. They also read high in humid conditions, because absorbed water makes particles optically larger than they are.
Metal oxide and photoionisation sensors respond to volatile organic compounds as a class. They are sensitive to change and largely unable to say what changed, because the response varies by compound and bears no relation to potency. Read as an indicator that something has altered, they earn their place. Read as a concentration, they overstate what the technology supports.
Every sensor class drifts. Rates and directions differ, but the consequence is the same: a dataset that is never checked slowly stops describing the building and starts describing the instrument. Drift matters more for absolute comparison than for pattern recognition, which is one reason monitoring is stronger at showing shape than at producing a figure to set against a reference value.
Carbon dioxide sensors are commonly fitted with an automatic baseline routine that assumes the lowest reading over a rolling period corresponds to outdoor concentration. In a building genuinely unoccupied at night, that assumption is reasonable. In a continuously occupied space, or one where ventilation stops overnight and concentrations never fall, it is false, and the routine progressively pushes readings downwards until a poorly ventilated space appears well ventilated. Whether the routine is enabled, and whether the space suits it, is worth establishing before any conclusion is drawn.
Co-location is the most practical field check. Placing the deployed sensors together for a period before installation, and again afterwards, shows how far they agree with one another and how far each has moved. Comparison against a calibrated reference instrument during a site visit anchors the whole record. Sensors also have a service life, and one that has passed it produces plausible-looking data that means nothing. None of this makes monitoring unreliable; it makes monitoring a system that requires maintenance, and a deployment without a calibration plan degrades quietly rather than failing visibly.
Carbon dioxide is measured indoors almost entirely for what it indicates rather than for what it does. People exhale it at a broadly predictable rate, so its concentration in an occupied space reflects the balance between how many people are present and how much outdoor air is arriving to dilute what they produce. Read that way it is an indicator of ventilation performance rather than a pollutant measurement, and that distinction changes how a reading is used.
The practical readings follow. A concentration climbing through a morning shows supply failing to keep pace with occupancy. A concentration sitting close to outdoor levels in a busy room shows generous ventilation, or a sensor reading incorrectly. A trace that does not respond to occupancy at all usually indicates a fault or an automatic baseline routine that has taken over the record.
The proxy has limits that matter. It says nothing about contaminants people do not exhale: a space with excellent carbon dioxide figures may still be accumulating compounds from materials or activities, because the same air change rate acts on a different generation rate. It says nothing about filtration, which removes particles without touching carbon dioxide. And it is weak in sparsely occupied large volumes, where dilution by sheer volume holds concentrations down regardless of the outdoor air rate.
Dubai Municipality's indoor air quality guideline lists carbon dioxide at 800 ppm over eight hours in its table for existing buildings. These are the values Dubai Municipality's indoor air quality guideline sets for buildings that opt into its indoor air quality certificate. The guideline describes itself as guidance rather than as a binding standard, and the certificate route is expressly optional. They are a published reference point within a voluntary Dubai Municipality scheme, not an enforceable UAE limit.
The first thing a record exposes is the daily and weekly rhythm. Ventilation plant that starts too late, stops too early or runs at reduced output on particular days is visible immediately in the carbon dioxide trace, and a pattern repeated across many days is more persuasive to those who control the plant than any single reading.
The second is the unoccupied condition. In this climate the interesting excursions often occur at night and at weekends, when cooling is set back: internal temperatures rise, moisture that entered during the day remains in the space and its furnishings, and relative humidity climbs for hours at a time. Those excursions never appear in a working-hours test, and they bear directly on the surface conditions that permit fungal growth.
The third is the response to events. A record spanning an outdoor dust episode, a filter change, a fit-out, a commissioning adjustment or a change in occupancy density shows how the building reacted, which is the nearest available approximation to a controlled test of whether an intervention worked.
Density follows zoning rather than floor area. One device per air handling unit zone is a reasonable starting point, with additional units in areas that have generated complaints. An outdoor reference device earns its place in monitoring for the same reason an outdoor sample does in testing, particularly for particulates, because it separates infiltration from indoor generation across the whole record rather than at a single moment.
Alert thresholds deserve care. A level set too tightly generates constant notifications, and a system that alerts constantly is ignored within a fortnight. Thresholds that account for duration as well as level, and that are revisited once the normal behaviour of the building is understood, survive longer. Someone also has to act on the output: a deployment with no route from a reading to a plant adjustment produces an archive rather than an improvement.
Monitoring does not cover the biological family. Fungal and bacterial burdens are not measurable by any continuous sensor in general use, and establishing them still requires sampling followed by laboratory analysis. What monitoring can do is track the conditions that govern them, and a long humidity record is frequently more useful for understanding fungal risk than an occasional spore count taken on an unremarkable afternoon.
Monitoring is also not a substitute for laboratory-grade analysis where a result is to be compared against a published reference value. On that point Dubai Municipality's Technical Guidelines for Indoor Air Quality for Healthy Life (Doc Ref DM-HSD-GU119-IAQ, Version 4, issue date 11 December 2024) states that "Air Quality testing must be carried out by an air testing accredited company or laboratory by Emirates International Accreditation center (EIAC), and the Compliant test results must be submitted to DM."
The parameters a monitoring deployment typically covers appear in that guideline as follows. Table 2, for existing buildings, gives carbon dioxide of 800 ppm over eight hours, PM2.5 of 35 μg/m³ over twenty-four hours, PM10 of 150 μg/m³ over twenty-four hours and total volatile organic compounds of 0.6 mg/m³ over eight hours. Table 3, covering thermal comfort in new and existing buildings, gives relative humidity of 20–60% with the remark "Never to exceed 56%RH at 80F (27C)", dry bulb temperature of 22.5–25.5 °C and air movement of 0.2–0.3 m/s, and states that the HVAC system "must be capable of providing the following range of conditions for ninety five percent (95%) of the year".
These are the values Dubai Municipality's indoor air quality guideline sets for buildings that opt into its indoor air quality certificate. The guideline describes itself as guidance rather than as a binding standard, and the certificate route is expressly optional. They are a published reference point within a voluntary Dubai Municipality scheme, not an enforceable UAE limit. There is no enforceable UAE exposure limit for mould, and no published Abu Dhabi reference value for it either.
Measures carbon dioxide by absorption at a characteristic wavelength. The most dependable low-cost class, and the basis of any credible ventilation indicator.
Infer mass from scattered light. Good at showing when a particle event started and stopped, weaker on absolute mass, and prone to reading high in humid air.
Respond to volatile organic compounds as a class. Useful as an indicator that something has changed, unable to identify what changed or how potent it is.
Used for carbon monoxide and similar gases. Reasonable performance with a finite service life, because the electrolyte degrades whether or not the cell is exposed.
The most reliable elements in a typical deployment, though humidity sensors drift with age and are damaged by condensation.
A device recording the same parameters outdoors, which separates infiltrated from internally generated fractions across the whole record rather than at one moment.
Monitoring data describes pattern, duration and frequency, and is best used for those purposes. Where a result is to be compared against Dubai Municipality's published reference values, the guideline states that air quality testing is to be carried out by a company or laboratory accredited by the Emirates International Accreditation Centre, with compliant results submitted to Dubai Municipality. There is no enforceable UAE exposure limit for mould, and no published Abu Dhabi reference value for it either.
These are the values Dubai Municipality's indoor air quality guideline sets for buildings that opt into its indoor air quality certificate. The guideline describes itself as guidance rather than as a binding standard, and the certificate route is expressly optional. They are a published reference point within a voluntary Dubai Municipality scheme, not an enforceable UAE limit.
No. Monitoring shows how parameters vary over time using sensors; testing produces a smaller number of higher-quality values, including laboratory analysis for compounds and biological material that no continuous sensor measures.
It is monitored as an indicator rather than as a hazard in its own right. A rising concentration shows that outdoor air supply is not keeping pace with the number of people present, which is the finding that matters.
It depends on the class and the manufacturer's stated interval, but every deployment needs a plan. Co-locating the devices before installation and again later, and anchoring the record against a calibrated reference instrument, are the practical field checks.
Not directly. No continuous sensor in general use measures fungal burden. Monitoring can track the temperature and humidity conditions that allow growth, which is often the more actionable information.
Because they size particles optically. Water absorbed onto a particle makes it scatter light as though it were larger, so the inferred mass rises even though the particle burden has not changed.
Density follows the way the building is zoned rather than its area. One device per air handling unit zone is a reasonable starting point, with extra units where complaints have arisen and one outdoor reference device.