View all publications

6 Where to Monitor? – Selecting Monitoring Sites

Monitoring sites are established for different reasons based on the monitoring objectives. Improperly located sites will provide unsuitable results for the intended purpose of monitoring and will in turn lead to incorrect decisions.

This chapter provides good-practice guidance on properly siting monitoring equipment. This Guide recommends the use of the Australian / New Zealand siting standard (AS/NZS 3580.1.1:2007, Methods for sampling and analysis of ambient air – Guide to siting air monitoring equipment) when establishing monitoring sites.

6.1 The Australian / New Zealand Siting Standard

The AS/NZS 3580.1.1:2007 is the new siting standard released in August 2007. It has superseded AS2922 – 1987, Ambient air – Guide for the siting of sampling units. The new standard provides useful guidance on siting individual monitoring stations.

AS/NZS 3580.1.1 recommends minimum distances of sites from roads so that the site is not unduly influenced by roadside emissions. It also recommends heights of sample inlets for monitoring specific contaminants at different site classifications. The importance of these considerations is discussed in section 6.3.

6.2 Site classification

This Guide recommends a combination of New Zealand, Australian and USEPA (USGPO, 2008) site classifications. The purpose of the site classification system is to distinguish between different types of monitoring sites and to allow data sets to be compared nationally and internationally.

The site classification should be recorded in the site metadata (see section 8.8). Site classifications are described in Table 3.

Table 3: New Zealand monitoring site classification and scales

Recommended site category Site scale equivalent Typical area
Traffic Peak (metres to 10s of metres) Typically very close to high-traffic-use roads and intersections. Site should be between 2 to 5 m from the roadside.
Industrial Peak (metres to 10s of metres)
Neighbourhood (10s of metres to 0.5 km)

Peak – close to one large point source or fugitive emissions. Typically used for compliance monitoring.

Neighbourhood – with large and varied point source industry emissions and high population density. Such areas may contain heavy commercial and processing industries.

Residential Peak (metres to 10s of metres)
Neighbourhood (0.5 to 10s of kilometres)

Peak – a monitoring site located somewhere not truly representative (so it is not neighbourhood scale) but does not exactly fit the ‘traffic’ or ‘industrial’ peak site descriptions.

Neighbourhood – suburban areas in larger cities with a relatively high population density, but not in the immediate vicinity of congested roads or industry. This category also includes residential areas in smaller towns in rural areas.

Special (site description) Regional (10s to 100s of kilometres) Airsheds that are distinct in their geographical, meteorological and emissions characteristics. Included are the effects of any point sources or urban plumes on the regional air quality. Could include places where natural emissions are significant (eg, Rotorua), in which case the category would be Special (Geothermal).
National National background sites that contribute to the global network; eg, Special (National).

6.3 Considerations for site selection2

Various factors can severely affect the validity of data from a monitoring site. The following issues should be thoroughly considered before selecting a site.

  • Sites with restricted air flows in the vicinity of the sample inlet
    Sites should not be adjacent to walls, buildings or trees that impede or distort the air flow. AS/NZS 3580.1.1 recommends a minimum clear sky angle of 120 degrees. The site must not be under overhanging foliage. In some cases, such as monitoring in urban canyons, these issues may prove difficult to avoid.

  • Sites prone to the effects of absorption and adsorption
    Some building materials may absorb some of the contaminant being monitored. Deposition of PM10 may occur on nearby leafy vegetation.

  • Sites where chemical interference may occur
    It is prudent not to monitor for ozone at roadside sites due to the possible interference of motor vehicle emissions with ozone measurements – unless their effect is what you are wanting to measure. Ammonia from refrigeration plants may interfere with sulphur dioxide monitoring.

  • Sites where physical interference can occur
    Avoid overhead high-voltage cables to prevent electrical interference with sampling equipment.

  • Sites affected by extraneous local emissions
    Do not choose sites where emissions from other sources will affect monitoring (eg, standby generators, unsurfaced roads, nearby boiler).

  • Sites that are vulnerable to vandalism
    This can be a major restriction to the available sites. Sample inlets are often the most vulnerable (since ideally they should be at breathing height to best assess exposure). The fabrication of a wide-gauge steel mesh enclosure with a lockable mesh lid for maintenance can be an effective way around this problem. If space is available, putting up a security fence is another option.

  • Sites that have poor access
    Consideration must be given to easy transportation of calibration cylinders, instrument repair and the movement of other bulky equipment to and from the site. Schools are often ideal sites in terms of security and power supply, but arrangements may be necessary to ensure access when the school is closed. This can pose problems during holidays.

  • Locating a suitable power supply
    This can be another obstacle, particularly with roadside sites. Sometimes it is possible to use a power supply from a nearby streetlight, running armoured cable under the pavement. Public safety is paramount when routing cables.

  • Changes around the monitoring site
    Demolition or construction activities, road works resulting in diversion of traffic or congestion can all have a considerable effect on data. A record of these events should be kept, including the date/s and intensity of the activity. It may be necessary to consider whether data during these periods is valid and representative.

    If the nature of a site becomes permanently changed, it may be necessary to consider a new location (if the site is no longer representative of the monitoring objectives). Another thing to take into account for long-term sites is the presence of small trees close to the site, which could grow to be very large over the period of monitoring.

Site selection for compliance monitoring should follow the Australian / New Zealand standard (AS/NZS 3580.1.1:2007) as far as possible. This is particularly important for regional and neighbourhood sites. A detailed list of site selection considerations for these sites can be found in the AS/NZS standard. Peak sites can deviate from the standards where necessary, but any non-conformity to standard requirements should be clearly detailed in the site metadata.

6.4 Representativeness of monitoring sites

6.4.1 Spatial representativeness

It is particularly important to select a site that is representative of the location being investigated. This can only be done if the objectives of the monitoring have been clearly identified. The site should not be unduly influenced by sources that are not being investigated. For example, it would be preferable not to locate a site intended to monitor PM10 from domestic heating near a busy road (another source of PM10).

The NES for air quality stipulate that monitoring must be undertaken where people may be exposed and where the standard is breached by the greatest margin or most frequently.

As an example, a site located in the centre of a residential area primarily for the purpose of monitoring ‘worst case’ PM10 emissions from domestic heating may be considered representative of exposure of the entire residential area. Alternatively, a busy roadside site for the purpose of monitoring carbon monoxide may be considered representative of public exposure in that area. Table 3 (see section 6.2) identifies spatial scales adapted from USEPA 40 CFR 58, Appendix D.

Existing air quality monitoring data, emissions inventories and atmospheric dispersion modelling are useful tools and should be used wherever possible when determining worst locations. Regional exposure (eg, how many people may be exposed) should also be taken into consideration when determining the ‘worst’ location.

In the absence of dispersion modelling information, additional guidance can be found in AS/NZS 3580.1.1:2007, Methods for sampling and analysis of ambient air – Guide to siting air monitoring equipment, and the US CFR 58, Appendix D.

6.4.2 Temporal representativeness

Ambient air monitoring should also take into account temporal variations in contaminant concentrations. The NES for air quality and AAQG specify time averages for particular contaminants (see section 3.1.1). Meteorological conditions during monitoring should also be considered, as these conditions change with time.

As an example, monitoring benzene continuously for four months (January to April) would not give a temporally representative result if the objective of monitoring is to determine seasonal variations. Monitoring continuously for one month in each of the four seasons would provide better seasonal data.

Recommendation 13: Representativeness of monitoring

The use of existing air quality monitoring data, emissions inventories and atmospheric dispersion modelling is recommended when determining ‘worst’ locations.

Regional exposure should also be taken into consideration when determining the ‘worst’ location.

Ambient air monitoring should take into account temporal variations in contaminant concentrations.

6.5 Further monitoring considerations

6.5.1 Distances from sources

It is recommended that monitoring at peak traffic sites should be no less than 5 metres away from the source. In the case of roadside monitoring, the distance between kerbside and building façade is often less than 5 metres. For this reason monitoring at the building façade is practical, since the monitoring station is less likely to create an obstruction to pedestrians and will represent air quality where people are likely to be exposed for longer periods of time.

Monitoring should not be conducted at road intersections (unless the intersection is a specific objective of the monitoring), but instead mid-way along a block. Air quality at intersections is generally unrepresentative, and may be better or worse than the rest of the road, depending on congestion and air flows.

In contrast, monitoring at neighbourhood and background sites needs to ensure that the effects of roadside emissions are not monitored (otherwise they will become peak sites). AS/NZS3580.1.1:2007 includes a table that recommends minimum distances from roads, specific for each contaminant, and takes account of different road flows.

To determine locations for peak industrial sites it is often necessary to use dispersion modelling. Such modelling would require local meteorological data.

6.5.2 Monitoring for multiple contaminants

Emissions from a source usually include a number of contaminants, and so it is often desirable to house a number of analysers together. The housing should be large enough to carry out routine calibration and basic repairs to the equipment.

The temperature inside instrument enclosures should remain constant (ideally around 25 degrees Celsius or at a temperature specified by the instrument manufacturer) and will require an air-conditioning unit with heat and cool cycles. A constant temperature is important as most analysers are temperature sensitive, with moderate variations in temperature affecting data. Smaller housings suffer temperature drops during winter calibrations, when their doors have to remain open for long periods. Walk-in housings do not suffer these problems but can be harder to locate and represent a higher level of expenditure.

Monitoring for multiple contaminants can help in assessing the validity of peak pollution levels. An unusually high level of PM10 at a roadside site, in association with high levels of NOx and possibly CO, would indicate a genuine event rather than an isolated set of anomalous PM10 values.

Recommendation 14: Instrument enclosure temperatures

The temperature inside instrument enclosures should remain constant (ideally around 25ºC or at any other temperature stated in the instrument’s user manual).

Installation of an air-conditioning unit with heat and cool cycles inside instrument enclosures is recommended to prevent variations in temperature.

6.5.3 Sample inlet considerations

Inlet heights should be sited at the breathing zone in order to represent human exposure. This is often not possible, because inlet heights have to be compromised to avoid the risk of vandalism or to accommodate the specific requirements of particular instruments. AS/NZS 3580.1.1:2007 provides guidance on acceptable inlet heights for different contaminants at different types of sites. Table 2 of this standard should be used to assess the acceptable range of parameters where it is not possible to monitor at, or near, the breathing zone.

Some considerations include:

  • keeping the inlet at least 20 metres away from tree drip-lines

  • ensuring an unrestricted airflow of 270 degrees around the sample inlet (or 180 degrees if at the side of a building)

  • ensuring a clear sky angle of 120 degrees.

6.5.4 Relevant averaging periods

It is important to use the specified averaging periods when reporting for the NES for air quality (such as an eight-hour running mean for CO, one-hour mean for NO2, and 24-hour mean for PM10). The averaging period for the contaminant should be considered when choosing a site. For example, the eight-hour average for CO is based on the known health effects of exposure above the threshold concentration for periods of eight hours or more, so it is appropriate to site the monitoring in locations where people are likely to be exposed to elevated CO over this averaging period.

Schedule 1 of the NES for air quality provides a concise definition of the averaging periods (see Appendix A). It should be noted that the 24-hour mean for PM10 runs from midnight to midnight.

6.6 Long-term viability of sites

Relocating monitoring sites is not desirable if the monitoring objective is to obtain long-term monitoring data for trend analysis. If this cannot be avoided, then it is important to choose an alternative site that will be available in the long term. Areas not subject to future development or change in ownership are potential long-term monitoring sites.

Where equipment must be moved to a new site, efforts should be made to:

  • select a site with similar sources and meteorology

  • have overlapping data between the old and new sites.

At least one year of simultaneous data recording from both sites is recommended. This data will be needed when trends in air quality are assessed over the period of site relocation.

Recommendation 15: Site selection

Regional, neighbourhood and compliance sites should follow the Australian / New Zealand standard (AS/NZS 3580.1.1.2007).

Peak sites can deviate from the standards when necessary.

Whenever site locations do not meet the standard requirement, this should be clearly detailed in the site metadata.

A minimum co-location period of one year is recommended when relocating monitoring sites.


2 Reproduced from AS/NZS 3580.1.1:2007 with the permission of Standards New Zealand under Licence 000713.