Source Attribution - Stage 5

01 Perform source apportionment using online mass spectrometry data to get time-resolved information on sources 

The data used here were collected during an integrated field campaign conducted over several weeks to improve understanding of specific emission sources, such as assessing the influence of crop burning on air quality. Check the resources below to understand how to perform receptor modelling on high temporal resolution mass spectrometry data (Atabakhsh et al., 2023; see also AMS Spectral Database (Unit Mass Resolution); PMF-AMS Analysis Guide; Protocol to Perform Source Apportionment on ACSM Data Using SoFi Pro). Follow the guidelines outlined in Stages 2 and 3 of the Source Attribution Guidance to determine the optimum number of factors. The identified factor profiles (e.g. traffic-related or biomass burning) can be validated by comparing them with published results. For example, the hydrocarbon-like organic aerosol (HOA) factor, which represents traffic-related emissions, can be evaluated against NOx, a well-known tracer for traffic emissions. While factors related to biomass burning organic aerosol (BBOA) can be validated based on their temporal evolution and their association with other biomass burning markers. Combining these results with back trajectories could also help to understand the origin of these sources, as well as their chemical characteristics.

The detailed information obtained on sources, along with how their contributions change over time, provides a strong evidence base for policy and decision-making. It enables policymakers to design short-term, targeted, and actionable measures, such as traffic management or restrictions on biomass burning or industrial emissions, based on real-time source behaviour.  

02 Combine high-temporal resolution mass spectrometry data with other measurements and perform PMF

Combining data can resolve many sources, which is especially important during pollution episodes to investigate sources and is helpful in pollution management plans. Follow recent studies such as Shukla et al., 2023 and Manousakas et al., 2026, to combine data from various advanced aerosol instruments (e.g. AMS/ ACSM with Xact (metals) and Aethalometer (BC). This approach provides a more complete picture of aerosol composition and improves source apportionment by reducing uncertainties and separating sources that are often missed when analysed individually. It is particularly useful for identifying event-based emissions, such as fireworks, which are often missed or misclassified in routine measurements. In such cases, the combined chemical signatures allow clearer identification and quantification of their contribution. In terms of policy making, this approach provides more reliable source attribution, which supports targeted and timely mitigation measures rather than relying on general reductions.   

03 Advance receptor modelling using extended (secondary markers) offline data

As described in Stage 5 of the Air Quality Monitoring Guidance, offline data on secondary markers (e.g. products formed from the oxidation of aromatics emitted by various anthropogenic sources or oxidation products of VOCs emitted by trees) are now also available. These can support advanced source apportionment studies by helping to distinguish between primary and secondary contributions. Incorporating this information into receptor modelling can improve decision-making by distinguishing between local sources that can be directly controlled from those influenced by regional transport (He et al., 2025). It also improves understanding of secondary formation, which cannot be controlled directly but can be mitigated by reducing precursor emissions such as VOCs through appropriate emission control strategies. 

04 Use source apportionment analysis to identify long-range transport 

To identify the influence of long-range transport, source attribution results from Step 2 and Step 4 above should be combined with meteorological data, particularly wind direction and speed. High time-resolution source information (e.g. hourly) is especially valuable, as it allows source contributions to be linked directly to specific air masses, which can provide a clear picture of how regional or transboundary air masses influence local concentrations. In the absence of such measurements, source apportionment results derived from filter-based measurements can still be used effectively, although you would not have high temporal information on source behaviour.

In both cases, you should focus on secondary inorganic components, typically characterised by ammonium, nitrate, and sulphate, or on sources whose profiles contain secondary organic species, which are often associated with aged and regionally transported pollution (Velásquez-García et al., 2024; Poulain et al., 2021). Elevated contributions from these sources, especially when air masses indicate transport from outside your jurisdiction, can be used as an indicator of long-range transport influences. These results can be shared with other regions or neighbouring nations, depending on the origin of emission sources. This approach complements the airshed approach discussed in the Legal Framework and Policy Design Guidance (Stage 4, Step 6) and supports evidence-based discussions on regional collaboration and emission control for managing transboundary air pollution. 

05 Perform source apportionment on data collected using mobile ground -based measurements

The mobile monitoring approach (see Air Quality Monitoring Guidance, Stage 5, Step 6) offers a rapid assessment of air quality issues, such as traffic-related emissions or wood burning, and enables the evaluation of policy effectiveness. By providing high spatial resolution measurements, it complements the reference monitoring stations, enhances spatial coverage, and helps identify local pollution sources that may otherwise remain unobserved (Bauer et al., 2026; Kolb et al., 2004; Zhang et al., 2022). Importantly, it supports policy interventions by enabling the identification of pollution hotspots, assessing the real-world impact of traffic management or emission control measures, and providing evidence to design targeted, location-specific mitigation strategies. For example, BC data collected through mobile monitoring can be used for source apportionment to separate traffic and biomass burning contributions, following the protocols discussed in Stage 4, Step 4 of the Source Attribution Guidance. This highly spatially resolved information on traffic and biomass burning emissions can then be combined with tools such as ArcGIS to create detailed maps of a city. These maps can highlight source-specific hotspots, allowing targeted interventions to be implemented to improve air quality in those areas. 

06 Carry out source apportionment using PM physical property data 

By this stage, long-term data on particle size distributions and spectrally resolved light absorption at different wavelengths (measured using the same instrument used for BC data) are available. These parameters are typically measured using instruments commonly employed by regulatory authorities and environmental protection agencies, making them readily available across most monitoring stations. Using a new source apportionment approach called RASPBERRY (Real-time Aerosol Source apportionment using Physics-Based Experimental data and multivaRiate factoR analYsis) (Diemoz et al., 2026), which relies only on PM physical measurements, offers a more cost-effective way to investigate PM sources. In this approach, receptor modelling was applied to input data collected using reference-grade optical aerosol size spectrometers and light absorption-based black carbon monitors (measuring BC mass concentrations). These instruments are relatively low cost compared to online/offline chemical speciation monitors (typically between USD 30k-50k, depending on the configurations) and are already widely deployed within regulatory air quality monitoring networks. Using this data, the approach avoids the need for expensive and labour-intensive chemical analyses, reducing both instrument and operational costs, particularly when resources are limited. This makes the approach especially useful for air quality management, as it can provide rapid insights into pollution sources (especially biomass vs traffic) and help inform more timely and targeted mitigation strategies. While RASPBERRY represents a cost-effective alternative to traditional source apportionment approaches, it is still an emerging methodology. Although the initial results are encouraging, further testing and validation are needed to assess its robustness across various different environments before it can be considered a routine approach. 

07  Establish relationship with PM oxidative potential and find sources 

PM Oxidative Potential (OP) was recently included in the revised EU directive on ambient air quality alongside other “pollutants of emerging concern” such as BC and NH3 (see Air Quality Monitoring Guidance, Stage 5, Step 4). PM OP data obtained from your monitoring network can be combined with the chemical composition of PM, following approaches such as those described in Camman et al., 2024 and Daellenbach et al., 2020, to identify the sources contributing to oxidative stress in lungs following exposure. Such analysis provides valuable insights into the health relevance of different emission sources and supports evidence-based policymaking.  

Additionally, these toxicity indicators represent emerging approaches for better characterising health impacts. However, research in this field has not yet advanced to the point where risk can be reliably attributed to individual chemical species. While there is suggestive evidence that some species may be more toxic than others in certain contexts, the current knowledge is still very limited for drawing definitive conclusions. Therefore, the relationship between OP and emission sources should be carefully evaluated before any mitigation measures are implemented. This can be further strengthened through integration with epidemiological studies that investigate associations between source-specific pollution and health outcomes, including mortality. 

08 Integrate PM oxidative potential values into Chemical Transport Models

Integrating source-specific oxidative potential (OP) values derived from field measurements into Chemical Transport Models (CTMs) offers a powerful approach to comprehensively map the spatial and temporal distribution of PM OP across large regions (Pekel et al., 2025; Vida et al., 2025). By linking detailed observational data with modelling frameworks, this approach allows for a more nuanced understanding of how different emission sources contribute to health-relevant air pollution. To apply this approach, the accuracy of emission inventories used within CTMs, along with information on source-specific OP measurements, is highly important. This approach has the potential to provide a promising pathway toward more targeted, health-oriented air quality management strategies.