In this blog series titled Measuring the Microns, we examine why greater attention to fine particulate matter is essential for improving air quality management in India. This first blog discusses the need to prioritise PM2.5 (particulate matter with aerodynamic diameter ≤ 2.5 µm) within the regulatory framework, explores the role of real-time source apportionment (RTSA) in monitoring PM2.5, and presents insights from a Bengaluru-based case study.
Despite ongoing mitigation efforts, India continues to face severe PM pollution, disrupting economic activities and posing a risk to public health. To address this challenge, the National Clean Air Programme (NCAP), launched in 2019, aims to achieve a 40% reduction in PM10 by 2026. Source apportionment studies tell us about the sources of pollution and their contributions, representing a crucial tool to identify and design appropriate mitigation measures.
The finer fraction of particles, or PM2.5 (aerodynamic diameter ≤ 2.5 µm), is mainly contributed by combustion-related sources and secondary particle formation in the atmosphere. Due to their smaller size, these particles can penetrate deeper into the respiratory system, including the lungs and even the alveoli, making them particularly harmful to humans. Further, owing to their long residence time in the atmosphere, PM2.5 can travel farther and contribute to regional pollution. On the other hand, PM10 includes all particles with aerodynamic diameters ≤ 10 µm and therefore contains PM2.5 in addition to the coarse fraction spanning 2.5–10 µm. Contribution to this coarse fraction is largely associated with mechanical processes, such as road dust resuspension and construction activities, as well as natural sources, such as dust storms, sea salt, and pollen. Strategies focusing on PM₁₀ often emphasise contributions from the coarse fraction, such as those from road dust and construction. However, the relative importance of coarse and fine particles varies significantly across locations and seasons. In many combustion-influenced urban environments, PM₂.₅ tends to dominate and is more strongly associated with adverse health impacts, whereas in dust-prone or coastal regions, coarse particles can contribute substantially to PM₁₀. Therefore, alongside existing PM₁₀-focused strategies, greater emphasis on PM₂.₅ is particularly important for improving air quality management in urban settings.
As PM2.5 comprises a wide range of chemical constituents from multiple sources that show distinct behaviour and properties, it is impractical to measure all components using a single instrument. Thus, comprehensive chemical characterisation of PM2.5 requires an integrated set of advanced instruments. A commonly deployed setup includes instruments for measuring non-refractory PM2.5 species (organics, nitrate [NO3−], sulphate [SO42−], ammonium [NH4+], and chloride [Cl−]), black carbon, and trace metals at high time resolution. Together, these instruments provide continuous insight into the chemical composition of PM2.5 at time resolutions of a few minutes. Such measurement systems are already integral to air quality research networks in the United States (ASCENT), Europe (ACTRIS), and China and are now increasingly being deployed in India. They enable RTSA using receptor models such as positive matrix factorisation, allowing identification and quantification of pollution sources with high temporal resolution in contrast to traditional source apportionment, as discussed in a previous blog.
Newly developed RTSA tools, such as SoFi-RT (Source Finder–Real Time), can transform source apportionment from a retrospective exercise into a near-real-time decision-support tool. SoFi-RT connects to near-real-time PM2.5 chemical composition instruments via a cloud server and performs the source apportionment analysis—work that traditionally takes weeks to months—in just minutes, without requiring user intervention. A key prerequisite for RTSA, however, is the availability of site- and season-specific source profiles, which must be derived from traditional source apportionment conducted on a representative dataset from the location of interest. In other words, real-time intelligence is only as good as the foundational science beneath it. This groundwork is now being laid in Bengaluru by the Center for Study of Science, Technology and Policy (CSTEP).
The Air Quality Observations team at CSTEP, Bengaluru, equipped with advanced instruments for real-time measurements of PM2.5 chemical composition, has been implementing and validating the RTSA setup using SoFi-RT to better understand the sources shaping the city’s air.
CSTEP’s RTSA results from the Bengaluru study site highlight how the balance between PM2.5 sources changes from day to day. On days when PM2.5 concentrations were relatively low (30 Nov. 2025 to 5 Dec. 2025), primary sources such as traffic and local combustion made up a larger percentage share of the pollution. But on high-PM2.5 days (25 Nov. 2025 to 29 Nov. 2025 and 6 Dec. 2025 to 9 Dec. 2025), the picture shifts completely. A major fraction, around 60%–70% of PM2.5, came from secondary sources, with dominant contributions from ammonium sulphate and secondary organics. These species are generally linked to regional-scale processes, associated with transported air masses and atmospheric chemical transformation, rather than local emission sources. This raises an important policy-relevant question: in such situations, would mitigation strategies focusing only on local emission sources be effective?
It is important to note that this analysis did not include metal components, which may contribute a significant fraction of PM2.5 in Bengaluru, particularly from sources such as road dust, brake and tyre wear, and industrial activities. Our Xact metal monitor, which became operational recently, enables metal measurement in real time. Its integration into the RTSA framework will further improve source identification.
In conclusion, as India’s air quality management evolves, RTSA can complement existing frameworks by providing timely insights into the sources contributing to PM₂.₅ pollution. While approaches such as the Graded Response Action Plan are based on air quality index thresholds and predefined actions, RTSA enhances situational awareness by identifying dominant sources at a given time. By translating real-time chemical measurements into near-real-time source information, the insights generated can support more informed decision-making, help interpret pollution episodes, and evaluate interventions. Although not predictive, RTSA provides valuable evidence to guide more targeted and adaptive air quality management over time.
The next blog in this series will discuss the submicron fraction of PM (PM1) and explore how environmental dust monitors enable real-time monitoring of PM1 alongside other particle fractions.
Acknowledgement: We sincerely thank Dr Prakash Doraiswamy for his valuable insights as a reviewer during the preparation and analysis of the blog.
Umer Ali, PhD, Former Senior Associate in the Air Quality Observations team at CSTEP, also co-authored the article.
India’s transition to a cleaner power system has moved from a supply...
In the first blog in ‘The AI shift’ series, we examined what...
With growing electricity demand, renewable energy penetration, and storage requirements, Indian distribution...
In the first blog in ‘The AI shift’ series, we looked into...
Following our photo essay from the field, this blog looks at the...
In this three-part blog series titled ‘The AI shift’, I examine how...
