Measurement report: Unravelling the Mechanisms and Chemical Drivers of New Particle Formation in Chennai, a Tropical Coastal Megacity in India Influenced by an Urban Forest

Devaprasad, Mangattayil; Sathyan, Amrutha M.; Joshi, Tanmay Kishor; Dey, Supriya; Malavika, Geethanjali Rajendran; Sapkota, Yougal; Salim, Rizana; Srivastava, Shailina; Amin, Insha; Pandithurai, Govindan; Raghunathan, Ravikrishna; Liu, Pengfei; Gunthe, Sachin S.

doi:10.5194/egusphere-2026-4244

Preprints

https://doi.org/10.5194/egusphere-2026-4244

Preprints

31 Jul 2026

| 31 Jul 2026

Measurement report: Unravelling the Mechanisms and Chemical Drivers of New Particle Formation in Chennai, a Tropical Coastal Megacity in India Influenced by an Urban Forest

Mangattayil Devaprasad, Amrutha M. Sathyan, Tanmay Kishor Joshi, Supriya Dey, Geethanjali Rajendran Malavika, Yougal Sapkota, Rizana Salim, Shailina Srivastava, Insha Amin, Govindan Pandithurai, Ravikrishna Raghunathan, Pengfei Liu, and Sachin S. Gunthe

Abstract. Chemically resolved gas- and particle-phase measurements capable of constraining new particle formation (NPF) remain scarce in South Asia. Here, we investigate the precursors driving NPF in a coastal megacity and examine why only a subset of nucleation events evolves into sustained particle growth. We present the first comprehensive 20-day field study in a Bay of Bengal coastal megacity, integrating aerosol microphysical, chemical, gas-phase, and VOC measurements with positive matrix factorization (PMF). Photochemically driven formation of 10–30 nm particles occurred between 07:00 and 11:00 IST, whereas sustained growth into the accumulation mode occurred only 2 of 20 days, with growth rates of 4.53 and 3.43 nm h⁻¹. The particle formation rate, condensation sink (CS), and coagulation sink did not differ between growth and non-growth days, indicating that neither formation intensity nor scavenging distinguished the two regimes. Instead, sustained growth occurred when northerly continental air masses prevailed, whereas growth ceased when a strong sea breeze shifted transport to marine air influenced by power-plant emissions. PMF resolved six factors, including an NPF factor dominated by 10–30 nm particles that was associated with biogenic VOCs. Its invariance between weekdays and weekends, and negative correlation with vehicular traffic factor, indicates that NPF was driven primarily by photo-oxidation of biogenic precursors, rather than by anthropogenic activity. These findings suggest that, in coastal cities, precursor availability and air-mass origin may exert a stronger influence on particle growth than CS, although long-term observations are needed to establish its broader applicability.

Received: 16 Jul 2026 – Discussion started: 31 Jul 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Preprint (PDF, 1659 KB)

Supplement (3401 KB)

Download & links

Status: final response (author comments only)

RC1:
'Comment on egusphere-2026-4244', Anonymous Referee #1, 25 Aug 2026
The manuscript presents a detailed observational study of new particle formation (NPF) and subsequent particle growth at a tropical coastal megacity in Chennai, India. The study combines particle size distributions, aerosol chemical composition, trace gases, VOCs, meteorological observations, back trajectories, and PMF analysis to investigate the mechanisms controlling NPF and particle growth. The results indicate that 10–30 nm particle formation occurred frequently, whereas sustained growth was observed on only 2 of the 20 campaign days. The authors attribute the occurrence of sustained growth primarily to the continuity of biogenic precursor supply associated with continental air masses and sea-breeze dynamics.
The manuscript is generally well organized and provides an interesting dataset. However, I recommend major revision before publication, mainly because some of the mechanistic conclusions appear stronger than can be supported by the limited duration of the campaign and by the absence of direct measurements of key nucleation and growth precursors. My main comments are:
Abstract: please indicate explicitly that the study is based on a 20-day campaign and that only two complete growth events were observed. This limitation is important for interpreting the generality of the proposed mechanism.

L197–208: boundary-layer height, solar radiation and precipitation are obtained from ERA5 rather than measured in situ. Please quantify or discuss the uncertainty introduced by using the nearest ERA5 grid point, particularly for boundary-layer dynamics and precipitation at an urban coastal site where strong spatial variability may occur.

L211–220: the H₂SO₄ concentration is estimated using a proxy because direct measurements were unavailable. Since the manuscript uses the H₂SO₄ proxy to argue that sulfuric acid availability does not distinguish growth from non-growth days, please provide a sensitivity/uncertainty assessment of this proxy and moderate the mechanistic interpretation accordingly.

L369–386: the conclusion that the morning 10–30 nm enhancement is predominantly secondary photochemical formation is well supported by the tracer-normalized analysis. However, the occurrence of two pre-dawn episodes associated with traffic emissions indicates that primary ultrafine particle emissions can also be important. Please clarify how these episodes are treated in the overall NPF classification and whether excluding them affects the conclusions.

L407–466: the central conclusion is that precursor-supply continuity, rather than condensation/coagulation sinks, controls sustained particle growth. However, this conclusion is based on only two growth days versus 18 non-growth days. Please provide additional statistical analysis of the differences between the two groups and clearly distinguish correlation from causality.

L420–436: the conclusion that H₂SO₄ availability does not control growth relies on a proxy and on the absence of significant differences between growth and non-growth days. Given that direct H₂SO₄ and HOM measurements are unavailable, I recommend moderating statements that exclude sulfuric acid as a possible contributor and instead state that no evidence for enhanced H₂SO₄ was found under the present observational constraints.

L494–557: the PMF analysis provides useful information on the association between NPF and biogenic VOCs. However, PMF identifies statistical covariance rather than uniquely resolving physical sources. Please avoid interpreting the factors as definitive source attribution, particularly for the broad “biogenic/photochemical precursor reservoir” factor containing both biogenic and inorganic species.

L657–682 / Conclusions: the manuscript appropriately acknowledges the main limitations, including the 20-day duration, the 10 nm SMPS detection limit, the lack of direct H₂SO₄, HOM and precipitation measurements, and the limitations of PMF. I recommend emphasizing these limitations more strongly in the conclusions and avoiding extrapolation to tropical coastal megacities more generally until longer, multi-season datasets are available.

The manuscript would be substantially strengthened by a more quantitative treatment of the statistical significance of the growth/non-growth comparison and by moderating the mechanistic conclusions given the limited number of growth events and the absence of direct measurements of H₂SO₄ and HOMs.
Citation: https://doi.org/10.5194/egusphere-2026-4244-RC1
RC2: 'Comment on egusphere-2026-4244', Anonymous Referee #2, 12 Sep 2026

The presented dataset is an intriguing study of secondary particle formation and growth in a large, tropical urban center uniquely surrounded by ocean, preserved forest, and anthropogenically-influenced surface types. The dataset and analyses appear sound and follow literature standards. However, the key conclusions drawn in the manuscript cannot be supported by the data. For this reason, rejection is recommended.

Importantly, the manuscript does not properly address the time between particle nucleation and measurement once the particle has reached a diameter between 10-30 nm. At a growth rate of 4 nm/hr, which falls in the range seen over 20-day campaign, it would take a 20 nm particle (center of 10-30 nm bin) 5 hours to reach its observed size assuming a constant growth rate upon inception. Further, the initial growth rate over the 1-3 nm range in particular is likely lower than that of the 10-30 nm bin due to Kelvin barrier effects. This could place the formation event sometime in the early morning hours, with growth occurring to 10-30 nm sizes over the morning rush hour before measurement at 11:00. The ‘lag’ time may be alluded to in line 545 in the manuscript, but the discussion needs to be concretely presented if this is the case. The key finding of NPF is most associated with photo oxidation of specifically biogenic precursors (and not aerosol scavenging or sulfuric acid availability) is inherently time-dependent, and loses credibility when the event timing is this uncertain. The analysis would benefit from instead framing these findings as physical/chemical conditions promoting initial growth of fresh NPF into the 10-30 nm bin.

Second, the conclusion about the two sustained growth events being the result of continuous air parcel (ie, Lagrangian-style) sampling is technically true, but I am skeptical of the corollary to this statement that the events are not the result of unique CS/CoagS conditions. Although the magnitude of the computed CS/CoagS is high, the absence of the Aitken/accumulation mode is notable. This is especially because the computed CS and CoagS are self-sampling; these values are computed including the 10-30 nm size bin, so high activity in this bin will always produce an elevated CS and CoagS. Re-calculating the CS for 30+ nm particles may reveal differences.

Despite its limitations, the dataset is a nice look into the conditions leading to sustained growth of recent NPF into Aitken and accumulation mode sizes. I recommend the authors continue to make use of HYSPLIT back trajectory analysis to investigate both the fate of the airmass each day around the 12:00 regime change (using HYSPLIT in the forward mode with trajectories initiated every few minutes in time) and the source of the incoming air that replaces it (using reverse trajectories with the same time frequency). It is possible that growth into the Aitken mode is occurring downwind of the sampling site and that Chennai could be a source of secondary aerosol to nearby locations. With this recommendation, I caution against drawing conclusions on what percent of the observed secondary formation events survive into CCN sizes using this dataset.

Citation: https://doi.org/10.5194/egusphere-2026-4244-RC2

Supplement

https://doi.org/10.5194/egusphere-2026-4244-supplement

Data sets

Dataset for atmospheric new particle formation measurements over the IIT Madras campus during May–June 2025 Mangattayil Devaprasad, M. S. Amrutha, T. K. Joshi, Supriya Dey, G. R. Malavika, Yougal Sapkota, Rizana Salim, Shailina Srivastava, Insha Amin, Govindan Pandithurai, Ravikrishna Raghunathan, Pengfei Liu, and Sachin S. Gunthe https://doi.org/10.5281/zenodo.21355747

Viewed

Total article views: 436 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
265	139	32	436	72	32	33

HTML: 265
PDF: 139
XML: 32
Total: 436
Supplement: 72
BibTeX: 32
EndNote: 33

Views and downloads (calculated since 31 Jul 2026)

Month	HTML	PDF	XML	Total
Jul 2026	70	23	5	98
Aug 2026	110	52	7	169
Sep 2026	85	64	20	169

Cumulative views and downloads (calculated since 31 Jul 2026)

Month	HTML	PDF	XML	Total
Jul 2026	70	23	5	98
Aug 2026	110	52	7	169
Sep 2026	85	64	20	169

Viewed (geographical distribution)

Total article views: 399 (including HTML, PDF, and XML) Thereof 399 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 30 Sep 2026

Short summary

We studied how tiny airborne particles form and grow in Chennai using detailed measurements of air pollutants, natural gases released by trees, weather, and particle composition over 20 days. New particles formed almost every morning, but only a few grew large enough to affect clouds and climate. Growth depended mainly on a continuous supply of natural gases carried by the wind, showing that changing air flow can be more important than pollution levels alone.


Total:	0
HTML:	0
PDF:	0
XML:	0