Universal Features in Atmospheric Particulate Matter Dynamics
Suchismita Banerjee, Koyena Ghosh, Urna Basu, Banasri Basu · 2026
Read Universal Features in Atmospheric Particulate Matter Dynamics by Suchismita Banerjee, Koyena Ghosh, Urna Basu, Banasri Basu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-25386.txt.
Introduction
Banerjee and three colleagues study PM2.5 swings in fifty-four Indian cities over six years. Despite varied settings and climates, the swings show shared traits in spread and time patterns. Once slow trends and seasonal parts are out, scaled spread curves of leftover swings fall onto one curve, fit by an exponentially modified Gaussian curve. Scaled leftover series also share decay traits, with matching auto-correlation decay and a similar 1/f fall at the tails. A spare random model explains the shared traits: steady spread, time links, and spectral scaling.
Catalog record
| CREATOR | Suchismita Banerjee, Koyena Ghosh, Urna Basu, Banasri Basu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 48816 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.25386
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Universal Features in Atmospheric Particulate Matter Dynamics Suchismita Banerjee, Koyena Ghosh, Urna Basu, Banasri Basu Abstract We study statistical properties of atmospheric particulate matter fluctuations using six years of daily PM 2.5 concentration data from fifty-four Indian cities. Despite diverse urban settings and heterogeneous climatic conditions, we find that the fluctuations show strikingly universal behaviour in both the distributional properties and temporal dynamics. After removing slow trends and seasonal components, the rescaled probability density functions of the residual fluctuations collapse onto a single curve and are well described by an exponentially modified Gaussian distribution. The rescaled residual time-series for all the cities further exhibit certain robust dynamical features, with similar decay of auto-correlation functions, and power spectral densities displaying a ∼ 1 / f \sim 1/f decay at the tails. Finally, we propose a minimal stochastic model for the residual dynamics, which explains the observed universal features—the stationary distribution, temporal correlation, and spectral scaling. I Introduction Atmospheric particulate matter, particularly fine particles with aerodynamic diameter less than 2.5\,\mu m (PM 2.5 ), exhibits irregular and intermittent fluctuations arising from emissions, chemical transformations, meteorology and turbulent transport 36 . These fluctuations reflect a complex system, where nonlinear interactions across multiple spatial and temporal scales generate variability that defies traditional deterministic or linear statistical descriptions 23 ; 43 ; 19 . Moreover, variability in PM 2.5 time series, manifested in amplitude and temporal structure driven by local structure and meteorology, poses challenges for unified description and motivates the identification of underlying common patterns across disparate systems. From the perspective of nonlinear dynamics and statistical physics, identification of universal features , shared across diverse systems irrespective of underlying microscopic details, provides a powerful framework for understanding complexity in natural systems. Universality and scaling concepts have been successfully applied to a wide range of real-world complex and non-equilibrium systems, including turbulence 11 ; 10 , finance 39 ; 35 , network dynamics 30 , urban systems 4 ; 13 ; 46 , and diverse environmental and geophysical systems 34 ; 9 ; 38 ; 33 ; 1 ; 7 . Equally importantly, universality is central to stochastic models describing noise-driven complex systems, including multiplicative-noise processes, nonlinear Langevin equations, and systems exhibiting heavy-tailed distributions and 1/f -type spectral behavior 12 ; 45 ; 21 . This broad success of universality in characterizing complex non-equilibrium systems raises a natural question: given the substantial heterogeneity in emissions and meteorology, do PM 2.5 fluctuations share common stochastic features that transcend location-specific details? Although environmental research has extensively documented the health and atmospheric impacts of PM 2.5 44 ; 20 ; 5 ; 14 , relatively few studies 31 ; 43 ; 18 ; 37 have examined detailed fluctuation statistics, scaling properties, and potential universality in PM 2.5 dynamics across geographically diverse urban settings. To address this limitation, we investigate whether universality and signatures of scaling emerge in PM 2.5 fluctuations across multiple cities in India. Using six years of daily PM 2.5 concentration data from fifty-four Indian cities, we demonstrate that the empirical probability density functions (PDFs) of the full time series collapse onto a common curve under appropriate rescaling, despite pronounced heterogeneity in emissions and meteorological conditions. To probe the intrinsic stochastic dynamics, we next isolate short-term fluctuations by removing slow seasonal and baseline components from the original time series. The resulting residuals exhibit a striking data collapse: their rescaled PDFs from most cities converge onto a single universal curve , well described by an exponentially modified Gaussian indicative of asymmetric fluctuations and intermittency. We further characterize the temporal structure of the residual fluctuations through auto-correlation functions (ACFs) and power spectral densities (PSDs). The ACFs reveal persistent correlations with a power-law decay (long-range correlations with power-law decay) over short to intermediate lags, while the PSDs display a robust scaling regime S(f)\sim 1/f . This behavior is characteristic of scale-invariant dynamics and is reminiscent of 1/f noise observed in a wide class of complex stochastic systems 12 ; 29 . Finally, we propose a minimal stochastic differential equation (SDE) model that reproduces the observed residual distributional shape, correlation decay, and spectral scaling. The model provides a compact dynamical description of the PM 2.5 fluctuations and links the empirical findings to well-studied universality classes of noise-driven nonlinear systems. These results demonstrate that urban PM 2.5 fluctuations exhibit robust and consistent statistical and dynamical signatures across diverse cities, suggesting the existence of universal features in their underlying stochastic dynamics. More broadly, our findings illustrate how tools from statistical physics and stochastic dynamical systems can uncover the unifying patterns in complex environmental time series. The remainder of the paper is organized as follows: Section II describes the data compilation, processing procedures and the preliminary statistical analysis of the daily PM 2.5 series. In Section III , we investigate the probability distributions of daily PM 2.5 across cities and demonstrate universality through collapse of the rescaled probability density functions (PDFs). Section IV analyses residual fluctuations and their universal features through PDFs, auto-correlation functions, and power spectral densities. In Section V we propose a (an effective) stochastic dynamical modeling framework for residual fluctuations, linking the observed statistics to the underlying dynamics. Finally, Section VI concludes with a summary of the main findings and their implications for air-quality modeling. Figure 1: Map of India indicating the locations of the cities used in this analysis II Data Compilation and Preliminary Analysis Data Compilation : For this study, the data is collected from the Central Pollution Control Board (CPCB) of India 6 . The data set comprises daily average concentrations of PM 2.5 , \rho(t) , recorded at multiple monitoring stations ( N_{\alpha} ) located in various cities throughout India. In total, the data is collected from 187 monitoring stations spread across 54 cities [see Fig. 1 and Table 1 ]. The time span of the dataset ranges from December 2018 to February 2025. Data has been collected based on measurement availability at individual monitoring station, which varies across locations and time periods. Nevertheless, the selected cities provide broad geographic coverage across India and represent areas with a wide range of population sizes. Figure 2: Temporal Variation of daily PM 2.5 concentrations of two representative monitoring stations over the study period. Each color in the background corresponds to a specific level of health concern, with the highest level designated as hazardous 2 . Data Preprocessing : The dataset contained missing values (on average, 6.54% of daily PM 2.5 observations across 54 cities.), outliers, and noisy measurements. To address these issues we employed Adaptive Kalman Filtering (AKF), an enhanced variant of the standard Kalman Filter 15 , for imputation and smoothing of time-series data 1 1 1 Sensitivity analysis with uniformly small root mean square error (RMSE) 22 , indicates robustness of AKF for handling missing values of PM 2.5 . . AKF dynamically updates both the process and measurement noise covariance matrices to accommodate changing data characteristics over time 26 and incorporates Mahalanobis-distance-based outlier detection. This approach smooths the data while preserving its inherent temporal structure and variability of the atmospheric PM 2.5 time series. Temporal Trend and Seasonality : We now analyze the time-series data for PM 2.5 concentrations, \rho(t) . Fig. 2 presents the