Frequency-Domain clustering of Non-Attainment cities of India using CEEMDAN decomposition of aerosols and Coupled meteorological drivers
Year: 2026
Type: article
Abstract: Understanding the temporal variability of aerosols and meteorological drivers is essential for assessing urban air quality dynamics across diverse climatic regimes. This study applies an Earth observation–driven frequency-domain framework to 131 non-attainment cities under India’s National Clean Air Programme (NCAP). Aerosol Optical Depth (MODIS) and meteorological parameters (ERA5: column water vapor, land surface temperature, planetary boundary layer height, and vertical wind shear) were decomposed using Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN). Normalized intrinsic mode energy was used to construct multi-parametric feature vectors that were clustered to identify coherent regimes of atmospheric variability. Distinct seasonal groupings emerged: during winter, over 90% of cities exhibited clusters dominated by high-frequency local emissions or persistent synoptic constraints, whereas pre-monsoon and monsoon periods reflected dust and convective influences, respectively. Post-monsoon clusters were characterized by shallow boundary layers and biomass burning effects. The framework transforms EO-derived variability signals into interpretable spatial clusters, bridging local emission variability with large-scale meteorological dispersion dynamics. This integrative approach demonstrates how signal decomposition and geospatial clustering of multi-source Earth observation data can inform regional-scale air quality management and dynamic airshed characterization under NCAP.
Author: Dibyendu Dutta
Institution: Indian Space Research Organisation
Language: English
Subfield: Atmospheric Science
Field: Earth and Planetary Sciences
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $3,550
OpenAlex ID: W7130508678