IDENTIFICATION OF AIR POLLUTION HOTSPOTS IN CHENNAI METROPOLITAN AREA

Abstract

This thesis analyzes the spatial dynamics of air pollution in the Chennai Metropolitan Area (CMA) from 2003 to 2023 using an integrated satellite remote sensing and GIS-based 1 km grid framework to identify and characterize air pollution hotspots. Motivated by rapid urbanization, industrialization, and a sparse ground-monitoring network, it addresses key gaps: the absence of a grid-based multi-indicator Environmental Quality Index (EQI) for Chennai, limited longitudinal remote-sensing studies, and a lack of empirical evidence on how planned versus unplanned development influences environmental quality. The study aims to identify and spatially characterize air pollution hotspots in the CMA by jointly analyzing CO emission intensity, land surface temperature (day and night), vegetation cover (NDVI), nighttime light radiance, built-up density, and population distribution for three benchmark years: 2003, 2013, and 2023. Landsat, MODIS, VIIRS, DNB, ODIAC CO emissions, GHSL population datasets, and CMDA planning layers are harmonised into a 1 km fishnet and processed using Google Earth Engine and ArcGIS/QGIS. The methodology follows a sequential design: data cleaning and min-max normalization of indicators; Pearson correlation to quantify relationships between urban form and emissions; multiple linear regression to model CO and nocturnal LST as functions of built-up, NDVI, population, nighttime lights, and road density; and construction of a composite Environmental Risk Index (ERI) and categorical stress classes to delineate and rank hotspots. Findings reveal a more than threefold rise in mean grid-level CO emissions, with strong positive associations between CO, built-up density, and nocturnal LST, and an increasingly strong inverse relationship with NDVI, highlighting the buffering role of green cover. Critical hotspots cluster in the Manali–Ennore industrial–port corridor, emerging nodes such as Sriperumbudur and dense transport and commercial corridors, while vegetated peri-urban and wetland grids function as low-risk ecological buffers. Planned growth areas generally exhibit better environmental outcomes than unplanned sprawl, underscoring the planning–environment nexus. The thesis proposes a tiered, spatially explicit framework for Local Area Action Plans and metropolitan-scale strategies aligned with national environmental mandates and CMDA master plans, offering a replicable methodology for other rapidly urbanizing Indian cities.