Deciphering air quality index through sample entropy: A nonlinear time series analysis.

[Display omitted] • Proposes a surrogate sample entropy-based method to assess air quality index. • Air quality of most polluted state capital in the world, Delhi, is investigated. • Temporal data of the pollutants are subjected to time series analysis. • Unveils the dependence of PM2.5 concentration on humidity and rain. • Phase portrait's complexity and sample entropy increase with pollutants level. Pollution and its impacts on human health have become a crisis in regions with poor air quality index (AQI), which is an indicator of concentrations of pollutants, prompting the United Nations (UN) to set sustainable development goals (SDG). The present study proposes a surrogate sample entropy-based method in tune with UN's SDG, to assess AQI from the time series of any of the pollutants. New Delhi is one of the world's most polluted state capital, with a higher level of particulate matter (PM). The temporal data of the pollutants in New Delhi, recorded in the one-hour interval during the years 2016 and 2017, are subjected to time series analysis. The data collected from the Central Pollution Control Board of India are analyzed with special reference to PM and compared with the World Air Quality Report 2021 and University of Washington data. The dependence of PM2.5 concentration on humidity and rain is also studied. The study reveals the increase in complexity with the concentration of pollutants through the phase portrait. The sample entropy analysis of the nonlinear time series of the pollutants exhibits a linear relation with AQI suggesting the possibility of using sample entropy as a surrogate measure of AQI. [ABSTRACT FROM AUTHOR]