The quantitative assessment of air pollution induced visibility impairment is a prevalent concern across Indian sub-continent. In present study, efforts have been made to identify thresholds of aerosol (PM2.5), meteorological variables (relative humidity, temperature at 2 m, wind speed at 10 m) and boundary layer for shallow (Category (CAT) I: 550–1000 m), moderate (CAT II: 300–550 m), dense (CAT IIIA: 175–300 m & CAT IIIB: 50–175 m) and very dense (CAT IIIC: <50 m) fog categories. A 5-year dataset from 2018 to 2023 for wintertime (December–February) is used to determine thresholds using statistical methods. A unique two-way rolling window correlation analysis is performed for linear, inverse and logarithmic functions considering described visibility classes, PM2.5, RH and temperature. Different sliding windows and subsequent step sizes are taken to ascertain interrelatedness based on significant correlation coefficients. Further, a frequency distribution based averaging method is used for wind speed and boundary layer thresholds. Based on the findings, identified thresholds for PM2.5, RH, temperature, wind speed and boundary layer are >270 μg/m3, >70%, <13 °C, <1.5 m/s, <100 m for CAT I; >250 μg/m3, >70%, <13 °C, <1.2 m/s, <80 m for CAT II; >250 μg/m3, >70%, <11 °C, calm winds, <80 m for CAT IIIA; >220 μg/m3, >80%, <11 °C, calm winds, <70 m for CAT IIIB while 180 μg/m3, >90%, <9 °C, calm winds and <60 m for CAT IIIC. The determined thresholds have been validated using 2023–2024 data based on which shallow, moderate, dense and very dense categories are found to be 65%, 77%, 88%, 90% and 93% compliant with the thresholds of CAT I, CAT II, CAT IIIA, CAT IIIB and CAT IIIC. Trajectory clustering is also included for ascertaining the potential pollution source regions. The study can aid policymakers in predicting fog events. Moreover, appropriate policy interventions can be formulated in wake of the early warning system and information dissemination.
Read full abstract