Probing into the wintertime meteorology and particulate matter (PM_2.5 and PM₁₀) forecast over Delhi

The Indian Institute of Tropical Meteorology (IITM), in partnership with the National Center for Atmospheric Research (NCAR), has developed a high resolution (400 m) air-quality Early Warning System (EWS) for Delhi using an advanced approach of assimilating aerosol optical depth, fire emissions from a space-borne platform, and real-time aerosol observations from in-situ network in WRF-Chem model to produce a 72-h forecast. The present study summarizes the performance of EWS forecast and prevailing meteorological conditions for winter months of 2020–2021. We examined the performance of model-simulated meteorology by comparing against the in-situ observations. The model shows a positive bias in downwelling shortwave radiation (≈ 34 Wm⁻²) and warm bias in near-surface temperature (≈ 3 K). The simulated winds while having realistic magnitudes depict more southwesterly behavior compared to the observations. The model struggles to accurately predict Planetary Boundary Layer Height. The model realistically simulates the relationship between wind and air quality parameters. The air quality forecast from the model is found to be skillful in three different AQI categories, with accuracy > 88% for critical category events. The model also exhibits good statistical performance in predicting AQI, with low mean bias (¡ 8%), high correlation (¿ 0.68), and low variance (NMSE ¡ 0.09).