Improving air quality predictions in New Delhi during the crop-residue burning season via chemical data assimilation

Air pollution has become one of the most important environmental concerns for Delhi. In order to reduce the exposure of public to polluted air via advanced warnings, the National Center for Atmospheric Research (NCAR) and the Indian Institute for Tropical Meteorology (IITM) have collaboratively developed a high-resolution air quality prediction system for Delhi. The newly developed system assimilates the satellite aerosol optical depth (AOD) retrievals and ground-based observations of particulate matter of aerodynamic diameter smaller than 2.5 mm (PM_2.5) to improve initialization of aerosol particles in the model and provides air quality forecasts for 72 hours in advance. The system was developed and tested during the post-monsoon crop-residue burning period (10 October - 20 November) of 2016 and 2017, and went operational in October 2018. During the development phase, the assimilation of AOD and PM_2.5 was found to reduce biases of more than 200 mg m^-3 in surface PM_2.5 mass concentrations against independent observations. A website has been developed to disseminate these forecasts to the public and decision-making authorities. An analog-based post-processing system is being developed to further improve the accuracy of deterministic forecast and quantify the uncertainties in PM2.5 forecasts. This presentation will discuss in detail the development of the system and its role in enhancing the air quality decision-making activity in New Delhi.