Impact of post monsoon crop residue burning on PM_2.5 over North India: Optimizing emissions using a high-density in situ surface observation network

Abstract. The impact of post monsoon crop residue burning (CRB) on surface PM_2.5 concentrations over the Punjab–Haryana–Delhi (PHD) region in North India was investigated using a regional meteorology–chemistry model, NHM-Chem, and a high-density in situ surface observation network comprising Compact and Useful PM_2.5 Instrument with Gas Sensors (CUPI-G) stations. We optimized CRB emissions from November 1 to 15, 2022 using NHM-Chem and surface PM_2.5 observational data. The CUPI-G data from Punjab was found to be crucial for CRB emission optimization, as the CRB emissions in North India in October and November are predominantly originating from Punjab, accounting for 80 %. The new emission inventory is referred to as OFEv1.0, with 12 h time resolution, in daytime (5:30–17:30 IST) and nighttime (17:30–5:30 IST). The total emissions in OFEv1.0, such as PM_2.5, organic carbon, and black carbon, were consistent with previous studies, except CO, which was overestimated. OFEv1.0 substantially boosted emissions, which were underestimated in satellite data due to clouds or thick haze on November 8 and 10, 2022. Large differences in optimized daytime and nighttime emissions indicated the importance of diurnal variations. Daytime emissions were larger than nighttime emissions on some days but not on others, indicating that diurnal variation shape may have differed each day. The mean contribution of CRB to surface PM_2.5 over PHD was 30 %–34 %, which increased to 50 %–56 % during plume events that transported pollutants from Punjab, to Haryana, to Delhi. Due to low performance of the meteorological simulation on November 8 and 9, 2022, emission optimization was not successful in the case of increased PM_2.5 concentrations observed in Haryana on these days. The results of this study were obtained using a single transport model. Multi-model analysis is indispensable for better predictions and quantification of uncertainties in prediction results.