资源环境科技发展态势分析平台

Air quality in India during the winter months is particularly bad, due to the meteorological conditions limiting dispersal of pollutants. However, investigations of this period using regional air quality models have, so far, been limited. Air quality simulations, using the Weather Research and Forecasting with Chemistry (WRF-Chem) model, at a high grid resolution of 12 km × 12 km, have been carried out for the 2015–16 winter period over the Indian subcontinent. Gas and aerosol chemistry are simulated using the CBM-Z and MOSAIC (4-bin) modules. Emissions from the EDGAR-HTAPv2.2 database are used, scaled to the simulation year based on changes in activity data for each Indian state, which increases national emissions of PM_2.5, BC, OC, NOx, SO₂ and NMVOCs by 2.6, 0.3, 0.5, 6, 13.5 and 3.6 Gg/day, respectively, in winter. Model performance was evaluated with respect to ground-based observations of PM_2.5, CO, NO₂ and O₃ from available monitoring stations in the cities Ahmedabad, Bangalore, Chennai, Delhi, Agra, Lucknow, Patna, Mumbai, Kolkata, and Hyderabad. PM_2.5 predictions at most of the monitoring cities fell within the excellent model performance criteria (Mean Fractional Bias range − 0.15 to 0.11). The NO₂ concentrations were reproduced well by the model, with a Mean Fractional Bias range of −0.17 to 0.25. CO concentrations were generally underpredicted, but with relatively smaller biases over northern Indian cities. Ozone was reasonably reproduced at Delhi, but modelled day time peaks are much higher than observations in the other cities (~ 42–62%). Patterns of diurnal cycles of pollutants were found to be broadly similar for both observations and WRF-Chem predictions. The 24 h averaged PM_2.5 mass loading over most Indian cities and states were found to exceed the Indian National Ambient Air Quality Standard (NAAQS). PM_2.5, CO and NO₂ mass loadings were highest in northern and eastern India, particularly over the Indo-Gangetic Plain. Overall in this study, wintertime pollution loading in India is reproduced well by WRF-Chem. However, creating diurnal emission cycles based on local activity data, updated national emission inventories for recent years at higher resolution, and extra air quality monitoring stations, especially in rural areas, are needed to increase the accuracy and validation of model predictions, to better inform policy making.