Impact of chemical initial and lateral boundary conditions on air quality prediction

Accurate representation of initial and boundary conditions of chemical species in numerical models is a major challenge for air quality prediction. Model for Ozone and Related chemical Tracers, version 4 (MOZART-4) is an offline global chemical transport model simulating various gases and particulates in the atmosphere. In this work, MOZART-4 analysis is used to modify the chemical boundary conditions over India to improve the forecast in an online-coupled model. A sensitivity study has been performed using the Weather Research and Forecasting (WRF) model coupled with Chemistry model (WRF-Chem) for the highly polluted winter month of December 2016. Daily two parallel experiments are performed from the WRF-Chem model with and without updating chemical initial and lateral boundary conditions using MOZART-4 analyses. Results show that the tropospheric ozone (O₃) and surface carbon monoxide (CO) forecasts are improved by 6% and 20%, respectively when MOZART-4 analysis is used to update chemical initial and lateral boundary conditions. Moreover, prediction of O₃ and CO is also improved vertically (reduction in root mean square difference by 58% and 26%, respectively) in different forecast lengths. The percentage improvement in surface CO, particulate matter of size less than 2.5 μm (PM_2.5) and particulate matter of size less than 10 μm (PM₁₀) forecasts is ∼22%, ∼4.3% and ∼20%, respectively when compared against ground observations from Central Pollution Control Board (CPCB), New Delhi stations. Overall, we observed improvement in O₃, CO, PM_2.5 and PM₁₀ forecast using WRF-Chem model after initialization of chemical conditions from MOZART-4 analysis.