Downscaling Emissions and Chemistry Transport Model Simulations with Multi-Sensor Satellite Data

The past decade has seen a growing interest of using satellite observations to constrain emissions of aerosols and their precursors. However, the spatial resolutions of these satellite observations, such as OMI SO2 and NO2 products, are often coarser than the resolution of typical air quality models. Furthermore, the computational cost also increases as the spatial resolution of inverse model gets finer. Therefore, a practical solution is to conduct the inverse modeling of emissions at global scale with coarser resolution, and then either downscale these top-down emissions for air quality modeling at finer spatial scale or downscale the global model simulations that use the coarse resolution top-down emissions. This study will showcase these two approaches by using multiple satellite data from OMI, OPMS, VIIRS, and TROPOMI. We will show the design of our methods to use VIIRS day-and-night band to downscale both NO2 emissions and GEOS-chem simulated NO2 concentration and evaluate both approaches. We will also compare these approaches with the downscale approach that uses the bottom-up emission inventories at fine spatial resolution. Finally, an outlook is provided regarding the use of future geostationary satellite observations (such as TEMPO) to improve both emissions and air quality model simulations.