The Spatial Resolution Needed from Models and Satellite-based Measurements to Accurately Predict NOx Emissions

The tropospheric NO2 column, a quantity accessible from space-based measurements, varies in space and time due to emissions, transport and also the non-linear response of OH to NOx concentration. We show that in a simple plume model describing the near field of cities there are steep nonlinearities in (column NO2)/(unit emissions) because of this OH feedback. At spatial scales of 10-50 km such effects are observable in 3-dimensional chemical transport models. We derive the model resolution necessary to infer emissions from space-based observations to 10% accuracy. We show that observations from OMI, which provide daily global coverage with urban to regional-scale resolution, can be averaged over multiple summertime observations to provide 5-10 km resolution, sufficient for accurately assessing current inventories.