Inverse modeling of emissions and their time profiles

The paper presents a version of the 4DVar method, capable of optimizing diurnal time profiles of emissions. It is a generalization of existing inverse methods that optimize emission daily totals. The core of the method is formed by the CMAQ adjoint model with SAPRC99 mechanism. Measurements from both ground–level stations (NO2 and O3), and satellites (retrieved columns of NO2 from GOME2 and OMI and the lowest layer of O3 retrieved from IASI) have been used as a data source for the inverse modeling procedure. The method can be used for detection of bias or errors in the emission model. It also can assist in development of data–driven emission model with location–specific time profiles of emissions. Different aspects of the method are illustrated on simulation experiments. Forecasting performance of the optimized model is evaluated for O3 and NO2 concentrations.