We developed a twin dispersion model based on AERMOD (Cimorelli et al., 2004) which simulates surface concentrations of NO2 at street-level resolution and column concentrations of NOx matching the satellite footprints. The CAMS regional data is used to provide background concentrations and to approximate NOx chemistry. Depending mainly on open datasets, the method is generic and versatile, meaning that it is applicable to other cities relatively easily. The twin model successfully creates a link between ground observations and space observations, see Figure 1, capable of recreating column concentration patterns observed from space.

Figure 1  Comparison of observed columns and simulated columns. The error bars in the scatter plot indicate the retrieval error.

The emission factors used in the model are daily calibrated to best match the observations. Calibration can be done with surface observations or with space observations, providing two different ways of estimating the urban emissions of the area. The yearly (2019) estimates of Madrid emissions look realistic, see Figure 2.

Figure 2  Comparison of space and surface derived NOx emissions for Madrid with the DECSO and CAMS inventories. Note that the estimates depend on the assumed lifetime of NOx.

Validation results show improved simulation of surface concentrations when emissions are calibrated daily. In Figure 3, we compare hourly NO2 concentrations at a street station with our uncalibrated dispersion model, and after calibration with either surface or space observations.

Figure 3 Validation of three different calibration scenarios against hourly NO2 observations at Madrid street station Castellana, for the first week of March 2019.


Cimorelli, A. J., Perry, S. G., Venkatram, A., Weil, J. C., Paine, R. J., Wilson, R. B., Lee, R. F., Peters, W. D., and Brode, R. W.: AERMOD: A dispersion model for industrial source applications Part I: General model formulation and boundary layer characterization, J. Appl. Meteor., 44, 682–693, 2004.