The observing requirements for the prediction of ozone

We present a prototype air quality forecasting model and use it to determine observing requirements for GEO-CAPE and to make recommendations for inverse emission estimation. This prototype framework comprises a combination of sequential 4D-var data assimilation and an analytical modelling framework used to determine assimilation solution diagnostics such as posteriori ozone prediction error and posteriori emission parameter uncertainty. We assess the optimal time of observation, the required precision of the observations and the diversity of observed chemical species required to provide acceptable forecast estimates of ozone concentrations. As the photochemical regime changes depending on NOx and VOC concentrations different observing strategies become favourable. Both ozone or formaldehyde observations can significantly improve ozone predictive skill under VOC limited conditions, and a combination of NO2 and ozone observations improve ozone predictive skill the most under strongly NOx limited conditions. These results are also relevant for emission inversion studies since the tools we present are mechanically very similar to the state of the art in this field. We show that combinations of ozone and formaldehyde reduce VOC emission uncertainties by up to 16% over observing scenarios using only formaldehyde under VOC limited conditions, and combinations of ozone and NO2 observations reduce NOx emission uncertainties by up to a factor 3 over scenarios utilizing only observations of NO2 whilst under NOx limited conditions.