Ozone over St. Petersburg: Comparison of experimental data and numerical simulation

Abstract

Comparison of numerically simulated ozone fields with various experimental data makes it possible to evaluate the quality of models for their use in reliable prediction of variations of the ozone layer. We compared the satellite (SBUV) and ground-based (IFS Bruker 125HR) measurements of ozone content in two atmospheric layers (0–25 and 25–60 km) with the numerical simulation data (obtained using the lowand middle-atmosphere models) over St. Petersburg for the period between 2011 and 2014, namely, the daily and monthly average values of ozone content for 3.5 years (June 2011–December 2014). In general, model describes the experimental ozone content with good or satisfactory accuracy in the two layers. Nevertheless, some systematic differences are found out between the satellite and ground-based data and the results of simulation. In the autumn–winter period, the model usually overestimates the ozone column in the 0–25 km layer as compared to the satellite measurements, and underestimates it in the 25–60 km layer. The same features are observed for daily and monthly average values. In some cases, the model shows strong and high-frequency oscillations in the ozone content, which are not observed in the measurements.